AndrewS.Grove

HIGHOUTPUTMANAGEMENT

AndrewS.GroveemigratedtotheUnitedStatesfromHungaryin1956.HeparticipatedinthefoundingofIntel,andbecameitspresidentin1979andchiefexecutiveofficerin1987.HewaschosenasTimemagazine’sManoftheYearin1997.In1998,hesteppeddownasCEOofIntel,andretired as chairman of the board in 2004.Grove taught at the StanfordUniversityGraduateSchoolofBusiness for twenty-fouryears.He livesintheSanFranciscoBayArea.

Acknowledgments

Theideasinthisbookaretheresultofacollectiveeffort—mycollaborationwithmany,many Intelmanagers over the years. I am very grateful to all of them,because I learned everything I know about how to manage from them. I amespecially grateful to Gordon Moore, one of the founders of Intel, whorecognizedthebuddingmanagerundermyengineer’sskinlongbeforeIhadanyinklingmyself.Thanks are also due to a group of the company’s middle managers who

cheerfully accepted the role of guinea pig, who suffered through my firstattempts to articulate these ideas, and who also generously providedmewithexperiences from their daily lives as managers. I’ve used their examples toillustratecertainpointsinthebook.ThesemanagersareacknowledgedbynameintheNotes.I owe special thanks to Grant Ujifusa, my Random House editor, who

tirelessly hammered away at the rough edges of my ideas and prose andtranslated the latter into English—from the original engineeringese; to PamJohnson,whoranthevariousrevisionsthroughthewordprocessor;andmostofalltoCharleneKing,myassistant,whonotonlyhelpedtopullthewholeprojecttogether, from capturing class discussions to gathering illustrations, but alsomade sure that I didmywork of running Intel even as I was busily splittinginfinitives.

Contents

CoverAbouttheAuthorTitlePageCopyrightAcknowledgments

IntroductionForeword

PARTITHEBREAKFASTFACTORY1TheBasicsofProduction:DeliveringaBreakfast(oraCollegeGraduate,

oraCompiler,oraConvictedCriminal…)2ManagingtheBreakfastFactory

PARTIIMANAGEMENTISATEAMGAME3ManagerialLeverage4Meetings—TheMediumofManagerialWork5Decisions,Decisions6Planning:Today’sActionsforTomorrow’sOutput

PARTIIITEAMOFTEAMS7TheBreakfastFactoryGoesNational8HybridOrganizations9DualReporting10ModesofControl

PARTIVTHEPLAYERS11TheSportsAnalogy12Task-RelevantMaturity

13PerformanceAppraisal:ManagerasJudgeandJury14TwoDifficultTasks15CompensationasTask-RelevantFeedback16WhyTrainingIstheBoss’sJob

OneMoreThing…Notes

Introduction

I.Whathappenedafter1983

Iwrotethisbookin1983.Itwastheresultoftwentyyearsofmanagerialworkduring which I learned a variety of ways to make things take place moreeffectively.What I learnedwere thebasicsofmanagerialwork,particularlyastheypertainedtomiddlemanagers.Morethanadecadehaspassedsince,butIfindthatmostofthethingsthatwereusefulthenarestillusefulnow;thebasicsofmanagementremainlargelyunaffected.However, two critical events took place in the 1980s that altered the

environmentinwhichwemanagersdoourwork—andthismademerealizethatan updated Introduction to this book was necessary. Those events were theJapanesememoryonslaughtande-mail.Letmeexplaintheirimplications.By the mid-eighties, the Japanese producers of Dynamic Random Access

Memories,orDRAMsfor short—themostpopularcomputermemorydevices,usedincomputersofallkinds—hadperfectedtheirtechnologicalcapabilityandhoned their manufacturing prowess to the extent that they could take on theAmericanproducers(whohadpioneeredthemarketandtotallydominateditforthe first fifteen years of its existence). The mid-eighties were also when thepersonal computer revolution took place. And because personal computersrequirealotofmemory,theJapaneseDRAMjuggernauthadareadymarketforitsproductscenteredintheUnitedStates.Everythingwasprimedforanattack.Intel, where I work, was one of the companies that got caught up in this

assault.Infact,IntelwasoneoftheearlyproducersofDRAMs.Morethanthat,initsearliestyears,wehadpracticallythewholemarkettoourselves.However,bythemid-eighties,competitionbothfromtheUnitedStatesand, increasingly,fromJapanesemanufacturerswhittleddownourshareofthemarket.Undertheferociousattackofaggressivelypriced,high-qualityJapaneseDRAMs,wewereforced to retreat and cut prices to a levelwhere being in theDRAMbusiness

brought us major losses. Ultimately, the losses forced us to do somethingextraordinarily difficult: to back out of the business that the company wasfoundedupon,andtofocusonanotherbusinessthatwethoughtwewerebestat—themicroprocessorbusiness.While this adjustment sounds quite logical and straightforward in theory, in

reality its implementation required us to move and redeploy a lot of ouremployees, let someof themgo,andshutteranumberof factories.Wedidallthis because under this strong attack, we learned that we must lead with ourstrength.Beingsecondbestinatoughenvironmentisjustnotgoodenough.Ultimately, we—Intel and the U.S. semiconductor industry—prevailed over

the onslaught of the Japanesemanufacturers. Intel grew to become the largestsemiconductor manufacturer in the world, and U.S. manufacturers recentlysurpassed their Japanese counterparts overall. Nevertheless, in retrospect it’sclear that this assault was just one wave of a much larger tide—the tide ofglobalization.Globalization simply means that business knows no national boundaries.

Capital andwork—yourwork andyour counterparts’work—cango anywhereonearthanddoajob.Someofusarefortunatetoberesidentsofacountry, theUnitedStates, that

enjoys one of the highest standards of living. TheU.S.market for goods andservices is the largest in the world. And until recently, it has been easier tosupplythatmarketfrominsidetheUnitedStatesthanfromabroad.Today, many markets outside the United States are growing faster than

marketsinsidetheU.S.Andthedomesticmarketcanbesuppliedfromanywherein theworld.Forexample, I recentlyboughtaGore-tex jacket fromPatagonia(theclothingmanufacturer,not theregion inSouthAmerica),andIsawthat itwasmadeinChina:Americanbrand,Americantechnology(thehigh-techfabricwasinventedandmadeintheUnitedStates),andassembledtothespecificationsofthereseller(Patagonia)inaforeigncountry.Theconsequenceof all this isvery simple. If theworldoperates asonebig

market,everyemployeewillcompetewitheverypersonanywhereintheworldwhoiscapableofdoingthesamejob.Therearealotofthem,andmanyofthemareveryhungry.Another consequence also follows. When products and services become

largelyindistinguishablefromeachother,allthereisbythewayofcompetitiveadvantage is time. And that’s where the second critical development of the

eightiescomesin—e-mail.JustastheJapaneseDRAMattackwasthefirstwaveofamuchgreatertide,

e-mail is also the firstmanifestation of a revolution in how information flowsandhowitismanaged.The informed use of e-mail—short for computer-to-computer electronic

messaging—results in two fundamentally simple but startling implications. Itturns days intominutes, and the originator of amessage can reach dozens ormoreofhisorherco-workerswiththesameeffortittakestoreachjustone.Asaresult,ifyourorganizationusese-mail,alotmorepeopleknowwhat’sgoingoninyourbusinessthandidbefore,andtheyknowitalotfasterthantheyusedto.Letmeinterjectabitofirony.Backintheeighties,whentheJapaneseseemed

invincible, one explanation advanced for their ability to act quickly anddecisively was the way Japanese offices were set up. In a Japanese office, amanager and his subordinates all sit around a big long table. Peoplework ontheirownassignmentsbutwhentheyneedtoexchangeinformation,everybodytheyworkwith sitswithin speaking distance, right around the same table. Soinformation is exchanged in minutes and everybody can be reached with thesameeffort.Asaresult,becauseoftheeasewithwhichJapaneseofficeworkerscommunicate,theyhave,infact,beenslowtoembraceelectronicmail.But now the pendulum is swinging in the other direction. As businesses

becomemorewidelyspreadoutaroundtheglobeandastimebecomesthekeycompetitive weapon, American organizations are often better positioned thantheirJapanesecounterparts.Why?BecausethesameeaseofcommunicationthatprevailedbynaturalmeansintheJapaneseofficenoweffectivelytravelsaroundtheworldthroughelectronicmeans.And e-mail is only the first wave. Everything today is going to a digital

format:sound,photos,movies,books,financialservices.Andeverythingthat’sdigitalcanbeshippedaroundtheworldjustasfastasitcanbeshippeddownthehallatyourworkplace.Hereisaninterestingillustrationoftheconsequenceofsuchacapability.Iam

told that thepostofficesorts90percentofall lettersautomatically.For the10percentthemachinescan’tdecipher,ahumanreadertypestheaddressesintoamachine.Recently,tolowerthecostofthiswork,thepostalservicetriedanewsystem. A machine takes a digital photograph of the illegible envelopes,instantly ships the digital image to a lower-labor-cost region where someonereadsandkeysintheaddressfromthedigitalimage,thenelectronicallyshipsthe

addressbacktotheregionalpostalcenter.This is thebeginningofa trendthatwillbecomeall-encompassinginthenexttwenty-fiveyears.Simply put, the information revolution does away with hiding places

anywhere,inanylineofwork.Sothequestionsare:Whatarebusinessestodoandwhataremanagersthemselvestodo?

II.Operatinginthenewenvironment

Let’s back off for amoment and consider whom this book is aimed at. I amespecially eager to reach the middle manager, the usually forgotten man orwomanofanyorganization.Thefirst-linesupervisoron theshopfloorand thechiefexecutiveofficerofacompanyarebothwellappreciated.You’llfindmanycourses designed to teach the former the fundamentals of his work, whilepracticallyallofourleadingbusinessschoolsaresetuptoturnoutthelatter.Butbetween the two is a large group of people—the middle managers, whosupervise the shop-floor foremen, orwhowork as engineers, accountants, andsalesrepresentatives.Middlemanagersarethemuscleandboneofeverysizableorganization, no matter how loose or “flattened” the hierarchy, but they arelargelyignoreddespitetheirimmenseimportancetooursocietyandeconomy.Middlemanagers are not confined to big corporations. In fact, they can be

found inalmostanybusinessoperation. Ifyourunasmall taxdepartmentatalaw firm, you are a middle manager. The same is true if you are a schoolprincipal,anownerofadistributorship,orasmall-towninsuranceagent.Whenpeoplefromeachoftheseenterprisesreadtheoriginalmanuscriptofthisbook,their reactions confirmed what I suspected: the managerial ideas that weredevelopedatIntelasitgrewfromaverysmalltoaverylargeorganizationwerebroadlyapplicable.Anothergroupshouldalsobeincludedamongmiddlemanagers—peoplewho

may not supervise anyone directly but who even without strict organizationalauthorityaffectandinfluencetheworkofothers.Theseknow-howmanagersaresources of knowledge, skills, and understanding to people around them in anorganization. They are specialists and experts of some sort who act asconsultantstoothermembersoftheorganization;theyare,ineffect,nodesinalooselydefinednetworkofinformation.Teachers,marketresearchers,computermavens,andtrafficengineersshapetheworkofothersthroughtheirknow-howjustasmuchasormorethanthetraditionalmanagerusingsupervisoryauthority.Thusaknow-howmanagercanlegitimatelybecalledamiddlemanager.Infact,

as our world becomes evermore information-and service-oriented, know-howmanagerswillacquiregreaterimportanceasmembersofmiddlemanagement.Inshort,know-howmanagersshouldalsoreadon.Whether you are a know-how manager or a traditional manager, your

companyhasnochoicebuttooperateinanenvironmentshapedbytheforcesofglobalization and the information revolution. Companies today basically havetwo choices: Adapt or die. Some have died in front of our eyes; others arestrugglingwiththeadaptation.Astheystruggle,themethodsofdoingbusinessthatworked verywell for them for decades are becoming history.Companiesthathavehadgenerationsofemployeesgrowingupunderano-layoffpolicyarenowdumpingtenthousandpeopleatatimeontothestreet.Unfortunately,that’sallpartoftheprocessofadaptation.Allmanagersinsuchcompaniesneedtoadapttothenewenvironment.What

aretherulesofthenewenvironment?First,everythinghappensfaster.Second,anythingthatcanbedonewillbedone,ifnotbyyou,thenbysomeoneelse.Lettherebenomisunderstanding:Thesechangesleadtoalesskind,lessgentle,andlesspredictableworkplace.Again, as a manager in such a workplace, you need to develop a higher

tolerance for disorder.Now, you should still not accept disorder. In fact, youshoulddoyourbest todrivewhat’saroundyou toorder.Thebreakfastfactorymetaphorofthisbook—theideathatyoushouldrunyourmanagerialprocesseslikeawell-oiled factory—iseverybitasmuch the idealnowas itwaswhen Iwrote this book. But you as a manager need to be mentally and emotionallyready to be tossed into the turbulence generated by amega-merger that takesplaceinyourindustry—perhapsinthiscountry,perhapsontheothersideoftheglobe.Youshouldbepreparedfortheshockwavesengenderedbyabrand-newtechniquepioneeredbysomeoneyouhadneverevenheardofbefore.Youneedtotrytodotheimpossible,toanticipatetheunexpected.Andwhen

theunexpectedhappens,youshoulddoubleyoureffortstomakeorderfromthedisorder it creates inyour life.Themotto I’madvocating is“Let chaos reign,thenreininchaos.”Now,I’msurethatatvarioustimesyouwilltakeexceptiontowhatyouread

inthisbook.“ThismaybefineatIntel,”youwillsay,“butitwouldneverflyatPDQ,whereIwork.NothingdoesuntiltheOldManhimselfdecreesit.Shortofa palace revolution, I can’t use anythingyou recommend.”Letme assure youthatyouwillbeabletousemostofwhatIsay.Asamiddlemanager,ofanysort,

youareineffectachiefexecutiveofanorganizationyourself.Don’twaitfortheprinciples and practices you find appealing to be imposed from the top.As amicro CEO, you can improve your own and your group’s performance andproductivity,whetherornottherestofthecompanyfollowssuit.Thisbookcontainsthreebasicideas.Thefirstisanoutput-orientedapproach

to management. That is to say, we apply some of the principles and thediscipline of themost output-oriented of endeavors—manufacturing—to otherformsofbusinessenterprise,includingmostemphaticallytheworkofmanagers.ConsiderIntel,whichisatruemanufacturingandproductioncompany,makinghighlycomplexsiliconchipsaswellascomputer-likeproductsbuiltfromthem.Our company now has over thirty thousand employees. Of these, about 25percent actuallywork tomake the products.Another 25 percent help them astheysupervisethepersonnel,maintainthemachines,andengineerandimprovethemanufacturing process. Another 25 percent work in administration, wherethey schedule production, keeppersonnel records, sendbills to our customers,andpayour suppliers. Finally, the remaining25percent designnewproducts,takethemtothemarketplace,sellthem,andservicethemafterthesale.As we founded, organized, and managed Intel, we found that all our

employees “produce” in some sense—some make chips, others prepare bills,whilestillotherscreatesoftwaredesignsoradvertisingcopy.Wealsofoundthatwhenwe approached anywork done at Intel with this basic understanding inmind, the principles and discipline of production gave us a systematicway ofmanaging it,muchas the languageandconceptsof financecreatedacommonapproachtoevaluatingandmanaginginvestmentsofanysort.Thesecondideaisthattheworkofabusiness,ofagovernmentbureacracy,of

most formsofhumanactivity, is somethingpursuednotby individualsbutbyteams. This idea is summed up inwhat I regard as the singlemost importantsentence of this book: The output of a manager is the output of theorganizationalunitsunderhisorhersupervisionorinfluence.Thequestionthenbecomes, what can managers do to increase the output of their teams? Putanother way, what specifically should they be doing during the day when avirtuallylimitlessnumberofpossibletaskscallsfortheirattention?Togiveyouaway toanswer thequestion, I introduce theconceptofmanagerial leverage,whichmeasurestheimpactofwhatmanagersdotoincreasetheoutputoftheirteams. Highmanagerial productivity, I argue, depends largely on choosing toperformtasksthatpossesshighleverage.

A team will perform well only if peak performance is elicited from theindividualsinit.Thisisthethirdideaofthebook.Canbusinessusewhateveritis thatmotivates an athlete toput out his “personal best” consistently? I thinkbusiness can, which is why I examine the sports analogy and the role ofsomething called task-relevant feedback to get and to sustain a high level ofperformancefromthemembersofabusinessteam.Wemustrecognizethatnoamountofformalplanningcananticipatechanges

such as globalization and the information revolutionwe’ve referred to above.Doesthatmeanthatyoushouldn’tplan?Notatall.Youneedtoplanthewayafiredepartmentplans.Itcannotanticipatewherethenextfirewillbe,soithastoshape an energetic and efficient team that is capable of responding to theunanticipatedaswellastoanyordinaryevent.Second, a responsive company should have fewer levels ofmanagers. This

concept is easier to apply today as electronic mail can carry information toanyone in the organization. One basic role of management—the role ofdisseminatinginformation—isnolongerasimportantamanagerialfunctionasitwasinthepast.With fewer levels in today’s organization, each manager will have larger

numbersofemployeesreportingtohimthanwasthecasetenyearsago.Oneofthe fundamental tenets of Intel’s managerial philosophy is the one-on-onemeetingbetweena supervisoranda subordinate. Itsmainpurposesaremutualeducationandtheexchangeofinformation.Bytalkingaboutspecificproblemsandsituations, thesupervisor teaches thesubordinatehisskillsandknow-how,and suggests ways to approach things. At the same time, the subordinateprovides the supervisor with detailed information about what he is doing andwhatheisconcernedabout.Obviously,one-on-onestaketime,bothinpreparingfor themand in actuallyholding them—time that today’sbusiermanagermaynothave.Are one-on-one meetings still needed? Absolutely. Can you have them as

often with ten direct reports as with five? No. Do you need to? No again,becauseforthemostpart,theseemployeesaremoreawareofwhat’sgoingonintheir business through their computer network than their counterparts were adecade ago; theyno longer rely onyou to bring themup to date.Nor doyouneedtorelyonone-on-onestocatchupwithwhatyoursubordinateshavefoundin their lab, factory, or sales region; you’ve already read about thosedevelopmentsonyourcomputerscreen,minutesaftertheychosetoinformyou.

NowconsidertheproverbialJapaneseemployeessittingaroundthetableinaJapanese office. They don’t need to get together with their supervisors to bebroughtuptodate.Theymaystillneedtoleavethetableandhaveatête-a-têtewiththemtodiscusstheirconcernsortobringupissuesthey’reuncomfortablewith,buta lotof thepurposesof theone-on-onesare takencareofminutebyminute.It’snodifferentwhenyouandyoursubordinatesareworkingaroundtheelectronicequivalentofthistable.So,yes,youstillneedone-on-ones.Butyouneed them for fewerof thepurposes I envisionedwhen Ioriginallywrote thisbook. Therefore, you can deal with more employees less frequently and inmeetingsofshorterduration.

III.Managingyourowncareer

Butwhataboutmanagerswhoare,afterall,employeesthemselves?I recently readanarticle saying thatmiddle-agedmenare twiceas likely to

lose their jobs today than they were in 1980, fifteen years ago. This trend isgoingtoincreaseintheyearsahead.Asageneralrule,youhavetoacceptthatnomatterwhereyouwork,youare

notanemployee—youareinabusinesswithoneemployee:yourself.Youareincompetitionwithmillionsofsimilarbusinesses.Therearemillionsofothersallover theworld,pickingup thepace, capableofdoing the samework thatyoucan do and perhaps more eager to do it. Now, you may be tempted to lookaroundyourworkplaceandpointtoyourfellowworkersasrivals,buttheyarenot.Theyareoutnumbered—athousandtoone,onehundredthousandtoone,amillion toone—bypeoplewhowork fororganizations thatcompetewithyourfirm. So if you want to work and continue to work, you must continuallydedicateyourselftoretainingyourindividualcompetitiveadvantage.Inasloworno-growthenvironment, thereisanotherfactorthatyouhaveto

contendwithaswell:ambitiousjunioremployeeswhodesiretomoveupwardintheorganization.Theymayverywellbereadytodosobutcan’tbecauseyou’rein theway. Sooner or later, your bosswill inevitably have tomake a choice:whethertoholdontoyou,whoisdoingagoodjobbutisinthewayofanotherperson.Theresponsibilitytoavoidsuchsituationsisyours.The recipe for success for the generation of managers who worked in the

sixties, seventies, andmuch of the eightieswas to join stable and enlightenedcompaniesandhelp themdowell; thesecompanies in turnwould rewardsuchmanagerswithacareer.Obviously,thatisnolongerthecase.

Thepointis,theclichésofglobalizationandtheinformationrevolutionhaverealmeaning—potentially deadlymeaning—for your career. The sad news is,nobodyowesyouacareer.Youownitasasoleproprietor.Youmustcompetewithmillions of individuals every day, and every day youmust enhance yourvalue,honeyourcompetitiveadvantage,learn,adapt,getoutoftheway,movefromjobtojob,evenfromindustrytoindustryifyoumustandretrenchifyouneedtodosoinorder tostartagain.Thekeytaskis tomanageyourcareersothatyoudonotbecomeacasualty.Icanofferyounosurefireformula.Buthereareafewquestionstoponder:1.Areyouadding realvalueormerelypassing informationalong?Howdo

youaddmorevalue?Bycontinuallylookingforwaystomakethingstrulybetterinyourdepartment.Youareamanager.Thecentralthoughtofmybookisthattheoutputofamanageristheoutputofhisorganization.Inprinciple,everyhourofyourdayshouldbespent increasingtheoutputor thevalueof theoutputofthepeoplewhomyou’reresponsiblefor.2. Are you plugged into what’s happening around you? And that includes

what’s happening inside your company as well as inside your industry as awhole. Or do you wait for a supervisor or others to interpret whatever ishappening?Areyouanodeconnectedtoanetworkofplugged-inpeopleorareyoufloatingbyyourself?3. Are you trying new ideas, new techniques, and new technologies, and I

meanpersonally trying them,not just readingabout them?Orareyouwaitingforotherstofigureouthowtheycanre-engineeryourworkplace—andyououtofthatworkplace?Iamanengineerbytrainingandamanagerofahigh-technologycompanyby

profession. As amanager, I am also amember of the group of individuals—manymillionsstrongintheUnitedStatesalone—thatholdsthekeytoincreasedproductivity: generatingmore and better goods and services to meet people’sneeds. I am an optimist and believe our potential to increase our wealth hashardlybeentapped.ButIalsothinkthatpeopledonotalwaysfaceuptothechangestheyhaveto

dealwith,soat timesI feel Ihave tobea realist, too.Youcan’tbeoptimisticabout the future until you have survived the crucible of change. The key tosurvivalistolearntoaddmorevalue—andthatultimatelyiswhatthisbookisabout.FrommyownexperienceatIntel,Istronglybelievethatapplyingthemethods

ofproduction,exercisingmanagerial leverage,andelicitinganathlete’sdesireforpeak performance can help nearly everyone—lawyers, teachers, engineers,supervisors,evenbookeditors;inshort,middlemanagersofallkinds—toworkmoreproductively.So,letusproceedbytakingafieldtriptoafactory….

AndrewS.GroveApril1995

ForewordtotheVintageBooksEdition

IfirstreadHighOutputManagementin1995.Inthosedays,therewerenoblogsor TED Talks teaching us about entrepreneurship. In fact, there was almostnothing of use written for people like me who aspired to build and run acompany.Against this backdrop,High Output Management had an almost legendary

status. All the best managers knew about it. The top venture capitalists gavecopies of it to their entrepreneurs, and aspiring leaders in Silicon Valleydevoured its contents. It amazed all of us that theCEOof Intel had taken thetimetoteachustheessentialskillofentrepreneurship:howtomanage.ThiswasnosmallthingbecauseIntelwasknownasthebestcompanyinthe

technology industry. Ithadpulledoff thegreatest transformation in thehistoryof the business: moving from the memory business to microprocessors morethanadecadeafteritsfounding.Beyondthat,Intelranwithlegendaryprecision,which gave it the ability to make multibillion dollar investments with highconfidence.Ifyouwantedtohireagreatoperationalmanager,thenIntelwastheplace togo—butgood luckgettingone to leave thebest-managedcompany inSiliconValley.Andyhimselfwas a legendary figure.Hehadgrownup Jewish inHungary

during a timewhen the countrywas occupied by theNazis and, later, by theSovietCommunists.ArrivinginNewYork,hespokenoEnglishandhadalmostnomoney.HeenrolledhimselfattheCityCollegeofNewYork,overcamehislanguage deficiency, and went on to get a PhD from UC Berkeley. Thisnonnative English speaker would then write an important textbook onsemiconductors in English while working at Fairchild Semiconductor. As aresult,hewasconsideredascientificpioneerevenbeforehelpingtolaunchIntelin 1968, building it into the seminal technology companyof the era.Later, in1997,Timemagazinewould recognize his nearly impossible accomplishmentsandnamehimManoftheYear.

This is inpartwhatmadeHighOutputManagement soextraordinary.AndyGrove,whobuilthimselffromnothingtorunIntel,stoppedwhathewasdoingto teach us hismagic.And not through some ghostwriter either—Andywrotethisbookhimself.Whatanincrediblegift.WhenIfinallygotmyhandsonthebook,thepaperbackcovertookmeaback.

The 1995 version featured a picture ofAndyGrove standing next to the Intelsign.UnlikeeveryotherCEOphotothatIhadeverseen,Andywasnotwearingadesignersuit.Hedidnothaveperfectlycombedhair,andhedidnotstrikeanarms-foldedpowerpose.No,AndyGrovewasdressedforworkrightdown tohiskeycardhangingfromhisbelt.Ididadoubletake.“Wasthatakeycard?Hedidn’tremovehiskeycardforthebook’scoverphoto?”Inretrospect,thecoverwasperfect.Asyouwillseewhenyoureadthisbook,

AndyGrovewasallsubstance.Hedidnothavetimeforprettyphotoshootsorself-promotion.Hewrotethebookforus,butifwehadtobesoldonitbyhowhe looked in the photo, then thatwould be our loss. The time that he did notspendstylingfancyphotos,heputintowritingthebook.Hedidnotjustgiveusthe lessons; he articulated them in a way that connected both logically andemotionally.Wewouldcometounderstandhimandfeelwhathemeantinourcore.I immediately got a jolt of this stylewith the title of the very first chapter:

“TheBasicsofProduction:DeliveringaBreakfast(oraCollegeGraduate,oraCompiler, or a Convicted Criminal…).” Okay, I am interested. What doesmakingasoft-boiledegghavetodowithhowmanyprisonswebuild?It turnsoutquiteabit.HighOutputManagementopensbyteachingustheimportanceofpropersystemdesignevenwhenwearedealingwithasystemofhumanbeings—especiallywhenwearedealingwithasystemofhumanbeings.Andythenshowsushowyoucanusethesesameprinciplestounderstandhow

society shouldoperate. Itdoesn’t accomplishanything todeclare thatweneedmorekidsgoingtocollegethantojailanddemandthatwebuildmoreschoolsthanjails. Infact, it’scounterproductive. Identifyingcomplexsystemproblemsisonething.Solvingthemissomethingelseentirely,andAndylaysoutthetoolstodojustthat.Over the years, I have come to considerHigh Output Management a true

masterpiece, and there are at least three core aspects to its genius. First, in aslittleasonesentence,itlucidlyexplainsconceptsthatrequireentirebooksfromlesserwriters.Second,itconsistentlyuncoversbrand-newmanagementideasor

finds new insights into old standards. Finally, whilemostmanagement booksattempttoteachbasiccompetency,HighOutputManagementteachesthereaderhowtobegreat.Andyintroducesmanagementwiththisclassicequation:

A manager’s output = the output of his organization + the output of theneighboringorganizationsunderhisinfluence.

On the surface it may seem simple, but he clarifies the essential differencebetween a manager and an individual contributor. A manager’s skills andknowledge are only valuable if she uses them to get more leverage from herpeople.So,Ms.Manager, youknowmore about our product’s viral loop thananyoneinthecompany?That’sworthexactlynothingunlessyoucaneffectivelytransfer that knowledge to the rest of the organization. That’s what being amanagerisabout.It’snotabouthowsmartyouareorhowwellyouknowyourbusiness;it’sabouthowthattranslatestotheteam’sperformanceandoutput.As a means to obtain this leverage, a manager must understand, as Andy

writes:“Whenapersonisnotdoinghisjob,therecanonlybetworeasonsforit.The person either can’t do it or won’t do it; he is either not capable or notmotivated.”Thisinsightenablesamanagertodramaticallyfocusherefforts.Allyoucandotoimprovetheoutputofanemployeeismotivateandtrain.Thereisnothingelse.Ashedescribestheplanningprocess,Andysumsuphisessentialpointwith

this eloquent nugget ofwisdom: “I have seen far toomany peoplewho uponrecognizingtoday’sgaptryveryhardtodeterminewhatdecisionhastobemadeto close it. But today’s gap represents a failure of planning sometime in thepast.”Hopefully,thevalueofthisshortinsightisnotlostontheyoungreader.Ifyouonlyunderstandonethingaboutbuildingproducts,youmustunderstandthatenergyputinearlyintheprocesspaysofftenfoldandenergyputinattheendoftheprogrampaysoffnegativetenfold.Thebookhasanentiresectiondedicatedtoanoftenneglected,butcritically

importantmanagementtool:meetings.Andymakesusseethisoldestofbusinessprinciples in a new light.He teachesmeetings from first principles, beginningwith how to conduct a one-on-one. It seems incredible that the CEO of Intelwouldtakethetimetoexplainhowtohaveaone-on-one.Why is he doing this? It turns out that the one-on-one is not only a

fundamentalelementinthemanager/employeerelationship,butperhapsthebest

sourcefororganizationalknowledgethatamanagercanget.Inmyexperience,managers who don’t have one-on-ones understand very little about what’shappeningintheirorganizations.It is byunderstanding the simple things thatAndygoes deep.For example,

whenpeoplevisit today’s technology companies theyoften remark about howcasualtheenvironmentsare,butwithverylittleexplanationaboutwhytheyarethatway.Infact,manyCEOsdonotunderstandwhyastheysimplyfollowthetrend,butAndyexplainsitperfectly:

A journalist puzzled by ourmanagement style once askedme, “Mr.Grove,isn’t your company’s emphasis on visible signs of egalitarianism such asinformal dress, partitions instead of offices…just somuch affectation?”Myanswerwasthatthisisnotaffectation,butamatterofsurvival.Inourbusinesswe have tomix knowledge-power peoplewith position-power people daily,andtogethertheymakedecisionsthatcouldaffectusforyearstocome.

Inthisfashion,thebookquicklygetstotheheartofcomplexissues.Itraisesand deals with the stickiest management issues. Andy asks the question ofwhetheryoushouldbefriendswiththepeopleyoumanage:

Everyonemustdecideforhimselfwhatisprofessionalandappropriatehere.Atestmight be to imagine yourself delivering a tough performance review toyourfriend.Doyoucringeatthethought?Ifso,don’tmakefriendsatwork.Ifyour stomach remains unaffected, you are likely to be someone whosepersonalrelationshipswillstrengthenworkrelationships.

Bybreakingdowntheprocess,hemakeshardthingsmanageable.Ultimately, the power ofHighOutputManagement is that it creates expert

ratherthanmerelycompetentmanagers.Agreatexampleof this is thesectionontask-relevantmaturity.Thispartof

thebookbecameverypersonalformeasittaughtmehowtoformulatethemostusefulmanagementquestionthatIuseininterviews:“Isitbettertobeahands-onorhands-offmanager?”It seems like a simple enough question, but it sorts out the 95 percent of

managerswhonever thinkdeeplyabout theircraft fromthe5percentwhodo.Theanswer,asAndyexplains,isthatitdepends.Specifically,itdependsontheemployee. If the employee is immature in the task, then hands-on training is

essential.Iftheemployeeismoremature,thenadelegateapproachiswarranted.Andy presents a great example of this: “The subordinate did poor work. Myassociate’sreaction:‘Hehastomakehisownmistakes.That’showhelearns!’Theproblemwiththisisthatthesubordinate’stuitionispaidbyhiscustomers.Andthatisabsolutelywrong.”PerhapsthechapterthatbestreflectsAndyGroveisthelast,“WhyTrainingIs

theBoss’sJob.”Often,peoplewhomanageintheso-calledknowledgeeconomybelieve their employees are so smart that they need no training at all. Andybrilliantly corrects this notion by explaining why as customers we areflabbergasted when we encounter employees who are insufficiently trained atrelativelysimpletaskssuchastakingrestaurantreservations.Hethenchallengesus to imagine how furious customers of complex jobswill be if an employeeisn’t properly trained. Finally, he reiterates his thesis that there are only twoways in which a manager can impact an employee’s output: motivation andtraining.Ifyouarenottraining,thenyouarebasicallyneglectinghalfthejob.Throughout the chapter, the reader feels Andy’s intense passion toward

training and teaching, because in the end—more than anything else—he is ateacher…intheverybestsenseoftheword.ManyyearsafterreadingHighOutputManagement,ImetAndyforthefirst

time. Upon seeing him, I was so excited that I immediately blurted out howmuchIlovedthebook.InclassicAndyGrovestyle,heshotback:“Why?”Ididnotexpectthat.Ithoughtthathewouldsay,“Thankyou”or“Iappreciatethat,”but not “Why?” But that was Andy. He was always teaching and alwaysexpectingmorefromeverystudent.Caught completely off guard, I scrambled for the reason and came across a

goodone:“EveryothermanagementbookthatI’vereadexplainsthetrivial,butyoursgetstotherealissues.”Uponhearingthat,themasterteachersoftenedandrepliedwithapricelessstory:

It’s funny that you say that aboutmanagement books. I recently ran out ofspaceonmybookshelfathome,soIwasfacedwithachoice.Ieitherhadtothrow away some books or buy a bigger house. Well, that was an easydecision, but which books to throw out? Then I thought, the managementbooks!ButIhadaproblem.NearlyeverymanagementbookthatI’dreceivedwassent tomebytheauthorandwasautographedwithakindinscription.Ifeltbadlyaboutthrowingawayall thosenicenotes.So,Iwentthrougheachbookand toreout the inscriptionpage then threwaway thebook.Sonow I

havealargestackofpagesofnicenotestomeandplentyofspaceforgoodbooks.

IhavenevermetanyoneotherthanAndyGrovewhowouldhaveastorylikethat. He uniquely balances the highest standards for clear thinking andperformancewithanundyingbelief in theunderlyingperson.Whoelsewouldrequiresohighabarforwritingthatyouhadtobegoodenoughtofitonhisonebookshelfandstillbesotouchedbythefact thatyouwantedhimtoreadyourworkthathewouldsavethepagethatyouinscribed?Later,in2001,ImetwithAndyagainandIaskedhimaboutarecentrunof

CEOsmissing theirnumbersdespitehaving told investors that theirbusinesseswerestrong.ThebubblehadburstforthefirstwaveofInternetcompaniesnearlya year prior, so it surprisedme that somanymany of themhad not seen thiscoming.AndyrepliedwithananswerthatIdidnotexpect:“CEOsalwaysactonleadingindicatorsofgoodnews,butonlyactonlaggingindicatorsofbadnews.”“Why?”Iaskedhim.Heansweredinthestyleresonantofhisentirebook:“In

ordertobuildanythinggreat,youhavetobeanoptimist,becausebydefinitionyou are trying to do something that most people would consider impossible.Optimistsmostcertainlydonotlistentoleadingindicatorsofbadnews.”Butthisinsightwon’tbeinanybook.WhenIsuggestedhewritesomething

on the topic,his responsewas:“Whywould Ido that? Itwouldbeawasteoftimetowriteabouthowtonotfollowhumannature.Itwouldbeliketryingtostop the Peter Principle.* CEOsmust be optimists and all in all that’s a goodthing.” This is classic Andy Grove. He is amazingly perceptive and can seeeveryflawineveryperson,yetdespitethathebelievesinhumanpotentialmorethan anyone.Maybe that’swhy he has spent somuch time teaching us to bebetter.IthasbeenanhonorformetolearnfromAndyGrovethroughtheyearsandI

amexcitedforeveryonewhoisnewtoHighOutputManagementtojoinmeinthisexperience. Iknowthatyouwillenjoy thismarvelousbookwrittenby thebestteacherthatIhaveeverknown.

BenHorowitz,2015

* The Peter Principle is a concept in management theory in which the selection of a candidate for apositionisbasedonthecandidate’sperformanceintheircurrentrole,ratherthanonabilitiesrelatedtothe

intendedrole.Thus,“managersrisetotheleveloftheirincompetence.”

I

TheBreakfastFactory

1TheBasicsofProduction:DeliveringaBreakfast

(oraCollegeGraduate,oraCompiler,oraConvictedCriminal…)

TheThree-MinuteEgg

Tounderstandtheprinciplesofproduction,imaginethatyou’reawaiter,whichIwaswhileIwenttocollege,andthatyourtaskistoserveabreakfastconsistingof a three-minute soft-boiled egg, buttered toast, and coffee. Your job is toprepareanddeliverthethreeitemssimultaneously,eachofthemfreshandhot.Thetaskhereencompassesthebasicrequirementsofproduction.Theseareto

build and deliver products in response to the demands of the customer at ascheduleddeliverytime,atanacceptablequalitylevel,andatthelowestpossiblecost. Production’s charter cannot be to deliver whatever the customer wantswheneverhewants it, for thiswould requirean infiniteproductioncapacityorthe equivalent—very large, ready-to-deliver inventories. In our example, thecustomermaywant to have a perfect three-minute eggwith hot buttered toastandsteamingcoffeewaitingforhimthemomenthesitsdown.Tofulfillsuchanexpectation,youwouldeitherhavetohaveyourkitchenidleandpoisedtoservethe customerwhenever he drops in, or have a ready-to-consume inventory ofperfectlyboiledeggs,hotbutteredtoast,andcoffee.Neitherispractical.Instead, a manufacturer should accept the responsibility of delivering a

productatthetimecommittedto—inthiscase,byimplication,aboutfivetotenminutesafterthecustomerarrivesatourbreakfastestablishment.Andwemustmakeourbreakfastatacostthatenablesustosellitatacompetitivepriceandstill make an acceptable profit. How are we going to do this in the mostintelligentway?Westartbylookingatourproductionflow.The first thing we must do is to pin down the step in the flow that will

determinetheoverallshapeofouroperation,whichwe’llcallthelimitingstep.

Theissuehereissimple:whichofthebreakfastcomponentstakesthelongesttoprepare? Because the coffee is already steaming in the kitchen and the toasttakesonlyaboutaminute, theanswer isobviously theegg,soweshouldplantheentire jobaround the timeneeded toboil it.Notonlydoes thatcomponenttakethelongesttoprepare,theeggisalsoformostcustomersthemostimportantfeatureofthebreakfast.What must happen is illustrated opposite. To work back from the time of

delivery, you’ll need to calculate the time required to prepare the threecomponents to ensure that they are all ready simultaneously. First you mustallowtimetoassembletheitemsonatray.Nextyoumustgetthetoastfromthetoasterandthecoffeefromthepot,aswellastheeggoutoftheboilingwater.Addingtherequiredtimetodothistothetimeneededtogetandcooktheeggdefines the lengthof theentireprocess—called, inproduction jargon, the totalthroughputtime.Nowyoucometothetoast.Usingtheeggtimeasyourbase,youmustallow

yourselftimetogetandtoasttheslicesofbread.Finally,usingthetoasttimeasyourbase,youcandeterminewhenyouneedtopourthecoffee.Thekeyideaisthat we construct our production flow by starting with the longest (or mostdifficult, or most sensitive, or most expensive) step and work our way back.Notice when each of the three steps began and ended.We planned our flowaround the most critical step—the time required to boil the egg—and westaggeredeachoftheotherstepsaccordingtoindividualthroughputtimes;againinproductionjargon,weoffsetthemfromeachother.

Makingtheeggsisthelimitingstep.

Theideaofalimitingstephasverybroadapplicability.Take,forexample,theneedtorecruitcollegegraduatestoworkforIntel.Certainofourmanagersvisitthe colleges, interview some of the seniors, and invite the more promisingcandidates to visit the company.We bear the expense of the candidates’ trip,whichcanbeconsiderable.Duringthetrip,thestudentsarecloselyinterviewedbyothermanagersandtechnicalpeople.Afterdueconsideration,employmentisoffered to some of the studentswhose skills and capabilitiesmatch our needsbest,andthosewhoaccepttheofferseventuallycometoworkforthecompany.To apply the basic principle of production, you need to build the sequence

herearound itsmostexpensivefeature,which is thestudents’ trip to theplant,thanks to the cost of travel and the time that Intel managers spend with thecandidates.Tominimizetheuseofthisstepperfinalcollegehire,weobviouslyhave to increase the ratio of accepted offers to applicants invited to visit theplant,whichwedobyusingphone interviews to screenpeoplebefore issuinginvitations.Thetechniquesavesmoney,substantiallyincreasestheratioofoffersextendedperplantvisit,andreducestheneedtousetheexpensivelimitingstepperhire.

Theprincipleoftimeoffsetsisalsopresenthere.Workingbackfromthetimethe studentswill graduate, the recruiter staggers the various steps involved toallowtimeforeverything—on-campusinterviews,phonescreening,plantvisits—totakeplaceattheappropriatetimesduringthemonthsprecedinggraduation.

ProductionOperations

Other production principles underlie the preparation of our breakfast. In themaking of it, we find present the three fundamental types of productionoperations: process manufacturing, an activity that physically or chemicallychangesmaterialjustasboilingchangesanegg;assembly,inwhichcomponentsare put together to constitute a new entity just as the egg, the toast, and thecoffeetogethermakeabreakfast;andtest,whichsubjectsthecomponentsorthetotaltoanexaminationofitscharacteristics.Thereare,forexample,visualtestsmadeatpointsinthebreakfastproductionprocess:youcanseethatthecoffeeissteamingandthatthetoastisbrown.Process, assembly, and test operations can be readily applied to other very

differentkindsofproductivework.Take,forinstance,thetaskoftrainingasalesforce to sell a new product. The three types of production operations can beeasilyidentified.Theconversionoflargeamountsofrawdataabouttheproductinto meaningful selling strategies comprehensible to the sales personnel is aprocess step, which transforms data into strategies. The combination of thevarioussalesstrategiesintoacoherentprogramcanbecomparedtoanassemblystep. Here the appropriate product-selling strategies and pertinentmarket data(such as competitive pricing and availability) are made to flow into onepresentation,alongwithsuchthingsasbrochures,handouts,andflipcharts.Thetest operation comes in the form of a “dry run” presentation with a selectedgroupoffieldsalespersonnelandfieldsalesmanagement.Ifthedryrunfailsthetest, thematerialmust be “reworked” (anotherwell-establishedmanufacturingconcept)tomeettheconcernsandobjectionsofthetestaudience.Thedevelopmentof a “compiler,” amajorpieceof computer software, also

demonstrates process, assembly, and test. A computer understands and useshuman instruction only if it receives such instruction in its own language. Acompiler is an interpreter, enabling the computer to translate into its languagematerial written in terms and phrases resembling English.With a compiler, aprogrammer can thinkmore or less like a human being rather than having toadapthimselftothewaythecomputerprocessesinformation.Thetaskofgetting

amachinetointerpretandtranslateinthisfashionisobviouslyformidable;thusthedevelopmentofacompiler takesstrenuouseffortonthepartofskilledandgiftedsoftwareengineers.Theeffort,however,isjustifiedbythesimplificationitbringstocomputeruse.Inanycase,thedevelopmentoftheindividualpiecesoutofwhichacompiler

isbuiltrepresentsaseriesofprocessingsteps.Actualworkingpiecesofsoftwarearegeneratedoutofspecificationsandbasicdesignknow-how.Eachpiecethenundergoes an individual operation called a “unit test.” When one fails, thedefectiveportionofthesoftwareisreturnedtotheprocessphasefor“rework.”Afterall thepiecespass their respectiveunit tests, theyareassembled to formthe compiler. Then, of course, a “system test” is performed on the completeproductbeforeitisshippedtothecustomer.Timeoffsetsareusedextensivelyinthe task. Because throughput times for the various engineering steps are wellestablished, the timing of the releases of various bodies of software from onestagetoanothercanallbecalculatedandstagedinadvance.Breakfastpreparation, college recruiting, sales training, andcompilerdesign

are verymuch unlike one another, but all of them possess a basically similarflowofactivitytoproduceaspecificoutput.

AFewComplications

Real life,asyouknow, is fullof thicketsandunderbrush. Ina schematic flowchart, our breakfast operation assumed infinite capacity,meaning that nobodyhad towait foranavailable toasteror forapot toboilanegg in.Butnosuchidealworldexists.Whatwouldhappenifyouhadtostandinalineofwaiters,waitingforyourturntousethetoaster?Ifyoudidn’tadjustyourproductionflowto account for the queue, your three-minute egg could easily become a six-minute egg. So limited toaster capacity means you have to redo your flowaroundthenewlimitingstep.Theeggstilldeterminestheoverallqualityofthebreakfast,butyourtimeoffsetsmustbealtered.Howwould our model reflect the change in manufacturing flow?Working

back from the time of breakfast delivery, let’s see how the production isaffected,asillustratedopposite.Theeggcycleremainsthesame,asdoestheonefor coffee.But limited toaster capacitymakes forquite adifference.Nowyoumust account for thedelivery timeof the toast and thewait for a free toaster.This means the whole production process has to be conceived differently.Toaster capacity has become the limiting step, and what you do has to be

reworkedaroundit.

Withlimitedtoastercapacity,makingthetoastbecomesthelimitingstep.

Nowlet’scomplicatethingsalittlefurther.Whathappensifyouarestuckinlinewaitingforatoasterwhenit’stimetostartboilingyouregg?Yourconflictis seemingly irreconcilable, but it really isn’t. If you were managing therestaurant, you could turn your personnel into specialists by hiring one egg-cooker, one toast-maker, one coffee-pourer, and one person to supervise theoperation.Butthat,ofcourse,createsanimmenseamountofoverhead,probablymakingittooexpensivetoconsider.Ifyouwereawaiter,youcouldaskthewaiterinlinenexttoyoutohelpout—

toputyour toast inwhileyouranoff tostartyouregg.Butwhenyouhave todepend on someone else, the results are likely to be less predictable. As themanager,youcouldaddanothertoaster,butthisbecomesanexpensiveadditionof capital equipment.You could run the toaster continuously and build up aninventory of hot toast, throwing away what you can’t use but always havingimmediate access to product. That means waste, which can also become tooexpensive for the operation. But at least you know that alternatives do exist:

equipment capacity, manpower, and inventory can be traded off against eachotherandthenbalancedagainstdeliverytime.Because each alternative costs money, your task is to find themost cost-

effective way to deploy your resources—the key to optimizing all types ofproductivework.Bearinmindthatinthisandinothersuchsituationsthereisarightanswer,theonethatcangiveyouthebestdeliverytimeandproductqualityatthelowestpossiblecost.Tofindthatrightanswer,youmustdevelopaclearunderstanding of the trade-offs between the various factors—manpower,capacity, and inventory—and you must reduce the understanding to aquantifiablesetofrelationships.Youprobablywon’tuseastopwatchtoconducta time-and-motion studyof thepersonbehinda toaster; norwill youcalculatethe precise trade-off between the cost of toast inventory and the added toastercapacity in mathematical terms. What is important is the thinking you forceyourselftogothroughtounderstandtherelationshipbetweenthevariousaspectsofyourproductionprocess.Let’stakeourmanufacturingexampleastepfurtherandturnourbusinessinto

a high-volume breakfast factory operation. First, you buy a continuous egg-boilerthatwillproduceaconstantsupplyofperfectlyboiledthree-minuteeggs.Itwill look something likewhat’s drawn in the figure opposite.Note that ourbusinessnowassumesahighandpredictabledemandfor three-minuteeggs; itcannotnowreadilyprovideafour-minuteegg,becauseautomatedequipmentisnot very flexible. Second, you match the output of the continuous egg-boilerwiththeoutputofacontinuoustoaster,asspecializedpersonnelloadeachpieceof equipment and deliver the product. We have now turned things into acontinuousoperationattheexpenseofflexibility,andwecannolongerprepareeach customer’s order exactlywhen andhowhe requests it. Soour customershave to adjust their expectations if theywant to enjoy thebenefitsofournewmode:lowercostandmorepredictableproductquality.

Thecontinuousegg-boiler:aconstantsupplyofthree-minuteeggs.

Butcontinuousoperationdoesnotautomaticallymean lowercostandbetterquality. What would happen if the water temperature in the continuous egg-boiler quietly went out of specification? The entire work-in-process—all theeggsintheboiler—andtheoutputofthemachinefromthetimethetemperatureclimbed or dropped to the time the malfunction was discovered becomesunusable.Allthetoastisalsowastedbecauseyoudon’thaveanyeggstoservewithit.Howdoyouminimizetheriskofabreakdownofthissort?Performingafunctionaltestisoneway.Fromtimetotimeyouopenaneggasitcomesoutofthe machine and check its quality. But you will have to throw away the eggtested. A second way involves in-process inspection, which can take manyforms.You could, for example, simply insert a thermometer into thewater sothatthetemperaturecouldbeeasilyandfrequentlychecked.Toavoidhavingtopaysomeonetoreadthethermometer,youcouldconnectanelectronicgadgettoitthatwouldsetoffbellsanytimethetemperaturevariedbyadegreeortwo.Thepoint is thatwhenever possible, you should choose in-process tests over thosethatdestroyproduct.Whatelsecouldgowrongwithourcontinuousegg-machine?Theeggsgoing

into it could be cracked or rotten, or they could be over-or undersized,whichwouldaffecthowfasttheycook.Toavoidsuchproblems,youwillwanttolookat the eggs at the time of receipt, something called incoming or receivinginspection. If theeggsareunacceptableinsomeway,youaregoingtohavetosendthemback,leavingyouwithnone.Nowyouhavetoshutdown.Toavoidthat, you need a raw material inventory. But how large should it be? Theprinciple to be applied here is that you should have enough to cover your

consumption rate for the length of time it takes to replace your rawmaterial.Thatmeansifyoureggmancomesbyanddeliversonceaday,youwanttokeepaday’sworthofinventoryonhandtoprotectyourself.Butremember,inventorycostsmoney, so you have to weigh the advantage of carrying a day’s supplyagainstthecostofcarryingit.Besidesthecostoftherawmaterialandthecostofmoney,you shouldalso try togauge theopportunityat risk:whatwould itcostifyouhadtoshutyoureggmachinedownforaday?Howmanycustomerswould you lose?Howmuchwould it cost to lure them back? Such questionsdefinetheopportunityatrisk.

AddingValue

All production flows have a basic characteristic: the material becomes morevaluableasitmovesthroughtheprocess.Aboiledeggismorevaluablethanarawone,afullyassembledbreakfastismorevaluablethanitsconstituentparts,andfinally, thebreakfastplacedinfrontof thecustomerismorevaluablestill.The last carries the perceived value the customer associates with theestablishmentwhenhedrivesintotheparkinglotafterseeingthesign“Andy’sBetter Breakfasts.” Similarly, a finished compiler is more valuable than theconstituent parts of semantic analysis, code generation, and run time, and acollegegraduatetowhomwearereadytoextendanemploymentofferismorevaluabletousthanthecollegestudentwemeetoncampusforthefirsttime.Acommonruleweshouldalwaystrytoheedistodetectandfixanyproblem

inaproductionprocessatthelowest-valuestagepossible.Thus,weshouldfindand reject the rotten egg as it’s being delivered from our supplier rather thanpermittingthecustomertofindit.Likewise,ifwecandecidethatwedon’twanta college candidate at the timeof the campus interview rather thanduring thecourse of a plant visit, we save the cost of the trip and the time of both thecandidateandtheinterviewers.Andweshouldalsotrytofindanyperformanceproblematthetimeoftheunittestofthepiecesthatmakeupacompilerratherthaninthecourseofthetestofthefinalproductitself.Finally, at the risk of being considered hard-hearted, let’s examine the

criminal justice system as if it were a production process aimed at findingcriminals and putting them into jail. The production begins when a crime isreported to the police and the police respond. In many instances, after somequestionsareasked,nofurtheractioncanbetaken.Forthosecrimeswhichthepolicecanpursue,thesecondstepismoreinvestigation.Butthecaseoftenends

hereforlackofevidence,complaintsbeingdropped,andsoon.Ifthingsmovetothenextstage,asuspect isarrested,andthepolicetry tofindwitnessesandbuildacase,hopingtogetanindictment.Onceagain,anindictmentisoftennotreturned because of insufficient evidence. For the cases that actually do goahead, the next stage is trial. Sometimes the suspect is found not guilty;sometimesthecaseisdismissed.Butwhenaconvictionissecured,theprocessmovestothesentencingandappealsround.Attimesapersonfoundguiltyofacrime will be given a suspended sentence and probation, and at others theconvictionwillbeoverturnedonappeal.Forthesmallfractionthatremains,thefinalstageisjail.If we make some reasoned assumptions about the percentages that move

forward at each stage and the costs associated with each, we arrive at somestrikingconclusions.Ifwecompilethecostoftheeffortthatgoesintosecuringaconvictionandassignitonlytothosecriminalswhoactuallyendupinjail,wefind that the cost of a single conviction works out to be well over a milliondollars—an absolutely staggering sum. The number is so high, of course,becauseonlyaverysmallpercentageoftheflowofaccusedpersonsmakesitallthewaythroughtheprocess.Everyoneknowsthatprisonsareovercrowded,andthat many criminals end up serving shorter jail terms or no jail terms at allbecausecellsareinsuchshortsupply.Soaterriblyexpensivetrade-offresults,violatingthemostimportantproductionprinciples.Thelimitingstephereshouldclearlybeobtainingaconviction.Theconstructioncostofajailcelleventodayisonlysome$80,000.This,plusthe$10–20,000itcoststokeepapersoninjailforayear,isasmallamountcomparedtothemilliondollarsrequiredtosecureaconviction.Not to jail a criminal inwhomsocietyhas investedover amilliondollarsforlackofan$80,000jailcellclearlymisusessociety’stotalinvestmentin thecriminal justice system.And thishappensbecausewepermit thewrongstep(theavailabilityofjailcells)tolimittheoverallprocess.

2ManagingtheBreakfastFactory

IndicatorsasaKeyTool

Ahungrypublichaslovedthebreakfastyou’vebeenserving,andthankstothehelpofyourmanycustomersandafriendlybanker,you’vecreatedabreakfastfactory, which among other things uses specialized production lines for toast,coffee,andeggs.Asmanagerof thefactory,youhaveasubstantialstaffandalotofautomatedequipment.Buttorunyouroperationwell,youwillneedasetof good indicators, ormeasurements.Your output, of course, is no longer thebreakfasts you deliver personally but rather all the breakfasts your factorydelivers,profitsgenerated,and thesatisfactionofyourcustomers.Just togetafixonyouroutput,youneedanumberofindicators;togetefficiencyandhighoutput,youneedevenmoreofthem.Thenumberofpossibleindicatorsyoucanchoose is virtually limitless, but for any set of them tobeuseful, youhave tofocuseachindicatoronaspecificoperationalgoal.Let’s say that asmanager of the breakfast factory, youwillworkwith five

indicatorstomeetyourproductiongoalsonadailybasis.Whichfivewouldtheybe?Putanotherway,whichfivepiecesofinformationwouldyouwanttolookateachday,immediatelyuponarrivingatyouroffice?Herearemycandidates.First,you’llwanttoknowyoursalesforecastforthe

day. How many breakfasts should you plan to deliver? To assess how muchconfidence you should place in your forecast, you would want to know howmanyyoudeliveredyesterdaycomparedtohowmanyyouplannedondelivering—in other words, the variance between your plan and the actual delivery ofbreakfastsfortheprecedingday.Yournextkeyindicatorisrawmaterialinventory.Doyouhaveenougheggs,

bread,andcoffeeonhandtokeepyourfactoryrunningtoday?Ifyoufindyouhave too little inventory, you can still order more. If you find you have toomuch,youmaywanttocanceltoday’seggdelivery.Anotherimportantpieceofinformationistheconditionofyourequipment.If

anything broke down yesterday, you will want to get it repaired or rearrangeyourproductionlinetomeetyourforecastfortheday.Youalsomustgeta fixonyourmanpower. If twowaitersareoutsick,you

willhavetocomeupwithsomethingifyouarestillgoingtomeetthedemandforecasted.Shouldyoucallintemporaryhelp?Shouldyoutakesomeoneoffthetoasterlineandmakehimawaiter?Finally,youwanttohavesomekindofquality indicator.It isnotenoughto

monitorthenumberofbreakfastseachwaiterdelivers,becausethewaiterscouldhave been rude to the customers even as they served a record number ofbreakfasts.Becauseyourbusinessdependsonpeoplewantingwhatyousell,youmustbeconcernedwiththepublic’sopinionofyourservice.Perhapsyoushouldset up a “customer complaint log” maintained by the cashier. If one of yourwaiters elicitedmore than the usual number of complaints yesterday, youwillwanttospeaktohimfirstthingtoday.All these indicatorsmeasurefactorsessential torunningyourfactory.Ifyou

lookatthemearlyeveryday,youwilloftenbeabletodosomethingtocorrectapotentialproblembeforeitbecomesarealoneduringthecourseoftheday.Indicatorstendtodirectyourattentiontowardwhattheyaremonitoring.Itis

like riding a bicycle: youwill probably steer itwhere you are looking. If, forexample,youstartmeasuringyour inventory levelscarefully,youare likely totakeactiontodriveyourinventorylevelsdown,whichisgooduptoapoint.Butyourinventoriescouldbecomesoleanthatyoucan’treacttochangesindemandwithout creating shortages. So because indicators direct one’s activities, youshouldguardagainstoverreacting.Thisyoucandobypairingindicators,sothattogether both effect and counter-effect are measured. Thus, in the inventoryexample, you need to monitor both inventory levels and the incidence ofshortages. A rise in the latter will obviously lead you to do things to keepinventoriesfrombecomingtoolow.Theprincipleherewasevidentmanytimesinthedevelopmentofacompiler.

Measuringthecompletiondateofeachsoftwareunitagainstitscapabilityisoneexample.Watchingthispairofindicatorsshouldhelpustoavoidworkingontheperfectcompilerthatwillneverbeready,andalsotoavoidrushingtofinishonethat is inadequate. In sum, joint monitoring is likely to keep things in theoptimummiddleground.Nowhere can indicators—and paired indicators—be of more help than in

administrative work. Having come to this realization, our company has been

usingmeasurementsasakeytooltoimprovetheproductivityofadministrativeworkforseveralyears.Thefirstruleisthatameasurement—anymeasurement—isbetterthannone.Butagenuinelyeffectiveindicatorwillcovertheoutputofthework unit and not simply theactivity involved.Obviously, youmeasure asalesmanbytheordershegets(output),notbythecallshemakes(activity).Thesecondcriterionforagoodindicatoristhatwhatyoumeasureshouldbea

physical, countable thing. Examples of effective measures of administrativeoutput are shown below. Because those listed here are all quantity or outputindicators, theirpairedcounterparts should stress thequalityofwork.Thus, inaccountspayable, thenumberofvouchersprocessedshouldbepairedwith thenumber of errors found either by auditing or by our suppliers. For anotherexample, the number of square feet cleaned by a custodial group should bepairedwithapartiallyobjective/partiallysubjectiveratingofthequalityofworkasassessedbyaseniormanagerwithanofficeinthatbuilding.

ADMINISTRATIVEFUNCTION WORKOUTPUTINDICATOR

Accountspayable #VouchersprocessedCustodial #SquarefeetcleanedCustomerservice #SalesordersenteredDataentry #TransactionsprocessedEmployment #Peoplehired(bytypeof

hire)Inventorycontrol #Itemsmanagedininventory

Examplesofadministrativeworkoutputindicators.

Such indicators havemany uses. First, they spell out very clearlywhat theobjectives of an individual or group are. Second, they provide a degree ofobjectivitywhenmeasuringanadministrativefunction.Third,andasimportantas any, they give us a measure by which various administrative groupsperforming the same function indifferent organizations canbe comparedwitheachother.Theperformanceofacustodialgroupinonemajorbuildingcannowbe compared with that of another group in a second building. In fact, ifindicatorsareput inplace, thecompetitivespirit engendered frequentlyhasanelectrifyingeffectonthemotivationeachgroupbringstoitswork,alongwithaparallel improvement in performance.More about this laterwhenwe examine

the“sportsanalogy.”

TheBlackBox

Wecanthinkofourbreakfastfactoryasifitwerea“blackbox”:input(therawmaterials)andthelaborofwaiters,helpers,andyou,themanager,flowingintothebox,andtheoutput(thebreakfast)flowingoutofitasillustratedbelow.Ingeneral,wecanrepresentanyactivity that resemblesaproductionprocess inasimple fashion as a black box. Thus, we can draw a black box to representcollegerecruiting,wheretheinputistheapplicantsoncampusandtheoutputiscollege graduateswho have accepted our employment offers. The labor is theworkofouron-campusinterviewersandthemanagersandtechnicalpeoplewhointerviewbackat theplant.Similarly, theprocessoffieldsales trainingcanbeseenasablackboxwiththeinputbeingtherawproductspecificationsandtheoutputbeingtrainedsalespersonnel.Thelaborhereistheworkofthemarketingandmerchandisingpeoplewhoturnrawinformationintousablesalestoolsandtrainthefieldsalespersonneltoexploitthem.Infact,wecanrepresentmost,ifnotall,administrativeworkbyourmagicalblackbox.Agroupwhosejobistobillcustomershasasitsinputtheinformationaboutthecustomer—whathehaspurchased,thepricingdata,andtheshipmentrecords;andoutputisthefinalbillsenttothecustomerthroughwhichpaymentiscollected.Thelaboristheworkofallpersonnelinvolved.

Thebreakfastfactory—asa“blackbox.”

Theblackboxsortsoutwhat the inputs, theoutput,and the laborare in the

production process.We can improve our ability to run that process by cuttingsomewindowsinourboxsothatwecanseesomeofwhatgoesonwithinit.Bylookingthroughtheopenings,asillustratedbelow,wecanbetterunderstandtheinternalworkingsofanyproductionprocessandassesswhatthefutureoutputislikelytobe.

Bypeeringthroughthewindowsintheblackbox,wecangetanideaofwhatthefutureoutputislikelytobe.

Leadingindicatorsgiveyouonewaytolookinsidetheblackboxbyshowingyouinadvancewhatthefuturemightlooklike.Andbecausetheygiveyoutimeto take corrective action, theymake it possible for you to avoid problems.Ofcourse, for leading indicators to do you any good, you must believe in theirvalidity.Whilethismayseemobvious,inpractice,confidenceisnotaseasytocomebyasitsounds.Totakebig,costly,orworrisomestepswhenyouarenotyetsureyouhaveaproblemishard.Butunlessyouarepreparedtoactonwhatyourleadingindicatorsaretellingyou,allyouwillgetfrommonitoringthemisanxiety.Thus,theindicatorsyouchooseshouldbecredible,sothatyouwill,infact,actwhenevertheyflashwarningsignals.Leading indicators might include the daily monitors we use to run our

breakfast factory, from machine downtime records to an index of customersatisfaction—both of which can tell us if problems lie down the road. Agenerally applicable example of a “window” cut into the black box is thelinearityindicator.Inthefigurebelow,weprovideoneforthecollegerecruitingprocess.Plottedhereisthenumberofcollegegraduateswhohaveacceptedouroffersversusthemonthoftheyear.Ifallwentideally,wewouldmovealongthestraightlinethatwouldyieldourhiringtargetforthesemesterbythemonthof

June.IfbyApriltheactualprogressisasshownhere,wewillfindourselvesfarbelow the ideal straight line. So from reading the indicator,we know that theonlywaywecanhitourtargetisbygettingacceptanceatamuchhigherrateinthe remaining twomonths thanwehadgotten in thepreceding four.Thus, thelinearity indicatorflashesanearlywarning,allowingus timeto takecorrectiveaction.Without it, wewould discover that we hadmissed our target in June,whennothingcanbedoneaboutit.

Thelinearityindicatorcangiveusanearlywarningthatwearelikelytomissourtarget.

If we consider a manufacturing unit in this fashion, we may assume thatbecauseitmakesmonthlygoalsregularly,alliswell.Butwecancutawindowinto the black box here,measure production output against time as themonthproceeds, and compare that with the ideal linear output. We may learn thatoutputperformanceisspreadevenlythroughoutthecourseofthemonthorthatit is concentrated in the last week of the month. If the latter is the case, themanageroftheunitisprobablynotusingmanpowerandequipmentefficiently.Andifthesituationisnotremedied,oneminorbreakdowntowardmonth’sendcould cause the unit to miss its monthly output goal entirely. The linearity

indicatorwillhelpyouanticipatesuchaproblemandisthereforequitevaluable.Alsovaluableare trend indicators.Theseshowoutput (breakfastsdelivered,

software modules completed, vouchers processed) measured against time(performancethismonthversusperformanceoveraseriesofpreviousmonths),andalsoagainstsomestandardorexpectedlevel.Adisplayoftrendsforcesyoutolookatthefutureasyouareledtoextrapolatealmostautomaticallyfromthepast. This extrapolation gives us another window in our black box. Also,measurementagainsta standardmakesyou think throughwhy the resultswerewhattheywere,andnotwhatthestandardsaidtheywouldbe.Anothersoundwaytoanticipate thefuture is through theuseof thestagger

chart, which forecasts an output over the next several months. The chart isupdatedmonthly, so that eachmonth youwill have an updated version of thethen-current forecast information as compared to several prior forecasts. Youcanreadilyseethevariationofoneforecastfromthenext,whichcanhelpyouanticipatefuturetrendsbetterthanifyouusedasimpletrendchart.Inmyexperience,nowherehasthestaggerchartbeenmoreproductivethanin

forecasting economic trends. Theway it works is shown in the figure below,which gives us forecasted rates of incoming orders for an Intel division. Thestaggerchartthenprovidesthesameforecastpreparedinthefollowingmonth,inthemonthafter that, and soon.Sucha chart showsnotonlyyouroutlook forbusinessmonthbymonthbutalsohowyouroutlookvariedfromonemonthtothenext.Thiswayof lookingat incomingbusiness,ofcourse,makeswhoeverdoes the forecasting take his task very seriously, because he knows that hisforecast for any givenmonthwill be routinely comparedwith future forecastsandeventuallywiththeactualresult.Butevenmoreimportant,theimprovementordeteriorationof the forecastedoutlookfromonemonth to thenextprovidesthemostvaluableindicatorofbusinesstrendsthatIhaveeverseen.Iwouldgoas far as to say that it’s too bad that all economists and investment advisersaren’tobligedtodisplaytheirforecastsinastaggerchartform.Thenwecouldreallyhaveawaytoevaluatewhateveranyoneofthemchoosestosay.

(*meanstheactualnumberforthatmonth)

Ihavefoundthe“staggerchart”thebestmeansofgettingafeelforfuturebusinesstrends.

Finally, indicators can be a big help in solving all types of problems. Ifsomethinggoeswrong,youwillhaveabankofinformationthatreadilyshowsall the parameters of your operation, allowingyou to scan them for unhealthydeparturesfromthenorm.Ifyoudonotsystematicallycollectandmaintainanarchiveof indicators,youwillhavetodoanawful lotofquickresearchtogettheinformationyouneed,andbythetimeyouhaveit, theproblemislikelytohavegottenworse.

ControllingFutureOutput

Therearetwowaystocontroltheoutputofanyfactory.Someindustriesbuildtoorder.Forexample,whenyougoshoppingforasofa,youaregoingtohavetowaitalongtimetogetwhatyoubought,unlessyoubuyitrightoffthefloor.Afurniturefactorybuildstoorder.Whenitlearnswhatyouwant,thefactorylooksforaholeinitsmanufacturingscheduleandmakestheitemforyou.Ifyouorderanewcar rather thanbuyingonerightoff the lot, thesamethinghappens: the

plantwillpaintthecarinthecoloryouwantandprovidetheoptionsyouwant,but you will have to wait for it. And our breakfast factory, of course, buildsbreakfaststoorder.Butifyourcompetitioninthesofabusinessmakesthesameproductbuthasit

ready in four weeks while you need four months, you are not going to havemanycustomers.Soeventhoughyouwouldmuchratherbuildtoorder,youwillhavetouseanotherwaytocontroltheoutputofyourfactory.Inshort,youwillhavetobuildtoforecast,whichisacontemplationoffutureorders.Todothis,themanufacturersetsuphisactivitiesaroundareasonedspeculationthatorderswillmaterializeforspecificproductswithinacertaintime.Anobviousdisadvantagehereisthatthemanufacturertakesaninventoryrisk.

Since the forecast is an assessment of future requirements, which themanufacturercommitsresources tosatisfy, thefactorycouldbe inanimmenseamount of trouble if the orders do notmaterialize or if theymaterialize for aproductotherthantheoneanticipated.Ineithercase,unwantedinventoryistheresult. To build to forecast, you risk capital to respond to anticipated futuredemandingoodorder.AtIntel,webuildtoforecastbecauseourcustomersdemandthatwerespond

to their needs in a timely fashion, even though ourmanufacturing throughputtimesarequitelong.Ourbreakfastfactorymakesitsproducttocustomerorder,but buys from its suppliers—like the egg man—on the basis of forecasteddemand. Similarly, most companies recruit new college graduates to fillanticipated needs—rather than recruiting only when a need develops, whichwouldbefoolishbecausecollegegraduatesare turnedout inahighlyseasonalfashion. Computer software products, such as compilers, are also typicallydeveloped in response to an anticipated market need rather than to specificcustomerorder.So“building”toforecastisaverycommonbusinesspractice.Deliveringaproduct thatwasbuilt toforecast toacustomerconsistsof two

simultaneousprocesses,eachwithaseparatetimecycle.Amanufacturingflowmust occur inwhich the rawmaterialmoves throughvariousproduction stepsand finally enters the finished goods warehouse, as illustrated below.Simultaneously, a salesman finds a prospect and sells to that prospect, whoeventually places an order with the manufacturer. Ideally, the order for theproduct and the product itself should arrive on the shipping dock at the sametime.Because the art and science of forecasting is so complex, you might be

tempted to give all forecasting responsibility to a singlemanagerwho can bemadeaccountableforit.Butthisusuallydoesnotworkverywell.Whatworksbetter is to ask both themanufacturing and the sales departments to prepare aforecast, so that people are responsible for performing against their ownpredictions.At Intel we try to match the two parallel flows with as much precision as

possible. If there’s nomatch,we end upwith a customer order thatwe can’tsatisfyorwithafinishedproductforwhichwehavenocustomer.Eitherwaywehaveproblems.Obviously, if thematchdoescomeoff,withaforecastedorderbecomingarealorder,thecustomer’srequirementscanbenicelysatisfiedwiththefactory’sproductdelivery.

Theorderfortheproductandtheproductitselfshouldarriveattheshippingdockatthesametime.

Theidealisrarelyfoundintherealworld.Moreoften,customerordersdon’tdevelop in time or the customer changes his mind. As for the other flow,manufacturingcouldmissdeadlines,ormakemistakes,orencounterunforeseenproblems. Because neither the sales flow nor the manufacturing flow iscompletely predictable, we should deliberately build a reasonable amount of“slack”intothesystem.Andinventoryisthemostobviousplaceforit.Clearly,themoreinventorywehave,themorechangewecancopewithandstillsatisfyorders.But inventory costsmoney to build and keep, and therefore should becontrolledcarefully.Ideally,inventoryshouldbekeptatthelowest-valuestage,aswe’ve learnedbefore, like raweggskept at the breakfast factory.Also, thelowerthevalue,themoreproductionflexibilityweobtainforagiveninventory

cost.It is a good idea to use stagger charts in both themanufacturing and sales

forecasts. As noted, they will show the trend of change from one forecast toanother, aswell as the actual results.By repeatedly observing the variance ofone forecast from another, you will continually pin down the causes ofinaccuracyandimproveyourabilitytoforecastbothordersandtheavailabilityofproduct.Forecasting future work demands and then adjusting the output of an

“administrative factory” represents a very important way in which itsproductivity can be increased. Though an old and honored way of operating“widget factories,” theapplicationof forecasting techniques ishardlycommonas a way to control administrative work. Such work has up to now beenconsidered qualitatively different fromwork in awidget factory, and has alsolackedobjectiveperformancestandardsneededtosizeorscaletheworkunit.Butifwehavecarefullychosenindicatorsthatcharacterizeanadministrative

unit and watch them closely, we are ready to apply the methods of factorycontroltoadministrativework.Wecanusedefactostandards,inferredfromthetrend data, to forecast the number of people needed to accomplish variousanticipated tasks. By rigorous application of the principles of forecasting,manpowercanbereassignedfromoneareatoanother,andtheheadcountmadeto match the forecasted growth or decline in administrative activity. Withoutrigor,thestaffingofadministrativeunitswouldalwaysbeleftatitshighestleveland, given Parkinson’s famous law, people would find ways to let whateverthey’redoingfillthetimeavailableforitscompletion.Thereisnoquestionthathaving standards and believing in them and staffing an administrative unitobjectively using forecastedworkloadswill help you tomaintain and enhanceproductivity.

AssuringQuality

Aswehavesaid,manufacturing’scharteristodeliverproductataqualitylevelacceptable to the customer atminimumcost.To assure that the quality of ourproductwill in fact be acceptable, all production flows,whether they “make”breakfasts, college graduates, or software modules, must possess inspectionpoints. To get acceptable quality at the lowest cost, it is vitally important torejectdefectivematerialatastagewhereitsaccumulatedvalueisatthelowestpossible level.Thus,asnoted,wearebetteroffcatchingabadrawegg thana

cooked one, and screening out our college applicant before he visits Intel. Inshort,rejectbeforeinvestingfurthervalue.In the language of production, the lowest-value-point inspection where we

inspect raw material is called incoming material inspection or receivinginspection. If we again use a black box to represent our production process,inspectionsthatoccuratinterveningpointswithinitarecalled,logicallyenough,in-process inspections. Finally, the last possible point of inspection,when theproduct is ready to be shipped to the customer, is called final inspection oroutgoingqualityinspection.Thethreetypesaredepictedbelow.

Thekeyprincipleistorejectthedefective“material”atitslowest-valuestage.

Whenmaterialisrejectedatincominginspection,acoupleofchoicespresentthemselves.Wecansenditbacktothevendorasunacceptable,orwecanwaiveour specifications and use the substandardmaterial anyway. The latter wouldresult in a higher reject rate in our production process than if we had usedthoroughly acceptablematerial, but thatmight be less expensive than shuttingdown the factory altogether until our vendor provides better material. Suchdecisions canonlybemadeproperlybyabalancedgroupofmanagers,whichtypicallyconsistsof representatives from thequalityassurance,manufacturing,anddesignengineeringdepartments.Thisgroupcanweighalltheconsequencesofrejectingoracceptingsubstandardrawmaterial.Whileinmostinstancesthedecisiontoacceptorrejectdefectivematerialata

given inspection point is an economic one, one should never let substandardmaterialproceedwhen itsdefectscouldcauseacomplete failure—areliabilityproblem—for our customer. Simply put, because we can never assess theconsequences of an unreliable product, we can’t make compromises when itcomes to reliability.Thinkofacomponentgoing into themakingofacardiacpacemaker. If some of the components don’t work upon receipt by themanufacturer,hecanreplacethemwhiletheunitisstillinthefactory.Thiswillprobablyincreasecosts.Butifthecomponentfailslater,afterthepacemakerhasbeenimplanted,thecostofthefailureismuchmorethanafinancialone.Inspections,ofcourse,costmoneytoperformandfurtheraddtoexpenseby

interferingwiththemanufacturingflowandmakingitmorecomplicated.Somematerial has to be recycled through steps already performed, upsetting thesmoothnesswithwhichtherestofthematerialmoves.Accordingly,oneshouldapproach the need to inspect recognizing that a balance exists between thedesiredresultof theinspection, improvedquality,andminimumdisturbancetotheproductionprocessitself.Let’s consider a few techniques commonly used to balance the two needs.

Thereisagate-likeinspectionandamonitoringstep.Intheformer,allmaterialis held at the “gate” until the inspection tests are completed. If the materialpasses,itismovedontothenextstageintheproductionprocess;ifthematerialfails, it will be returned to an earlier stage, where it will be reworked orscrapped.Inthelatter,asampleofthematerialistaken,andifitfails,anotationismadefromwhichafailurerate iscalculated.Thebulkof thematerial isnotheld as the sample is taken but continues tomove through themanufacturingprocess. The smoothness of the flow ismaintained, but if, for example, threesuccessive samples fail themonitoring test, we can stop the line.What is thetrade-offhere?Ifweholdallthematerial,weaddtothroughputtimeandslowdownthemanufacturingprocess.Amonitorproducesnocomparableslowdownbut might let some bad material escape before we can act on the monitor’sresultsandshutthingsdown,whichmeansthatwemighthavetorejectmateriallateratahigher-valuestage.Clearly,forthesamemoneywecandoalotmoremonitoring than gate-type inspections; if we do the former, we may wellcontribute more to the overall quality of the product than if we choose lessfrequentgate-likeinspections.Thetrade-offhereisnotobvious,andanychoicehastobemadewithaspecificcaseinmind.Asaruleofthumb,weshouldleantowardmonitoringwhen experience showswe are not likely to encounter bigproblems.

Another way to lower the cost of quality assurance is to use variableinspections.Becausequality levelsvaryover time, it isonlycommonsense tovaryhowoftenweinspect.Forinstance,ifforweekswedon’tfindproblems,itwouldseemlogicaltochecklessoften.Butifproblemsbegintodevelop,wecantestevermorefrequentlyuntilqualityagainreturnstotheprevioushighlevels.The advantage here is still lower costs and even less interference with theproduction flow. Yet this approach is not used very often, even in widgetmanufacturing.Whynot?Probablybecausewearecreaturesofhabitandkeepdoingthingsthewaywealwayshave,whetheritbefromweektoweekoryeartoyear.Suitablythoughtthrough,intelligentinspectionschemescanactuallyincrease

theefficiencyandproductivityofanymanufacturingoradministrativeprocess.Let’stakeanexampleverydifferentfromthemakingofwidgetsorbreakfasts.I recently read a story in a news magazine that said that the American

EmbassyinLondoncouldnotdealwithadelugeofvisaapplications.Someonemillion Britons apply for visas each year, of which about 98 percent areapproved. The embassy employs sixty people, who process asmany as 6,000applicationsaday.Mostapplicationsarereceivedbymail,andatanytime,from60,000to80,000Britishpassportsareintheembassy’shands.Meanwhile,linesofonehundredormoreBritishandothernationalsstandinfrontofthebuilding,looking for an opportunity to walk their passports through. The embassy hastriedanumberofwaystohandlemattersmoreefficiently,includingnewspaperadvertisements asking tourists to apply early and to expect a three-weekturnaround.The embassy also installed boxeswhere applicants could drop offtheir passports and visa applications if they really needed same-day service.Evenso,thelinesattheembassyremainedlong.In fact, the embassy’s expediting schemes only made the problem worse,

becausenothingwasdonetoaddressthebasicissue:tospeedtheprocessingofvisas overall. Time and money were spent to classify various kinds ofapplications slated for different processing times, but this only created morelogisticaloverheadwithnoeffectonoutput.If our government wants British tourists to visit the United States, our

governmentshouldnotirritatethesewould-bevisitors.Andiftheembassycan’tgetthemoneytoincreaseitsstaff,asimplesolutioncanbeborrowedfrombasicproductiontechniques.Weneed,inshort,toreplacetheirpresentschemewithaqualityassurancetest.

For that, thebureaucraticmindsat theembassywouldneed toaccept that a100 percent check of the visa applicants is unnecessary. Some 98 percent ofthose applying are approvedwithout any question. So if the embassywere toinstituteasamplingtestofvisas(aqualityassurancetest),andathoroughoneatthat, the logjam of applications could be brokenwithoutmaterially increasingthe chance that the undesirablewill enter our country.Moreover, the embassycouldselect thesample tobecheckedaccording topredeterminedcriteria.Thevisa processing could then work rather like the Internal Revenue Service.Through the checks andaudits that the IRSperforms, thatgovernment agencyinduces compliance among most taxpayers without having an agent look ateverysinglereturn.Later, when we examine managerial productivity, we’ll see that when a

managerdigsdeeplyintoaspecificactivityunderhisjurisdiction,he’sapplyingthe principle of variable inspection. If the manager examined everything hisvarioussubordinatesdid,hewouldbemeddling,whichforthemostpartwouldbeawasteofhistime.Evenworse,hissubordinateswouldbecomeaccustomedto not being responsible for their own work, knowing full well that theirsupervisorwillcheckeverythingoutclosely.Theprincipleofvariableinspectionapplied tomanagerialwork nicely skirts both problems, and, aswe shall see,givesusanimportanttoolforimprovingmanagerialproductivity.

Productivity

Theworkingsofourblackboxcanfurnishuswiththesimplestandmostusefuldefinitionofproductivity.Theproductivityofanyfunctionoccurringwithinitistheoutputdividedbythelaborrequiredtogeneratetheoutput.Thus,onewaytoincreaseproductivityistodowhateverwearenowdoing,butfaster.Thiscouldbedonebyreorganizingtheworkareaorjustbyworkingharder.Herewe’venotchangedwhatworkwe do,we’ve just institutedways to do it faster—gettingmoreactivities per employee-hour to go on inside the black box.Because theoutputof theblackbox isproportional to theactivity thatoccurswithin it,wewillgetmoreoutputperhour.Thereisasecondwaytoimproveproductivity.Wecanchangethenatureof

theworkperformed:whatwedo,nothowfastwedoit.Wewanttoincreasetheratio of output to activity, thereby increasing output even if the activity peremployee-hour remains the same. As the slogan has it, we want to “worksmarter,notharder.”

Productivitycanbeincreasedbyperformingtheworkactivitiesatahigherrate…

…orbyincreasingtheleverageoftheactivities.

Here I’d like to introduce the concept of leverage, which is the outputgeneratedbyaspecifictypeofworkactivity.Anactivitywithhighleveragewillgenerate a high level of output; an activity with low leverage, a low level ofoutput.Forexample,awaiterabletoboiltwoeggsandoperatetwotoasterscandeliver twobreakfasts foralmost the sameamountofworkasone.Hisoutputperactivity,and thereforehis leverage, ishigh.Awaiterwhocanhandleonlyone egg and one toaster at a time possesses lower output and leverage. ThesoftwareengineerusingaprogramminglanguageratherlikeEnglish,latertobetranslated by a compiler, can solvemany problems per hour of programming.Hisoutputandleveragearehigh.Asoftwareengineerusingamorecumbersomeprogrammingmethodofonesandzeroswillrequiremanymorehourstosolvethe same number of problems.His output and leverage are low. Thus, a veryimportantway to increase productivity is to arrange thework flow inside ourblackboxsothatitwillbecharacterizedbyhighoutputperactivity,whichistosayhigh-leverageactivities.Automationiscertainlyonewaytoimprovetheleverageofalltypesofwork.

Havingmachines to help them, human beings can createmore output. But inboth widget manufacturing and administrative work, something else can alsoincreasetheproductivityoftheblackbox.Thisiscalledworksimplification.Toget leverage this way, you first need to create a flow chart of the productionprocess as it exists.Every single stepmust be shownon it; no step should beomittedinordertoprettythingsuponpaper.Second,countthenumberofstepsintheflowchartsothatyouknowhowmanyyoustartedwith.Third,setaroughtarget for reduction of the number of steps. In the first round of worksimplification,ourexperienceshowsthatyoucanreasonablyexpecta30to50percentreduction.To implement the actual simplification, youmust questionwhy each step is

performed.Typically,youwillfindthatmanystepsexistinyourworkflowfornogood reason.Often theyare thereby traditionorbecause formalprocedureordains it, and nothing practical requires their inclusion.Remember, the “visafactory”atourembassy inBritaindidn’t reallyhave toprocess100percentofthe applicants. So no matter what reason may be given for a step, you mustcriticallyquestioneachandthrowoutthosethatcommonsensesaysyoucandowithout. We found that in a wide range of administrative activities at Intel,substantial reduction—about 30 percent—could be achieved in the number of

stepsrequiredtoperformvarioustasks.Of course, the principle ofwork simplification is hardly new in thewidget

manufacturing arts. In fact, this is one of the things industrial engineers havebeendoingforahundredyears.Buttheapplicationoftheprincipletoimprovetheproductivityofthe“softprofessions”—theadministrative,professional,andmanagerialworkplace—isnewandslowtotakehold.Themajorproblemtobeovercomeisdefiningwhattheoutputofsuchworkisorshouldbe.Aswewillsee,intheworkofthesoftprofessions,itbecomesverydifficult todistinguishbetween output and activity. And as noted, stressing output is the key toimprovingproductivity,while looking to increaseactivitycanresult in just theopposite.

II

ManagementIsaTeamGame

3ManagerialLeverage

WhatIsaManager’sOutput?

Iaskedagroupofmiddlemanagersjustthatquestion.Igottheseresponses:

judgmentsandopinionsdirectionallocationofresourcesmistakesdetectedpersonneltrainedandsubordinatesdevelopedcoursestaughtproductsplannedcommitmentsnegotiated

Do these things really constitute the output of amanager? I don’t think so.Theyare instead activities, or descriptionsofwhatmanagersdo as they try tocreateafinalresult,oroutput.What,then,isamanager’soutput?AtIntel,ifsheisinchargeofawaferfabricationplant,heroutputconsistsofcompleted,high-quality, fully processed silicon wafers. If he supervises a design group, hisoutputconsistsofcompleteddesignsthatworkcorrectlyandarereadytogointomanufacturing.Ifamanageristheprincipalofahighschool,heroutputwillbetrainedandeducatedstudentswhohaveeithercompletedtheirschoolingorarereadytomoveontothenextyearoftheirstudies.Ifamanagerisasurgeon,hisoutput is a fully recovered, healed patient. We can sum matters up with thepropositionthat:

Amanager’soutput=

Theoutputofhisorganization

+

The output of the neighboringorganizationsunderhisinfluence

Why?Becausebusinessandeducationandevensurgeryrepresentworkdonebyteams.Amanagercandohis“own”job,hisindividualwork,anddoitwell,butthat

doesnotconstitutehisoutput.Ifthemanagerhasagroupofpeoplereportingtohim or a circle of people influenced by him, the manager’s output must bemeasured by the output created by his subordinates and associates. If themanager is a knowledge specialist, a know-how manager, his potential forinfluencing “neighboring” organizations is enormous. The internal consultantwhosuppliesneededinsighttoagroupstrugglingwithaproblemwillaffectthework and the output of the entire group. Similarly, if a lawyer acquires aregulatorypermit foradrugcompany,hewill release the flowof the resultofmany years of research at that company to the public.Or amarketing analystwho reviews mountains of product, market, and competitive information,analyzesmarket research, andmakes fact-finding visits can directly affect theoutputofmany“neighboring”organizations.Hisinterpretationsofthedataandhis recommendationswill perhaps guide the activities for thewhole company.Thus, thedefinitionof“manager”shouldbebroadened:individualcontributorswhogatheranddisseminateknow-howand informationshouldalsobeseenasmiddlemanagers,becausetheyexertgreatpowerwithintheorganization.Butthekeydefinitionhereisthattheoutputofamanagerisaresultachieved

by a group either under her supervision or under her influence. While themanager’s own work is clearly very important, that in itself does not createoutput.Herorganizationdoes.Byanalogy,acoachoraquarterbackalonedoesnotscoretouchdownsandwingames.Entireteamswiththeirparticipationandguidanceanddirectiondo.Leaguestandingsarekeptbyteam,notbyindividual.Business—andthismeansnotjustthebusinessofcommercebutthebusinessofeducation, the business of government, the business of medicine—is a teamactivity.And,always,ittakesateamtowin.Itisimportanttounderstandthatamanagerwillfindhimselfengaginginan

array of activities in order to affect output.As themiddlemanagers I queriedsaid, a manager must form opinions and make judgments, he must providedirection, hemust allocate resources, hemust detectmistakes, and so on.Allthesearenecessary toachieveoutput.Butoutputandactivityarebynomeans

thesamething.Considermy ownmanagerial role.As president of a company, I can affect

outputthroughmydirectsubordinates—groupgeneralmanagersandotherslikethem—by performing supervisory activities. I can also influence groups notundermy direct supervision bymaking observations and suggestions to thosewhomanagethem.Bothtypesofactivitywill,Ihope,contributetomyoutputasamanagerbycontributingtotheoutputofthecompanyasawhole.Iwasonceasked by a middle manager at Intel how I could teach in-plant courses, visitmanufacturingplants,concernmyselfwiththeproblemsofpeopleseverallevelsremovedfrommeintheorganization,andstillhavetimetodomyjob.Iaskedhimwhathethoughtmyjobwas.Hethoughtforamoment,andthenansweredhisownquestion,“Iguessthosethingsareyourjobtoo,aren’tthey?”Theyareabsolutelymyjob—notmyentirejob,butpartofit,becausetheyhelpaddtotheoutputofIntel.Let me give another example. Cindy, an engineer at Intel, supervises an

engineeringgroupinawaferfabricationplant.Shealsospendssomeofhertimeasamemberofanadvisorybodythatestablishesstandardproceduresbywhichall the plants throughout the company perform a certain technical process. Inbothroles,Cindycontributestotheoutputofthewaferfabricationplants.Asasupervisingengineer,sheperformsactivitiesthatincreasetheoutputoftheplantinwhichsheworks;asamemberoftheadvisorybody,sheprovidesspecializedknowledge that will influence and increase the output of “neighboringorganizations”—alltheotherIntelwaferfabricationplants.Let’sreferagaintoourblackbox.Ifthemachinerywithinanorganizationcan

becomparedtoaseriesofgears,wecanvisualizehowamiddlemanageraffectsoutput. In times of crisis, he provides power to the organization.When thingsaren’t working as smoothly as they should, he applies a bit of oil. And, ofcourse,heprovidesintelligencetothemachinetodirectitspurpose.

“Daddy,WhatDoYouReallyDo?”

Mostofushavehadtostruggletoanswerthatquestion.Whatweactuallydoisdifficulttopindownandsumup.Muchofitoftenseemssoinconsequentialthatour position in the business hardly seems justified. Part of the problem herestemsfromthedistinctionbetweenouractivities,whichiswhatweactuallydo,and our output, which is what we achieve. The latter seems important,significant, andworthwhile. The former often seems trivial, insignificant, and

messy.Butasurgeonwhoseoutputisacuredpatientspendshistimescrubbingandcuttingandsuturing,andthishardlysoundsveryrespectableeither.Tofindoutwhatwemanagersreallydo,let’stakealookatoneofmybusier

days, shown in the table below. Here I describe the activity in which I wasengaged, explain it a bit, and categorize it into typeswe shall examine in thebalanceofthechapter.

ADayfromMyLife

TimeandActivity Explanation(TypeofActivity)

8:00–8:30 Met with a managerwho had submitted hisresignation to leave foranothercompany.

Ilistenedtohisreasons(information-gathering),felt he could be turned around and saved forIntel. Encouraged him to talk to certain othermanagers about a career change (nudge), anddecided to pursue thismatter with themmyself(decision-making).

Incomingtelephonecallfromacompetitor.

Call was ostensibly about a meeting of anindustry-wide society, but in reality he wasfeelingouthowIsawbusinessconditions. Ididthesame.(Information-gathering.)

8:30–9:00 Read mail from thepreviousafternoon.

Iscribbledmessagesonabouthalfofit,someofwhich were expressions of encouragement ordisapproval, others exhortations to take certaintypesofaction(nudges).Onewasthedenialofarequesttoproceedwithaparticularsmallproject(decision-making). (Of course, information-gatheringtookplaceinallofthesecases,too.)

9:00–12:00 Executive StaffMeeting (a regularweekly meeting of thecompany’s senior

management). Subjectscovered at thisparticularone:—Review of the priormonth’sincomingorderandshipmentrates.

(Information-gathering)

—Discussion to setpriorities for theupcoming annualplanningprocess.

(Decision-making)

—Review of the statusof a major marketingprogram (scheduledsubject).

This came about through a prior decision thatthis programwas faltering and required review.We found it was doing a little bit better thanbefore (information-gathering), but thepresentation still elicited a lot of comments andsuggestions (nudges) from various members oftheaudience.

—Reviewofaprogramto reduce themanufacturing cycletime of a particularproduct line (scheduledsubject).

Thepresentation indicated that theprogramwasingood shape. (It representedonly information-gathering;nofurtheractionwasstimulated.)

12:00–1:00 Lunch in the companycafeteria.

I happened to sit withmembers of our trainingorganization, who complained about thedifficultytheyhadingettingmeandotherseniormanagerstoparticipateintrainingatourforeignlocations(information-gathering).Thiswasnewstome.Imadeanote tofollowupwithmyownschedule,aswellaswithmystaff,andtonudgethem into doing a better job supporting theforeigntrainingprogram.

1:00–2:00 Meeting regarding aspecific product-quality

The bulk of the meeting involved gettingsufficientinformationonthestatusoftheproduct

problem. and the corrective action that had beenimplemented (information-gathering). Themeetingendedinadecisionmadebythedivisionmanager, with my concurrence, to resumeshipmentoftheproduct.

2:00–4:00 Lecture at ouremployee orientationprogram.

This is a program inwhich seniormanagementgives all professional employees a presentationdescribing the objectives, history, managementsystems, etc., of the company and its majorgroups. Iamthe first lecturer in theseries.Thisclearlyrepresentedinformation-giving,andIwasa role model not only in communicating theimportanceweplaceontraining,butalso,bymyhandling of questions and comments, inrepresenting, in living form, someof the valuesof the company.The nature of the questions, atthesametime,gavemeafeelingfortheconcernsand understanding level of a large number ofemployees towhomIwouldnototherwisehaveaccess. So this also represented information-gathering,characteristicofthe“visit”typeinitsefficiency.

4:00–4:45 In the office, returningphonecalls.

I disapproved granting a compensation increaseto a particular employee, which I thought wasway outside of the norm (clearly a decision). Idecided to conduct a meeting with a group ofpeople todecidewhatorganizationwouldmoveto a new sitewewere opening in another state.(Thiswas adecision to hold a decision-makingmeeting.)

4:45–5:00 Metwithmyassistant. Discussedavarietyofrequestsformytimefora

number of meetings in the upcoming week.Suggested alternatives where I decided not toattend.

5:00–6:15 Read the day’s mail,including progressreports.

As with the morning’s mail reading, this wasinformation-gathering,interspersedwithnudginganddecision-makingthroughmyannotationsandmessagesscribbledonmuchofit.

Whenyoulookatwhathappened,youwon’tseeanyobviouspatterns.Idealtwiththingsinseeminglyrandomfashion.Mywife’sreactiontomydaywasthatitlookedverymuchlikeoneofherown.Shewasrightinnotingasimilarity.MydayalwaysendswhenI’mtiredandreadytogohome,notwhenI’mdone.Iamnever done. Like a housewife’s, a manager’s work is never done. There isalwaysmore to be done,more that should be done, alwaysmore than can bedone.Amanagermust keepmany balls in the air at the same time and shift his

energy and attention to activities that will most increase the output of hisorganization. In otherwords, he shouldmove to the pointwhere his leveragewillbethegreatest.Asyoucansee,muchofmydayisspentacquiringinformation.Andasyou

canalsosee,Iusemanywaystoget it.Ireadstandardreportsandmemosbutalso get information ad hoc. I talk to people inside and outside the company,managersatotherfirmsorfinancialanalystsormembersofthepress.Customercomplaints, both external and internal, are also a very important source ofinformation. For example, the Intel training organization, which I serve as aninstructor, is an internal customer ofmine. To cutmyself off from the casualcomplaints of people in that groupwould be amistake because Iwouldmissgettinganevaluationofmyperformanceasan internal“supplier.”Peoplealsotellus thingsbecause theywantus todosomethingfor them; toadvance theircase,theywillsometimesshoweruswithusefulinformation.Thisissomethingweshouldremember,apartfromwhetherwedoastheyask.Ihavetoconfess that theinformationmostuseful tome,andIsuspectmost

useful toallmanagers, comes fromquick,oftencasualverbal exchanges.Thisusuallyreachesamanagermuchfasterthananythingwrittendown.Andusuallythemoretimelytheinformation,themorevaluableitis.So why are written reports necessary at all? They obviously can’t provide

timely information. What they do is constitute an archive of data, help to

validateadhocinputs,andcatch,insafety-netfashion,anythingyoumayhavemissed. But reports also have another totally different function. As they areformulatedandwritten,theauthorisforcedtobemoreprecisethanhemightbeverbally.Hencetheirvaluestemsfromthedisciplineandthethinkingthewriterisforcedtoimposeuponhimselfasheidentifiesanddealswithtroublespotsinhis presentation. Reports are more amedium of self-discipline than a way tocommunicateinformation.Writingthereportisimportant;readingitoftenisnot.Therearemanyparallels to this.Aswewill see later, thepreparationofan

annualplan is in itself theend,not the resultingboundvolume.Similarly,ourcapitalauthorizationprocess itself is important,not theauthorization itself.Toprepareand justifyacapitalspendingrequest,peoplego througha lotofsoul-searching analysis and juggling, and it is thismental exercise that is valuable.Theformalauthorizationisusefulonlybecauseitenforcesthedisciplineoftheprocess.To improve and maintain your capacity to get information, you have to

understand the way it comes to you. There’s a hierarchy involved. Verbalsourcesarethemostvaluable,butwhattheyprovideisalsosketchy,incomplete,andsometimesinaccurate,likeanewspaperheadlinethatcangiveyouonlythegeneralideaofastory.Aheadlinecan’tgiveanyofthedetailsandmightevengive you a distorted idea of what the real story is. So you then read thenewspaperarticleitselftofindoutwho,what,where,why,andhow.Afterthis,you should have some reiteration and perspective, which can be compared toreading a news magazine or even a book. Each level in your informationhierarchy is important, and you can rely on none alone. Though the mostthorough information might come from the news magazine, you do not, ofcourse, want to wait a full week after an event to find out about it. Yourinformation sources should complement one another, and also be redundantbecausethatgivesyouawaytoverifywhatyou’velearned.There is an especially efficient way to get information, much neglected by

mostmanagers.That is to visit a particular place in the company andobservewhat’sgoingon there.Whyshouldyoudo this?Thinkofwhathappenswhensomebody comes to see a manager in his office. A certain stop-and-startdynamicsoccurswhenthevisitorsitsdown,somethingsociallydictated.Whileatwo-minutekernelofinformationisexchanged,themeetingoftentakesahalfhour.Butifamanagerwalksthroughanareaandseesapersonwithwhomhehasatwo-minuteconcern,hecansimplystop,coverit,andbeonhisway.Dittofor thesubordinatewhenheinitiatesconversation.Accordingly,suchvisitsare

anextremelyeffectiveandefficientwaytotransactmanagerialbusiness.Thenwhyaretheyunderutilized?Becauseoftheawkwardnessthatmanagers

feelaboutwalkingthroughanareawithoutaspecifictaskinmind.AtIntelwecombat this problem by using programmed visitsmeant to accomplish formaltasks,butwhichalsosetthestageforadhocmini-transactions.Forexample,weaskourmanagerstoparticipatein“Mr.Clean”inspections,inwhichtheygotoapartof the company that theynormallywouldn’tvisit.Themanagers examinethehousekeeping,thearrangementofthings,thelabs,andthesafetyequipment,and in so doing spend an hour or so browsing around and getting acquaintedwiththingsfirsthand.Ascanbeseenfrommyschedule,amanagernotonlygathersinformationbut

is also a source of it.Hemust convey his knowledge tomembers of his ownorganizationandtoothergroupsheinfluences.Beyondrelayingfacts,amanagermustalsocommunicatehisobjectives,priorities,andpreferencesastheybearonthewaycertaintasksareapproached.Thisisextremelyimportant,becauseonlyifthemanagerimpartsthesewillhissubordinatesknowhowtomakedecisionsthemselves that will be acceptable to the manager, their supervisor. Thus,transmittingobjectivesandpreferredapproachesconstitutesakeytosuccessfuldelegation. As we will see later, a shared corporate culture becomesindispensable to a business. Someone adhering to the values of a corporateculture—an intelligent corporate citizen—will behave in consistent fashionunder similar conditions,whichmeans thatmanagers don’t have to suffer theinefficiencies engendered by formal rules, procedures, and regulations that aresometimesusedtogetthesameresult.Thethirdmajorkindofmanagerialactivity,ofcourse,isdecision-making.To

besure,onceinawhilewemanagersinfactmakeadecision.Butforeverytimethathappens,weparticipateinthemakingofmany,manyothers,andwedothatinavarietyofways.Weprovidefactualinputsorjustofferopinions,wedebatetheprosandconsofalternativesandtherebyforceabetterdecisiontoemerge,we review decisions made or about to be made by others, encourage ordiscouragethem,ratifyorvetothem.Just howdecisions should bemade,we’ll talk about later.Meanwhile, let’s

saythatdecisionscanbeseparatedintotwokinds.Theforward-lookingsortaremade, for example, in the capital authorization process. Here we allocate thefinancial resources of the company among various future undertakings. Thesecond type ismadeaswerespond toadevelopingproblemoracrisis,which

can either be technical (a quality control problem, for example) or involvepeople(talkingsomebodyoutofquitting).It’s obvious that your decision-making depends finally on how well you

comprehendthefactsandissuesfacingyourbusiness.Thisiswhyinformation-gathering is so important in a manager’s life. Other activities—conveyinginformation,makingdecisions,andbeingarolemodelforyoursubordinates—areallgovernedbythebaseofinformationthatyou,themanager,haveaboutthetasks,theissues,theneeds,andtheproblemsfacingyourorganization.Inshort,information-gatheringisthebasisofallothermanagerialwork,whichiswhyIchoosetospendsomuchofmydaydoingit.Youoftendothingsattheofficedesignedtoinfluenceeventsslightly,maybe

making a phone call to an associate suggesting that a decision be made in acertainway,orsendinganoteoramemothatshowshowyouseeaparticularsituation, ormaking a comment during an oral presentation. In such instancesyoumaybeadvocatingapreferredcourseofaction,butyouarenotissuinganinstruction or a command. Yet you’re doing something stronger than merelyconveyinginformation.Let’scallit“nudging”becausethroughityounudgeanindividual or ameeting in thedirectionyouwould like.This is an immenselyimportantmanagerialactivityinwhichweengageallthetime,anditshouldbecarefully distinguished from decision-making that results in firm, cleardirectives. In reality, for every unambiguous decision we make, we probablynudgethingsadozentimes.Finally, somethingmore subtlepervades thedayof allmanagers.Whilewe

moveabout,doingwhatweregardasourjobs,wearerolemodelsforpeopleinourorganization—oursubordinates,ourpeers,andevenoursupervisors.Muchhasbeensaidandwrittenaboutamanager’sneedtobealeader.Thefactis,nosinglemanagerialactivitycanbesaidtoconstituteleadership,andnothingleadsas well as example. By this I mean something straightforward. Values andbehavioral norms are simply not transmitted easily by talk or memo, but areconveyedveryeffectivelybydoinganddoingvisibly.Allmanagersneedtoactsothattheycanbeseenexertinginfluence,butthey

shoulddosointheirownway.Someofusfeelcomfortabledealingwithlargegroupsandtalkingaboutourfeelingsandvaluesopenlyinthatfashion.Othersprefer working one-on-one with people in a quieter, more intellectualenvironment. These and other styles of leadership will work, but only if werecognizeandconsciouslystresstheneedforustoberolemodelsforpeoplein

ourorganization.Don’tthinkforamomentthatthewayI’vedescribedleadershipappliesonly

to largeoperations.An insuranceagent inasmallofficewhocontinually talkswith personal friends on the phone imparts a set of values about permissibleconduct toeveryoneworkingforhim.A lawyerwhoreturns tohisofficeafterlunchalittledrunkdoesthesame.Ontheotherhand,asupervisorinacompany,large or small, who takes hiswork seriously exemplifies to his associates themostimportantmanagerialvalueofall.A great deal of a manager’s work has to do with allocating resources:

manpower,money,andcapital.But thesinglemost importantresourcethatweallocate from one day to the next is our own time. In principlemoremoney,moremanpower, or more capital can always bemade available, but our owntime is the one absolutely finite resourcewe each have. Its allocation and usethereforedeserveconsiderableattention.Howyouhandleyourown time is, inmyview,thesinglemostimportantaspectofbeingarolemodelandleader.As you can see, in a typical day of mine one can count some twenty-five

separate activities in which I participated, mostly information-gathering and -giving,butalsodecision-makingandnudging.Youcanalsoseethatsometwothirdsofmytimewasspentinameetingofonekindoranother.BeforeyouarehorrifiedbyhowmuchtimeIspendinmeetings,answeraquestion:whichoftheactivities—information-gathering, information-giving, decision-making,nudging, and being a rolemodel—could I have performed outside ameeting?The answer is practically none.Meetings provide an occasion for managerialactivities. Getting together with others is not, of course, an activity—it is amedium. You as a manager can do your work in a meeting, in a memo, orthrougha loudspeaker for thatmatter.Butyoumustchoose themosteffectivemediumforwhatyouwanttoaccomplish,andthatistheonethatgivesyouthegreatestleverage.Moreaboutmeetingslater.

LeverageofManagerialActivity

We’ve established that the output of a manager is the output of the variousorganizations under his control and his influence.What can a manager do toincreasehisoutput?Tofindout,let’slookattheconceptofleverage.Leverageis the measure of the output generated by any given managerial activity.Accordingly, managerial output can be linked to managerial activity by theequation:

ManagerialOutput=

Outputoforganization

=

L1×A1+L2×A2+…

Thisequationsaysthatforeveryactivityamanagerperforms—A1,A2,andsoon—theoutputoftheorganizationshouldincreasebysomedegree.Theextenttowhich thatoutput is thereby increased isdeterminedby the leverageof thatactivity—L1,L2,andsoon.Amanager’soutputisthusthesumoftheresultofindividualactivitieshavingvaryingdegreesofleverage.Clearlythekeytohighoutputmeansbeingsensitivetotheleverageofwhatyoudoduringtheday.Managerial productivity—that is, the output of a manager per unit of time

worked—canbeincreasedinthreeways:

1.Increasingtheratewithwhichamanagerperformshisactivities,speedinguphiswork.

2.Increasingtheleverageassociatedwiththevariousmanagerialactivities.3. Shiftingthemixofamanager’sactivitiesfromthosewithlowertothose

withhigherleverage.

Letusconsiderfirsttheleverageofvarioustypesofmanagerialwork.

HIGH-LEVERAGEACTIVITIES

Thesecanbeachievedinthreebasicways:

•Whenmanypeopleareaffectedbyonemanager.•Whenaperson’sactivityorbehavioroveralongperiodoftimeisaffectedbyamanager’sbrief,well-focusedsetofwordsoractions.

•Whenalargegroup’sworkisaffectedbyanindividualsupplyingaunique,keypieceofknowledgeorinformation.

The first is the most obvious example. Consider Robin, an Intel financemanager,responsibleforsettinguptheannualfinancialplanningprocessforthecompany.WhenRobindefinesinadvanceexactlywhatinformationneedstobegatheredandpresentedateachstageoftheplanningprocessandlaysoutwhois

responsible forwhat, she directly affects the subsequentwork of perhaps twohundredpeoplewhoparticipate in theplanningprocess.Byspendingacertainamount of time in advance of the planning activities, Robin will help toeliminate confusion andambiguity for a largepopulationofmanagersover anextendedperiodoftime.Consequently,herworkcontributestotheproductivityoftheentireorganizationandclearlyhasgreatleverage,leveragethatdepends,however, on when it is performed. Work done in advance of the planningmeetingobviouslyhasgreat leverage. IfRobinhas to scramble later tohelp amanagerdefineguidelinesandmilestones,herworkwillclearlyhavemuchlessleverage.Another example of leverage that depends on timely action iswhat you do

whenyoulearnthatavaluedsubordinatehasdecidedtoquit.Insuchacase,youmust direct yourself to the situation immediately if you want to change theperson’smind.Ifyouputitoff,allyourchancesarelost.Thustomaximizetheleverageofhisactivities,amanagermustkeeptimeliness,whichisoftencritical,firmlyinmind.Leverage can also be negative. Some managerial activities can reduce the

outputof anorganization. Imean somethingvery simple.Suppose I amakeyparticipantatameetingandIarriveunprepared.NotonlydoIwastethetimeofthe people attending themeeting because ofmy lack of preparation—a directcostofmycarelessness—butIdeprivetheotherparticipantsoftheopportunitytousethattimetodosomethingelse.Eachtimeamanagerimpartshisknowledge,skills,orvaluestoagroup,his

leverage is high, asmembers of the groupwill carrywhat they learn tomanyothers.Butagain,leveragecanbepositiveornegative.AnexampleofleveragethatIhopeishighandpositiveismytalkintheorientationcourse.Duringthetwo hours I have, I try to impart a great deal of information about Intel—itshistory, its objectives, its values, its style—to a group of two hundred newemployees. Besides what I say specifically, my approach toward answeringquestionsandmyconduct ingeneral communicateourwayofdoing things totheseemployeeswhentheyaremostimpressionable.Hereisanotherexampleofthiskindofleverage.Totrainagroupofsalesmen,

Barbara, an Intel marketing engineer, sets out to teach them what theorganization’s products are. If she does her job well, the salespeople will bebetterequippedtoselltheline.Ifshedoesitpoorly,greatandobviousdamageisdone.

A final, less formal, example here: Cindy, as you recall, is amember of atechnicalcoordinatingbodyinwhichshetriestodisseminateherunderstandingofaspecifictechnologytoallofthecompany’smanufacturinggroups.Ineffect,she uses the coordinating body as an informal training vehicle to effect highleverageonhercounterpartsinneighboringIntelorganizations.Amanagercanalsoexerthighleveragebyengaginginanactivitythattakes

himonlyashorttime,butthataffectsanotherperson’sperformanceoveralongtime.A performance review represents a good example of this.With the fewhours’workthatamanagerspendspreparinganddeliveringthereview,hecanaffecttheworkofitsrecipientenormously.Heretooamanagercanexerteitherpositive or negative leverage. A subordinate can be motivated and evenredirected in his efforts, or the review can discourage and demoralize him forwhoknowshowlong.Anotherseeminglytrivialpieceofwork—creatingaticklerfile—canimprove

dailyworksignificantlyforalongtime.Settingupthesimplemechanicalaidisaone-timeactivity, yet it is likely to improve theproductivityof themanagerwhousesitindefinitely.Thustheleveragehereisvery,veryhigh.Examplesofhighnegative leverageabound.Aftergoing through theannual

planningprocess,anIntelmanagersawthat,inspiteofsuccessfulcostreductionefforts intheprioryear,hisdivisionwasstillnotgoingtomakeanymoneyinthecomingyear.Themanagerbecamedepressed.Thoughhedidn’trealizeit,healmost immediately began to affect people around him and soon depressionspreadthroughouthisorganization.Hesnappedoutofitonlywhensomeoneonhis staff finally toldhimwhathewasdoing to thepeopleunderhim.Anotherexampleiswaffling,whenamanagerputsoffadecisionthatwillaffecttheworkof other people. In effect, the lack of a decision is the same as a negativedecision; no green light is a red light, and work can stop for a wholeorganization.Both the depressed and the waffling manager can have virtually unlimited

negativeleverage.Ifpeoplearebadlyaffectedbyapoorsalestrainingeffort,thesituation can be handled by retraining the group. But the negative leverageproduced by depression and waffling is very hard to counter because theirimpactonanorganizationisbothsopervasiveandsoelusive.Managerial meddling is also an example of negative leverage. This occurs

when a supervisor uses his superior knowledge and experience of asubordinate’s responsibilities to assume command of a situation rather than

letting the subordinate work things through himself. For example, if a seniormanager sees an indicator showing an undesirable trend and dictates to theperson responsible a detailed set of actions to be taken, that is managerialmeddling. In general, meddling stems from a supervisor exploiting too muchsuperior work knowledge (real or imagined). The negative leverage producedcomes from the fact that after being exposed to many such instances, thesubordinatewillbegintotakeamuchmorerestrictedviewofwhatisexpectedofhim,showingless initiative insolvinghisownproblemsandreferring theminstead to his supervisor. Because the output of the organization willconsequentlybereducedinthelongrun,meddlingisclearlyanactivityhavingnegativemanagerialleverage.The third kind of managerial activity with high leverage is exercised by a

personwithuniqueskillsandknowledge.OnesuchpersonisanIntelmarketingengineer responsible for setting prices for the product line. Hundreds ofsalespeopleinthefieldcanbenegativelyaffectedifpricesaresettoohigh:nomatterhowhardtheymaytry,theywon’tbeabletogetanybusiness.Ofcourse,ifthepricesweresettoolow,wewouldbegivingmoneyaway.Take another example. An Intel development engineer who has uniquely

detailed knowledge of a particular manufacturing process effectively controlshowitisused.Sincetheprocesswilleventuallyprovidethefoundationforthework of many product designers all over the company, the leverage thedevelopmentengineerexertsisenormous.Thesameistrueforageologistinanoilcompanyoranactuaryinaninsurancefirm.Allarespecialistswhoseworkisimportant for the work of their organization at large. The person whocomprehends the critical facts or has the critical insights—the “knowledgespecialist” or the “know-how manager”—has tremendous authority andinfluenceontheworkofothers,andthereforeveryhighleverage.Theartofmanagementliesinthecapacitytoselectfromthemanyactivities

of seemingly comparable significance the one or two or three that provideleveragewellbeyondtheothersandconcentrateonthem.Forme,payingcloseattentiontocustomercomplaintsconstitutesahigh-leverageactivity.Asidefrommakinga customerhappy, thepursuit tends toproduce important insights intothe workings of my own operation. Such complaints may be numerous, andthoughallofthemneedtobefollowedupbysomeone,theydon’tallrequireorwouldn’tallbenefitfrommypersonalattention.Whichoneoutoftenortwentycomplaintstodiginto,analyze,andfollowupiswhereartcomesintotheworkofamanager.Thebasisofthatartisanintuitionthatbehindthiscomplaintand

nottheotherlurkmanydeeperproblems.

DELEGATIONASLEVERAGE

Becausemanagerial time has a hierarchy of values, delegation is an essentialaspectofmanagement.The“delegator”and“delegatee”must shareacommoninformationbaseandacommonsetofoperationalideasornotionsonhowtogoabout solving problems, a requirement that is frequently notmet.Unless bothpartiesshare therelevantcommonbase, thedelegateecanbecomeaneffectiveproxyonlywithspecific instructions.As inmeddling,wherespecificactivitiesareprescribedindetail,thisproduceslowmanagerialleverage.Picturethis.Iamyoursupervisor,andIwalkovertoyouwithpencilinhand

and tell you to take it.You reach for the pencil, but Iwon’t let go. So I say,“Whatiswrongwithyou?Whycan’tIdelegatethepenciltoyou?”Weallhavesomethingsthatwedon’treallywanttodelegatesimplybecausewelikedoingthemandwouldrathernotletgo.Foryourmanagerialeffectiveness,thisisnottoobadso longas it isbasedonaconsciousdecision thatyouwillholdon tocertain tasks that you enjoyperforming, even thoughyou could, if you chose,delegate them.But be sure to know exactlywhat you’re doing, and avoid thecharade of insincere delegation, which can produce immense negativemanagerialleverage.Givenachoice,shouldyoudelegateactivitiesthatarefamiliartoyouorthose

that aren’t? Before answering, consider the following principle: delegationwithoutfollow-throughisabdication.Youcanneverwashyourhandsofatask.Evenafteryoudelegateit,youarestillresponsibleforitsaccomplishment,andmonitoringthedelegatedtaskistheonlypracticalwayforyoutoensurearesult.Monitoring is not meddling, but means checking to make sure an activity isproceedingin linewithexpectations.Becauseit iseasier tomonitorsomethingwith which you are familiar, if you have a choice you should delegate thoseactivitiesyouknowbest.Butrecallthepencilexperimentandunderstandbeforethefactthatthiswillverylikelygoagainstyouremotionalgrain.Please turnback to the tableofmyday’sactivitieson thispage.During the

executivestaffmeetingweheardtwofollow-uppresentations,oneonthestatusofanextremelyimportantmarketingprogramandtheotherontheprogressofaprogram aimed at reducingmanufacturing throughput times.Both reviews areexamplesofmonitoring.Earlier,wehadassignedeachtoamiddlemanagerandmadesurethesemanagersandtheseniorstaffagreedaboutwhat theprograms

were to be. Themiddlemanagers thenwent about their business expecting toreportbacktotheexecutivestaff,thebodythatdelegatedtheprogramstothem.Monitoringtheresultsofdelegationresemblesthemonitoringusedinquality

assurance. We should apply quality assurance principles and monitor at thelowest-added-value stage of the process. For example, review rough drafts ofreports that you have delegated; don’twait until your subordinates have spenttime polishing them into final formbefore you find out that you have a basicproblem with the contents. A second principle applies to the frequency withwhich you check your subordinates’ work. A variable approach should beemployed, using different sampling schemes with various subordinates; youshould increase or decrease your frequency depending on whether yoursubordinateisperforminganewlydelegatedtaskoronethathehasexperiencehandling.Howoftenyoumonitorshouldnotbebasedonwhatyoubelieveyoursubordinatecandoingeneral,butonhisexperiencewithaspecifictaskandhispriorperformancewith it—his task-relevantmaturity,somethingI’ll talkaboutin detail later. As the subordinate’s work improves over time, you shouldrespondwithacorrespondingreductionintheintensityofthemonitoring.To use quality assurance principles effectively, themanager should only go

intodetailsrandomly,justenoughtotrytoensurethatthesubordinateismovingaheadsatisfactorily.Tocheck intoall thedetailsof adelegated taskwouldbelikequalityassurancetesting100percentofwhatmanufacturingturnedout.Makingcertaintypesofdecisionsissomethingmanagersfrequentlydelegate

to subordinates. How is this best done? By monitoring their decision-makingprocess.Howdoyoudothat?Let’sexaminewhatIntelgoesthroughtoapproveacapitalequipmentpurchase.Weaskasubordinate to thinkthroughtheentirematter carefullybeforepresentinga request for approval.And tomonitorhowgoodhisthinkingis,weaskhimquitespecificquestionsabouthisrequestduringa review meeting. If he answers them convincingly, we’ll approve what hewants. This technique allows us to find out how good the thinking iswithouthavingtogothroughitourselves.

IncreasingManagerialActivityRate:SpeedingUptheLine

Ofcourse,themostobviouswaytoincreasemanagerialoutputistoincreasetherate,orspeed,ofperformingwork.Therelationshiphereis:

whereListheleverageoftheactivity.The most common approach to increasing a manager’s productivity—his

output over time—has been time-management techniques,which try to reducethedenominatoronbothsidesofthisequation.Anynumberofconsultantswilltellamanagerthatthewaytohigherproductivityistohandleapieceofpaperonlyonce,toholdonlystand-upmeetings(whichwillpresumablybeshort),andtoturnhisdesksothathepresentshisbacktothedoor.These time-management suggestions can be improved upon, I think, by

applying our production principles. First, we must identify our limiting step:what is the“egg” inourwork? Inamanager’s life some things reallyhave tohappenonaschedulethatisabsolute.Forme,anexampleistheclassIteach.Iknowwhen it is going tomeet, and I know Imust prepare for it.There is no“give”inthetimehere,becauseovertwohundredstudentswillbeexpectingme.Accordingly, I have to create offsets and schedulemy otherwork around thislimiting step. In short, ifwedeterminewhat is immovable andmanipulate themoreyieldingactivitiesaroundit,wecanworkmoreefficiently.Asecondproductionprinciplewecanapply tomanagerialwork isbatching

similar tasks.Anymanufacturingoperationrequiresacertainamountofset-uptime.Soformanagerialworktoproceedefficiently,weshouldusethesameset-up effort to apply across a group of similar activities. Think about ourcontinuous egg-boiler, which was installed to produce fine-quality, identical,three-minute eggs. Shouldwe now decide to serve our customers four-minuteeggs,wewouldhavetoslowdowntheconveyorbeltmovingthemthroughthehotwater.Theadjustmenttakestime:notonlydoweadjustnutsandboltsonthemachine,wealsohavetoinspectthequalityofthefour-minuteeggsbysamplingafewofthem.Set-up time hasmany parallels in managerial work. For example, once we

have prepared a set of illustrations for a training class, we will obviouslyincreaseourproductivity ifwecanuse the same set over andover againwithotherclassesorgroups.Similarly,ifamanagerhasanumberofreportstoreadoranumberofperformancereviewstoapprove,heshouldsetasideablockoftimeanddoabatchofthemtogether,oneaftertheother,tomaximizetheuseofthementalset-uptimeneededforthetask.

Whatmakesrunningafactorydifferentfromrunningajobshop?Thelatterisprepared to service any customer who drops in; the owner handles the jobrequiredandmovesontothenextone.Afactory,ontheotherhand,isusuallyrunbyforecastandnotbyindividualorder.Frommyexperiencealargeportionofmanagerialworkcanbeforecasted.Accordingly,forecastingthosethingsyoucanandsettingyourselfuptodothemisonlycommonsenseandanimportantway to minimize the feeling and the reality of fragmentation experienced inmanagerial work. Forecasting and planning your time around key events areliterallylikerunninganefficientfactory.Whatisthemediumofamanager’sforecast?Itissomethingverysimple:his

calendar.Mostpeopleusetheircalendarsasarepositoryof“orders”thatcomein. Someone throws an order to a manager for his time, and it automaticallyshowsuponhiscalendar.This ismindlesspassivity.Togainbettercontrolofhis time, themanager shouldusehiscalendarasa“production”planning tool,takinga firminitiative toschedulework that isnot time-criticalbetween those“limitingsteps”intheday.Another production principle can be applied here. Because manufacturing

people trust their indicators, they won’t allow material to begin its journeythroughthefactoryif theythinkit isalreadyoperatingatcapacity.If theydid,materialmight go halfway through and back up behind a bottleneck. Instead,factorymanagerssay“no”attheoutsetandkeepthestartlevelfromoverloadingthe system. Other kinds of managers find this hard to apply because theirindicatorsofcapacityarenotaswellestablishedornotasbelievable.Howmuchtime do you need to read your mail, to write your reports, to meet with acolleague?Youmaynotknowprecisely,butyousurelyhaveafeelforthetimerequired.Andyoushouldexploitthatsensetoscheduleyourwork.To use your calendar as a production-planning tool, you must accept

responsibilityfortwothings:

1. You should move toward the active use of your calendar, taking theinitiative to fill theholesbetween the time-criticaleventswithnon-time-criticalthoughnecessaryactivities.

2.Youshouldsay“no”attheoutsettoworkbeyondyourcapacitytohandle.

Itisimportanttosay“no”earlierratherthanlaterbecausewe’velearnedthattowaituntilsomethingreachesahighervaluestageandthenabortduetolackofcapacitymeanslosingmoremoneyandtime.Youcanobviouslysay“no”either

explicitly or implicitly, because by not delivering you end up saying whatamountsto“no.”Remembertoothatyourtimeisyouronefiniteresource,andwhenyousay“yes”toonethingyouareinevitablysaying“no”toanother.The next production principle you can apply is to allow slack—a bit of

looseness in your scheduling. Highway planners, for example, know that afreewaycanhandleanoptimumnumberofvehicles.Havingfewercarsmeansthat the road isnotbeingusedatcapacity.Butat thatoptimumpoint, if justafew more cars are allowed to enter the traffic flow, everything comes to acrunching halt.With the newmetering devices that control access during therush hour, planners can get a fix on the right number.The same thing can bedoneformanagerialwork.Thereisanoptimumdegreeofloading,withenoughslackbuiltinsothatoneunanticipatedphonecallwillnotruinyourschedulefortherestoftheday.Youneedsomeslack.Anotherproductionprinciple is verynearly theopposite.Amanager should

carry a rawmaterial inventory in termsofprojects.This isnot tobe confusedwith his work-in-process inventory, because that, like eggs in a continuousboiler,tendstospoilorbecomeobsoleteovertime.Insteadthisinventoryshouldconsist of things you need to do but don’t need to finish right away—discretionaryprojects,thekindthemanagercanworkontoincreasehisgroup’sproductivity over the long term. Without such an inventory of projects, amanager will most probably use his free timemeddling in his subordinates’work.A final principle. Most production practices follow well-established

procedures and, rather than reinventing the wheel repeatedly, use a specificmethod that has been shown to work before. But managers tend to beinconsistentandbringawelterofapproachestothesametask.Weshouldworkto change that.Aswebecomemore consistent,we should also remember thatthe valueof an administrative procedure is containednot in formal statementsbutintherealthinkingthatledtoitsestablishment.Thismeansthatevenaswetrytostandardizewhatwedo,weshouldcontinuetothinkcriticallyaboutwhatwedoandtheapproachesweuse.

Built-InLeverage:HowManySubordinatesShouldYouHave…

Animportantcomponentofmanagerialleverageisthenumberofsubordinatesamanagerhas.Ifhedoesnothaveenough,his leverageisobviouslyreduced.Ifhe has too many, he gets bogged down—with the same result. As a rule of

thumb, amanagerwhosework is largely supervisory should have six to eightsubordinates;threeorfouraretoofewandtenaretoomany.Thisrangecomesfrom a guideline that amanager should allocate about a half day perweek toeachofhissubordinates.(Twodaysaweekpersubordinatewouldprobablyleadto meddling; an hour a week does not provide enough opportunity formonitoring.)The six to eight rule is right for the classically hierarchicalmanagerwhose

primaryworkisthesupervisionofothers.Whataboutaknow-howmanager,themiddle manager who mainly supplies expertise and information? Even if heworkswithoutasinglesubordinate,servicinganumberofvaried“customers”asaninternalconsultantcaninitselfbeafull-timejob.Infact,anyonewhospendsaboutahalfdayperweekasamemberofaplanning,advisory,orcoordinatinggrouphastheequivalentofasubordinate.Soasaruleofthumb,ifamanagerisbothahierarchicalsupervisorandasupplierofknow-how,heshouldtrytohaveatotalofsixtoeightsubordinatesortheirequivalent.Sometimesabusinessisorganizedinawaythatmakestheidealfan-outofsix

to eight subordinates hard to reach.Amanufacturing plant, for example,mayhave an engineering section and a production section, inwhich case the plantmanagerwould only have two people reporting directly to him. Themanagermight then choose to “act” asoneof the two subordinates, choosing tobehisown engineering manager, for instance. If he does that, the manufacturingmanagerwillstillreport tohim,andhewillhaveaddedthepeoplewhowouldordinarilyreport totheheadofengineering.Sotheplantmanagerwillactuallyhave six direct reports: five engineers and the manufacturing manager. Thearrangement,shownbelow,doesnothave theengineersappearingtobeat thesameorganizational levelas themanufacturingmanager—somethinghewouldsurelytakeexceptionto.

Thisarrangementwillavoidforcingtheplantmanagereitherintoon-the-jobretirementorintomeddling.

Interruptions—ThePlagueofManagerialWork

Thenext importantproductionconceptwecanapply tomanagerialwork is tostrive toward regularity. We could obviously run our breakfast factory moreefficiently if customers arrived in a steady and predictable stream rather thandropping inbyonesand twos.Thoughwecan’tcontrolourcustomers’habits,we should try to smooth out ourworkload asmuch as possible.As noted,weshouldtrytomakeourmanagerialworktakeonthecharacteristicsofafactory,not a job shop.Accordingly,we shoulddo everythingwe can toprevent littlestops and starts in our day as well as interruptions brought on by bigemergencies.Eventhoughsomeofthelatterareunavoidable,weshouldalwaysbelookingforsourcesoffuturehigh-prioritytroublebycuttingwindowsintotheblack box of our organization. Recognizing you’ve got a time bomb on yourhandsmeansyoucanaddressaproblemwhenyouwant to,notafter thebombhasgoneoff.Butbecauseyoumustcoordinateyourworkwiththatofothermanagers,you

can only move toward regularity if others do too. In other words, the sameblocks of timemust be used for like activities. For example, at IntelMondaymornings have been set aside throughout the corporation as the time whenplanninggroupsmeet.SoanybodywhobelongstoonecancountonMondayfor

thatpurposeandbefreeofschedulingconflicts.About twenty middle managers at Intel were once asked to be part of an

experiment.Afterpairingup,theytriedsomerole-playinginwhichonemanagerwas to define the problemmost limiting his output and the otherwas to be aconsultantwhowouldanalyzetheproblemandproposesolutions.The most common problem cited was uncontrolled interruptions, which in

remarkablyuniformfashionaffectedbothsupervisoryandknow-howmanagers.Everyone felt that the interruptions got in the way of his “own” work.Interruptionshadacommonsource,mostfrequentlycomingfromsubordinatesandfrompeopleoutsidethemanagers’immediateorganizationbutwhoseworkthe managers influenced. For those in manufacturing, the interruptions mostoftencame fromproductionoperators, and formarketingpeople, fromoutsidecustomers:inshort,fromtheconsumersofthemiddlemanagers’authorityandinformation.The most frequently proposed solutions were not very practical. The idea

mentionedmostoftenwastocreateblocksoftimeforindividualworkbyhidingphysically. But this is a less than happy answer, because the interruptersobviously have legitimate problems, and if themanager responded by hiding,thesewould pile up. One “solution”was a suggestion that customers not callmarketingmanagersatcertainhours.Nogood.Therearebetterways.Let’sapplyaproductionconcept.Manufacturers turn

out standard products. By analogy, if you can pin down what kind ofinterruptions you’re getting, you can prepare standard responses for those thatpop upmost often. Customers don’t come up with totally new questions andproblems day in and day out, and because the same ones tend to surfacerepeatedly, a manager can reduce time spent handling interruptions usingstandard responses. Having them available also means that a manager candelegatemuchofthejobtolessexperiencedpersonnel.Also, if you use the production principle of batching—that is, handling a

group of similar chores at one time—many interruptions that come fromyoursubordinatescanbeaccumulatedandhandlednot randomly,butatstaffandatone-on-onemeetings,thesubjectofthenextchapter.Ifsuchmeetingsareheldregularly,peoplecan’tprotesttoomuchifthey’reaskedtobatchquestionsandproblemsforscheduledtimes,insteadofinterruptingyouwhenevertheywant.The use of indicators, especially the bank of indicators kept over time, can

alsoreduce the timeamanagerspendsdealingwith interruptions.Howfasthe

can answer a question depends on how fast he can put his finger on theinformationheneedsforaresponse.Bymaintaininganarchiveofinformation,amanagerdoesn’thavetodoadhocresearcheverytimethephonerings.If the peoplewho interrupt you knew howmuch theywere disturbing you,

they would probably police themselves more closely and cut down on thenumber of times they felt they had to talk to you right away. In any case, amanagershouldtrytoforcehisfrequentinterrupterstomakeanactivedecisionaboutwhetheranissuecanwait.So,insteadofgoingintohiding,amanagercanhang a sign on his door that says, “I am doing individual work. Please don’tinterrupt me unless it really can’t wait until 2:00.” Then hold an open officehour,andbecompletelyreceptivetoanybodywhowantstoseeyou.Thekeyisthis: understand that interrupters have legitimate problems that need to behandled.That’swhythey’rebringingthemtoyou.Butyoucanchannelthetimeneeded to deal with them into organized, scheduled form by providing analternativetointerruption—ascheduledmeetingoranofficehour.Thepointistoimposeapatternonthewayamanagercopeswithproblems.

Tomakesomethingregularthatwasonceirregularisafundamentalproductionprinciple,andthat’showyoushouldtry tohandle the interruptionsthatplagueyou.

4Meetings—TheMediumofManagerialWork

Meetingshaveabadname.Oneschoolofmanagementthoughtconsidersthemthe curse of the manager’s existence. Someone who did a study found thatmanagersspendupto50percentoftheirtimeinmeetings,andimpliedthatthiswastimewasted.PeterDruckeroncesaidthatspendingmorethan25percentofhis time inmeetings isasignofamanager’smalorganization,andWilliamH.Whyte, Jr., in his book The Organization Man, described meetings as “non-contributorylabor”thatmanagersmustendure.Butthereisanotherwaytoregardmeetings.Earlierwesaidthatabigpartof

amiddlemanager’sworkistosupplyinformationandknow-how,andtoimparta sense of the preferred method of handling things to the groups under hiscontrolandinfluence.Amanageralsomakesandhelpstomakedecisions.Bothkindsofbasicmanagerial tasks canonlyoccurduring face-to-faceencounters,and therefore only duringmeetings. Thus Iwill assert again that ameeting isnothinglessthanthemediumthroughwhichmanagerialworkisperformed.Thatmeansweshouldnotbefightingtheirveryexistence,butratherusingthetimespentinthemasefficientlyaspossible.The twobasicmanagerial rolesproduce twobasickindsofmeetings. In the

first kind ofmeeting, called a process-oriented meeting, knowledge is sharedand information is exchanged. Such meetings take place on a regularlyscheduledbasis.Thepurposeofthesecondkindofmeetingistosolveaspecificproblem. Meetings of this sort, calledmission-oriented, frequently produce adecision.Theyareadhocaffairs,notscheduled long inadvance,because theyusuallycan’tbe.

Process-OrientedMeetings

To make the most of this kind of meeting, we should aim to infuse it withregularity.Inotherwords,thepeopleattendingshouldknowhowthemeetingisrun, what kinds of substantive matters are discussed, and what is to beaccomplished.Itshouldbedesignedtoallowamanagerto“batch”transactions,

tousethesame“production”set-uptimeandefforttotakecareofmanysimilarmanagerial tasks.Moreover, given the regularity, you and theothers attendingcanbegintoforecastthetimerequiredforthekindsofworktobedone.Hence,a“productioncontrol”system,as recordedonvariouscalendars,can takeshape,which means that a scheduled meeting will have minimum impact on otherthingspeoplearedoing.AtIntelweusethreekindsofprocess-orientedmeetings:theone-on-one,the

staffmeeting,andtheoperationreview.

ONE-ON-ONES

AtIntel,aone-on-oneisameetingbetweenasupervisorandasubordinate,anditistheprincipalwaytheirbusinessrelationshipismaintained.Itsmainpurposeis mutual teaching and exchange of information. By talking about specificproblems and situations, the supervisor teaches the subordinate his skills andknow-how, and suggests ways to approach things. At the same time, thesubordinateprovides thesupervisorwithdetailed informationaboutwhathe isdoingandwhatheisconcernedabout.FromwhatIcantell,regularlyscheduledone-on-ones are highly unusual outside of Intel.When I ask a manager fromanother company about the practice, I usually get an “Oh no, I don’t needscheduledmeetingswithmysupervisor[orsubordinate];Iseehimseveraltimesaday…”But there isanenormousdifferencebetweenacasualencounterbyasupervisorandasubordinate,orevenameeting(mission-oriented)toresolveaspecificproblem,andaone-on-one.WhenIntelwasayoungcompany,IrealizedthateventhoughIwasexpected

to supervise both engineering andmanufacturing, I knew very little about thecompany’s first product line,memorydevices. I alsodidn’t knowmuch aboutmanufacturing techniques, my background having been entirely insemiconductordeviceresearch.Sotwoofmyassociates,bothofwhomreportedtome,agreedtogivemeprivatelessonsonmemorydesignandmanufacturing.Thesetookplacebyappointment,andinvolvedateacher/subordinatepreparingfor each; during the session the pupil/supervisor busily took notes, trying tolearn.AsIntelgrew,theinitialtoneandspiritofsuchone-on-onesenduredandgrew.Who should have a one-on-one? In some situations a supervisor should

perhaps meet with all those who work under him, from professionals toproduction operators. But here I want to talk about one-on-ones between a

supervisorandeachoftheprofessionalswhoreporttohimdirectly.Howoften should you have one-on-ones?Or put anotherway, howdo you

decide how often somebody needs such ameeting?The answer is the job- ortask-relevantmaturityofeachofyoursubordinates.Inotherwords,howmuchexperiencedoesagivensubordinatehavewiththespecifictaskathand?Thisisnotthesameastheexperiencehehasingeneralorhowoldheis.Aswewillseelater,themosteffectivemanagementstyleinaspecificinstancevariesfromveryclose to very loose supervision as a subordinate’s task maturity increases.Accordingly, you should have one-on-ones frequently (for example, once aweek)with a subordinatewho is inexperienced in a specific situation and lessfrequently(perhapsonceeveryfewweeks)withanexperiencedveteran.Another consideration here is how quickly things change in a job area. In

marketing,forexample,thepacemaybesorapidthatasupervisorneedstohavefrequent one-on-ones to keep current on what’s happening. But in a researchenvironment,lifemaybequieter,andforagivenleveloftask-relevantmaturity,lessfrequentmeetingsmaysuffice.Howlongshouldaone-on-onemeetinglast?Therereallyisnoanswertothis,

but the subordinatemust feel that there is enough time to broach andget intothornyissues.Lookatitthisway.Ifyouhadabigproblemthatyouwantedtokickaroundwithyoursupervisor—thepersonwhoseprofessionalinterestinthematter is second only to yours—would you want to bring it up in a meetingscheduledtolastonlyfifteenminutes?Youwouldnot.Ifeelthataone-on-oneshould last an hour at aminimum.Anything less, inmy experience, tends tomake the subordinate confine himself to simple things that can be handledquickly.Where should a one-on-one take place? In the supervisor’s office, in the

subordinate’soffice,orsomewhereelse?Ithinkyoushouldhavethemeetinginor near the subordinate’s work area if possible. A supervisor can learn a lotsimply by going to his subordinate’s office. Is he organized or not? Does herepeatedly have to spend time looking for a document hewants?Does he getinterrupted all the time? Never? And in general, how does the subordinateapproachhiswork?Akeypoint about a one-on-one: It shouldbe regarded as the subordinate’s

meeting, with its agenda and tone set by him. There’s good reason for this.Somebody needs to prepare for the meeting. The supervisor with eightsubordinateswouldhave toprepareeight times; the subordinateonlyonce.So

thelattershouldbeaskedtoprepareanoutline,whichisveryimportantbecauseitforceshimtothinkthroughinadvancealloftheissuesandpointsheplanstoraise.Moreover,withanoutline,thesupervisorknowsattheoutsetwhatistobecoveredandcan thereforehelp to set thepaceof themeetingaccording to the“meatiness”of the itemson theagenda.Anoutlinealsoprovidesa frameworkfor supporting information, which the subordinate should prepare in advance.Thesubordinateshouldthenwalkthesupervisorthroughallthematerial.What should be covered in a one-on-one? We can start with performance

figures, indicators used by the subordinate, such as incoming order rates,productionoutput,orprojectstatus.Emphasisshouldbeonindicatorsthatsignaltrouble. Themeeting should also cover anything important that has happenedsince the last meeting: current hiring problems, people problems in general,organizationalproblemsandfutureplans,and—very,veryimportant—potentialproblems.Evenwhenaproblemisn’ttangible,evenifit’sonlyanintuitionthatsomething’swrong,asubordinateowesittohissupervisortotellhim,becauseittriggers a look into theorganizationalblackbox.Themost important criteriongoverningmatters tobe talkedabout is that theybe issues thatpreoccupyandnagthesubordinate.Theseareoftenobscureandtaketimetosurface,consider,andresolve.What is the role of the supervisor in a one-on-one?He should facilitate the

subordinate’s expression of what’s going on and what’s bothering him. Thesupervisoristheretolearnandtocoach.PeterDruckersumsupthesupervisor’sjobhereverynicely:“Thegoodtimeusersamongmanagersdonottalktotheirsubordinatesabouttheirproblemsbuttheyknowhowtomakethesubordinatestalkabouttheirs.”How is this done?By applyingGrove’s Principle ofDidacticManagement,

“Askonemorequestion!”Whenthesupervisorthinksthesubordinatehassaidallhewantstoaboutasubject,heshouldaskanotherquestion.Heshouldtrytokeep the flow of thoughts coming by prompting the subordinate with queriesuntilbothfeelsatisfiedthattheyhavegottentothebottomofaproblem.I’dliketosuggestsomemechanicalhintsforeffectiveone-on-onemeetings.

First,boththesupervisorandsubordinateshouldhaveacopyoftheoutlineandbothshouldtakenotesonit,whichservesanumberofpurposes.Itakenotesinjustaboutallcircumstances,andmostoftenendupneverlookingatthemagain.IdoittokeepmymindfromdriftingandalsotohelpmedigesttheinformationIhearandsee.SinceItakenotesinoutlineform,Iamforcedtocategorizethe

information logically,which helpsme to absorb it. Equally important iswhat“writing it down” symbolizes. Many issues in a one-on-one lead to actionrequired on the part of the subordinate. When he takes a note immediatelyfollowing the supervisor’s suggestion, the act implies a commitment, like ahandshake,thatsomethingwillbedone.Thesupervisor,alsohavingtakennotes,canthenfollowupatthenextone-on-one.A real time-saver is using a “hold” file where both the supervisor and

subordinateaccumulateimportantbutnotaltogetherurgentissuesfordiscussionatthenextmeeting.Thiskindoffileappliestheproductionprincipleofbatchingandsaves timeforboth involvedbyminimizing theneedforadhoccontact—likephonecalls,drop-invisits,andsoon—whichconstitutetheinterruptionsweconsideredearlier.The supervisor should also encourage thediscussionof heart-to-heart issues

duringone-on-ones, because this is theperfect forum for getting at subtle anddeep work-related problems affecting his subordinate. Is he satisfied with hisownperformance?Doessomefrustrationorobstaclegnawathim?Doeshehavedoubts about where he is going? But the supervisor should be wary of the“zinger,”which isaheart-to-heart issuebroughtupatanawkward time.Moreoftenthannot,thesecomeneartheendofameeting.Ifyouletthathappen,thesubordinatemight tell you something like he’s unhappy and has been lookingoutsideforajobandgiveyouonlyfiveminutestodealwithit.Long-distance telephone one-on-ones have becomenecessary becausemany

organizations are now spread out geographically. But these can work wellenough with proper preparation and attention: the supervisor must have theoutline before the meeting begins, both parties should take notes, and so on.Because you can’t see the other participant in the meeting, note-taking can’twork in thesamewayas inaface-to-facemeeting.Exchangingnotesafter themeetingisawaytomakesureeachknowswhattheothercommittedhimselftodo.One-on-onesshouldbescheduledonarollingbasis—settingupthenextone

asthemeetingtakingplaceends.Othercommitmentscantherebybetakenintoaccount and cancellations avoided. If the supervisor uses a set schedule for aone-on-one,suchaseverysecondWednesdaymorning,andifthesubordinate’svacation happens to fall on that date, the meeting is not going to occur. Byschedulingonarollingbasis,thiscanbeeasilyavoided.Whatistheleverageoftheone-on-one?Let’ssayyouhaveaone-on-onewith

your subordinate every two weeks, and it lasts one and a half hours. Ninetyminutesofyourtimecanenhancethequalityofyoursubordinate’sworkfortwoweeks,or for someeighty-plushours, andalsoupgradeyourunderstandingofwhat he’s doing. Clearly, one-on-ones can exert enormous leverage. Thishappensthroughthedevelopmentofacommonbaseofinformationandsimilarwaysofdoingandhandlingthingsbetweenthesupervisorandthesubordinate.And this, asnoted, is theonlyway inwhichefficient andeffectivedelegationcantakeplace.Atthesametime,thesubordinateteachesthesupervisor,andwhatislearned

is absolutely essential if the supervisor is to make good decisions. During arecentone-on-onemeeting,mysubordinate,whoisresponsibleforIntel’ssalesorganization,reviewedtrendindicatorsofincomingorders.WhileIwasvaguelyfamiliarwith them,he laidouta lotof specific informationandconvincedmethat our business had stopped growing. Even though the summer is typicallyslow,heprovedtomethatwhatwasgoingonwasnot justseasonal.Afterwepondered the data for a while and considered their relationship to otherindicators of business activity in our industry, we came to the reluctantconclusionthatbusinesswasinfactslowingdown.Thismeantweshouldtakeaconservativeapproachtonear-terminvestment—nosmallmatter.By sharing his base of information with me, the two of us developed a

congruent attitude, approach, and conclusion: conservatism in our expansionplans.Heleftthemeetinghavingdecidedtoscalebackgrowthinhisownareaofresponsibility.Ilefthavingdecidedtosharewhatwehadconcludedwiththebusiness groups I supervised. Thus, this one-on-one produced substantialleverage:theIntelsalesmanageraffectedalltheothermanagerswhoreportedtome.Todigressabit,Ialsothinkthatone-on-onesathomecanhelpfamilylife.As

thefatheroftwoteenagedaughters,Ihavefoundthattheconversationinsuchatimetogetherisverydifferentintoneandkindfromwhatwesaytoeachotherinothercircumstances.Theone-on-onemakeseachofus taketheotherseriouslyandallowssubtleandcomplicatedmatterstocomeupfordiscussion.Obviously,no notes are taken, as father and daughter usually go out for dinner at arestaurant,butafamilyone-on-oneverymuchresemblesabusinessone-on-one.Istronglyrecommendbothpractices.

STAFFMEETINGS

A staff meeting is one in which a supervisor and all of his subordinatesparticipate, andwhich therefore presents an opportunity for interaction amongpeers.Aswewill see later, peer interaction—especially decision-making by agroupofpeers—isnoteasy.Yetitiskeytogoodmanagement.Theapproachtodecision-makingthatweadvocateinthenextchapter,aswellastheworkingsoftheprincipleofdualreporting(Chapter9),dependonagroupofpeersworkingwelltogether.Bylearninghowthishappensinstaffmeetings,whereagroupofpeersget toknoweachother,andwhere thepresenceofacommonsupervisorhelpspeer interaction todevelop,managerswillbeprepared tobemembersofotherworkingbodiesbasedonpeergroups.Staffmeetings also create opportunities for the supervisor to learn from the

exchangeandconfrontation thatoftendevelops. Inmyowncase, IgetamuchbetterunderstandingofanissuewithwhichIamnotfamiliarbylisteningtotwopeoplewithopposingviewsdiscussitthanIdobylisteningtoonesideonly.My first experiencewith staffmeetingsdates back tomy early professional

yearswhen Iwas theheadofa smallgroupofengineersdoingsemiconductordevice research. Everyone in this group worked on an isolated aspect of aproblem or on a different problem altogether. I was supposed to be thesupervisor,butIfoundthatothersinthegroupwereoftenmorefamiliarwiththeworkofanotherresearcherthanIwas.Thus,agroupdiscussiononanysubjecttendedtogetmoredetailedandmoreheated,butalwaysmorerewarding, thananexchangebetweenmeandoneotherspecialist.Whatshouldbediscussedatastaffmeeting?Anythingthataffectsmorethan

two of the people present. If the meeting degenerates into a conversationbetween twopeopleworkingonaproblemaffectingonly them, thesupervisorshouldbreakitoffandmoveontosomethingelsethatwillincludemoreofthestaff,whilesuggestingthatthetwocontinuetheirexchangelater.Howstructuredshouldthemeetingbe?Afree-for-allbrainstormingsessionor

controlled with a detailed agenda? It should be mostly controlled, with anagenda issued far enough in advance that the subordinates will have had thechance toprepare their thoughts for themeeting.But it shouldalso includean“opensession”—adesignatedperiodof time for the staff tobringupanythingtheywant.Thisiswhenavariedsetofhousekeepingmatterscanbedisposedof,as well as when important issues can be given a tentative first look. If it isjustified,youcanprovidetimeforamoreformaldiscussionaboutanissueinthescheduledportionofafuturemeeting.

What is the role of the supervisor in the staffmeeting—a leader, observer,expediter, questioner, decision-maker? The answer, of course, is all of them.Please note that lecturer is not listed. A supervisor should never use staffmeetingstopontificate,whichisthesurestwaytounderminefreediscussionandhencethemeeting’sbasicpurpose.Thefigureoppositeshowsthatthesupervisor’smostimportantrolesarebeing

a meeting’s moderator and facilitator, and controller of its pace and thrust.Ideally,thesupervisorshouldkeepthingsontrack,withthesubordinatesbearingthebruntofworkingtheissues.Staffmeetingsareanidealmediumfordecision-making, because the group ofmanagers present has typicallyworked togetherforalongtime.Theformalaswellasinformalauthorityofeachindividualhasbeenwellestablished,andeverybodyknowswholikestospoutoff,whotendstodaydream,whoknowswhatstuffandsoon.Astaffmeetingislikethedinner-table conversation of a family, while other forums of interaction at work,involving people who don’t know each other very well, are like a group ofstrangershavingtomakeadecisiontogether.

Thesupervisor’seffortatastaffmeetingshouldgointokeepingthediscussionontrack,withthesubordinatesbearingthebruntofworkingtheissues.

OPERATIONREVIEWS

This is themedium of interaction for peoplewho don’t otherwise havemuchopportunity to deal with one another. The format here should include formalpresentationsinwhichmanagersdescribetheirworktoothermanagerswhoarenottheirimmediatesupervisors,andtopeersinotherpartsofthecompany.ThebasicpurposeofanoperationreviewatIntelistokeeptheteachingandlearninggoing on between employees several organizational levels apart—people whodon’thaveone-on-onesorstaffmeetingswitheachother.Thisisimportantforboth the junior and senior manager. The junior person will benefit from thecomments, criticisms, and suggestionsof the seniormanager,who in turnwillget a different feel for problems from people familiarwith their details. Suchmeetings are also a source of motivation: managers making the presentationswill want to leave a good impression on their supervisor’s supervisor and ontheiroutsidepeers.Who are the players at an operation review? The organizing manager, the

reviewingmanager,thepresenters,andtheaudience.Eachoftheseplayershasadistinctroletoplayifthereviewistobeausefulone.The supervisor of the presenting managers—an Intel divisional marketing

manager,let’ssay—shouldorganizethemeeting.Heshouldhelpthepresentersdecidewhatissuesshouldbetalkedaboutandwhatshouldnot,whatshouldbeemphasized,andwhatlevelofdetailtogointo.Thesupervisorshouldalsobeinchargeofhousekeeping(themeetingroom,visualmaterials,invitations,andsoon). Finally, he should be the timekeeper, scheduling the presentations andkeeping them moving along. Though it’s hard to judge in advance the timeneededforanydiscussion, thesupervisorhaspresumablyhadmoreexperiencerunning meetings. In any case, he should pace the presenters usinginconspicuous gestures, so that the manager talking doesn’t suddenly findhimselfoutoftimewithonlyhalfhispointscovered.Thereviewingmanageristheseniorsupervisoratwhomthereviewisaimed

—like the general manager of an Intel division. He has a very importantalthoughmoresubtleroletoplay:heshouldaskquestions,makecomments,andingeneralimparttheappropriatespirittothemeeting.Heisthecatalystneededtoprovokeaudienceparticipation,andbyhisexampleheshouldencouragefreeexpression.Heshouldneverpreviewthematerial,sincethatwillkeephimfromreacting spontaneously. Because the senior supervisor is a role model for thejunior managers present, he should take his role at the review extremely

seriously.The people presenting the reviews—a group of marketing supervisors, for

example—shouldusevisual aids suchasoverhead transparencies to theextentpossible.Peopleareendowedwitheyesaswellasears,andthesimultaneoususeofbothdefinitelyhelpstheaudienceunderstandthepointsbeingmade.Butcaremust be taken, because all too frequently a presenter gets so obsessed withgettingthroughallofhisvisualmaterialthathismessagegetslostevenwhileallhischartsgetflipped.Asaruleofthumb,Iwouldrecommendfourminutesofpresentation and discussion time per visual aid, which can include tables,numbers, or graphics. The presenter must highlight whatever he wants toemphasizewithacolorpenorpointer.Throughout,apresenterhastowatchhisaudience like a hawk. Facial expressions and body language, among otherthings,willtellhimifpeoplearegettingthemessage,ifheneedstostopandgooversomethingagain,orifheisboringthemandshouldspeedup.Theaudienceatanoperationreviewalsohasacrucialparttoplay.Oneofthe

distinguishing marks of a good meeting is that the audience participates byaskingquestionsandmakingcomments.Ifyouavoidthepresenter’seyes,yawn,or read the newspaper it’s worse than not being there at all. Lack of interestunderminestheconfidenceofthepresenter.Rememberthatyouarespendingabig part of your working day at the review. Make that time as valuable foryourself and your organization as you can. Pay attention and jot down thingsyou’veheardthatyoumighttry.Askquestionsifsomethingisnotcleartoyouandspeakupifyoucan’tgoalongwithanapproachbeingrecommended.Andifa presenter makes a factual error, it is your responsibility to go on record.Remember,youarebeingpaidtoattendthemeeting,whichisnotmeanttobeasiesta in themidstofanotherwisebusyday.Regardattendanceat themeetingforwhatitis:work.

Mission-OrientedMeetings

Unlikeaprocess-orientedmeeting,whichisaregularlyscheduledaffairheldtoexchange knowledge and information, themission-orientedmeeting is usuallyheldadhocandisdesignedtoproduceaspecificoutput,frequentlyadecision.The key to success here is what the chairman does. Very often no one isofficiallygiventhattitle,butbywhatevername,onepersonusuallyhasmoreatstake in the outcome of the meeting than others. In fact, it is usually thechairmanorthedefactochairmanwhocallsthemeeting,andmostofwhathe

contributesshouldoccurbeforeitbegins.Alltoooftenheshowsupasifhewerejust another attendee and hopes that thingswill develop as hewants.When amission-orientedmeetingfailstoaccomplishthepurposeforwhichitwascalled,theblamebelongstothechairman.Thusthechairmanmusthaveaclearunderstandingofthemeeting’sobjective

—whatneedstohappenandwhatdecisionhastobemade.Theabsolutetruthisthat if you don’t knowwhat you want, you won’t get it. So before calling ameeting,askyourself:WhatamItryingtoaccomplish?Thenask,isameetingnecessary?Ordesirable?Orjustifiable?Don’tcallameetingifall theanswersaren’tyes.An estimate of the dollar cost of a manager’s time, including overhead, is

about$100perhour.Soameetinginvolvingtenmanagersfor twohourscoststhe company $2,000. Most expenditures of $2,000 have to be approved inadvance by senior people—like buying a copying machine or making atransatlantictrip—yetamanagercancallameetingandcommit$2,000worthofmanagerial resources at awhim. So even if you’re just an invited participant,youshouldaskyourself if themeeting—andyourattendance—isdesirableandjustified.Tell thechairman—thepersonwhoinvitedyou—ifyoudonotfeel itis.Determine thepurposeof ameetingbefore committingyour timeandyourcompany’s resources. Get it called off early, at a low-value-added stage, if ameetingmakes no sense, and find a less costlyway (a one-on-onemeeting, atelephonecall,anote)topursuethematter.Assuming the meeting does need to be held, the chairman faces a set of

obligations.Thefirstonehastodowithattendance.Asthechairman,youmustidentifywho should attend and then try to get those people to come. It is notenough to ask people and hope for the best; you need to follow up and getcommitments.Ifsomeoneinvitedcan’tmakeithimself,seetoitthathesendsapersonwiththepowertospeakforhim.Keepinmindthatameetingcalledtomakeaspecificdecisionishardtokeep

moving if more than six or seven people attend. Eight people should be theabsolutecutoff.Decision-makingisnotaspectatorsport,becauseonlookersgetinthewayofwhatneedstobedone.Thechairmanisalsoresponsibleformaintainingdiscipline.Itiscriminalfor

him toallowpeople tobe lateandwasteeveryone’s time.Remember,wastingtime here really means that you are wasting the company’s money, with themetertickingawayattherateof$100perhourperperson.Donotworryabout

confrontingthelatearriver.Justasyouwouldnotpermitafellowemployeetostealapieceofofficeequipmentworth$2,000,youshouldn’t letanyonewalkawaywiththetimeofhisfellowmanagers.Thechairmanshouldfinallyberesponsibleforlogisticalmatters.Heshould,

forexample,makesurethatallnecessaryandaudiovisualequipmentispresentin themeetingroom.Heshouldalsosendoutanagenda thatclearlystates thepurposeofthemeeting,aswellaswhatroleeverybodythereisexpectedtoplaytogetthedesiredoutput.Anexampleofsuchanagendaisshownbelow.

To: FarEastPlantManager ManufacturingManager CorporateConstructionManager PresidentFrom: FarEastConstructionManagerSubject: PhilippinesPlantLocationDecisionMeeting

Friday,October111:00a.m.–1:00p.m.

SantaClaraConferenceRoom212TeleconferenceconnectiontoPhoenixConferenceRoom4

Purposeofmeeting:TodecidespecificlocationforPhilippineplantexpansion

Agenda

11:00–11:30 Manufacturingconsiderations (F.E.PlantManager)11:30–12:00 Constructionconsiderations (F.E.ConstructionManager)12:00–12:45 Reviewofalternatives,includingpreferred

choice(F.E.ConstructionManager)

12:45–1:00 Discussion (All)

Thismaysoundliketoomuchregimentationforyou,butwhetherit’sthatorneededdisciplinedependsonyourpointofview.Ifthechairmanforcesyoutoshow up at a meeting prepared and on time, you might consider him a drillsergeant.Butifyoushowupontime,readytowork,andsomeoneelsedoesn’tandisn’t,you’llprobablybegrudgethepersonresponsibleforwastingyourtime.Itmustbemuchthesameinanoperatingroom.Somepeopleworkingtheremaynot like a surgeon insisting upon precision, but I am one patient who wouldmuchpreferadisciplinedoperatingroomtoanyotherkind.Once the meeting is over, the chairman must nail down exactly what

happenedbysendingoutminutes that summarize thediscussion thatoccurred,

the decision made, and the actions to be taken. And it’s very important thatattendees get the minutes quickly, before they forget what happened. Theminutes should also be as clear and as specific as possible, telling the readerwhatistobedone,whoistodoit,andwhen.Allthismayseemliketoomuchtrouble,butifthemeetingwasworthcallinginthefirstplace,theworkneededto produce theminutes is a small additional investment (an activitywith highleverage)toensurethatthefullbenefitisobtainedfromwhatwasdone.Ideally, a manager should never have to call an ad hoc, mission-oriented

meeting, because if all runs smoothly, everything is taken care of in regularlyscheduled, process-oriented meetings. In practice, however, if all goes well,routinemeetingswilltakecareofmaybe80percentoftheproblemsandissues;the remaining 20 percent will still have to be dealt with in mission-orientedmeetings. Remember, Peter Drucker said that if people spend more than 25percentoftheirtimeinmeetings,itisasignofmalorganization.Iwouldputitanotherway: the real signofmalorganization iswhenpeoplespendmore than25percentoftheirtimeinadhocmission-orientedmeetings.

5Decisions,Decisions

Makingdecisions—ormoreproperly,participatingintheprocessbywhichtheyaremade—isanimportantandessentialpartofeverymanager’sworkfromoneday to the next. Decisions range from the profound to the trivial, from thecomplex to the very simple: Shouldwe buy a building or shouldwe lease it?Issue debt or equity?Shouldwe hire this person or that one?Shouldwe givesomeone a 7 percent or a 12 percent raise?Canwe deposit a phosphosilicateglasswith 9 percent phosphorus contentwithout jeopardizing its stability in aplasticpackage?CanweappealthiscaseonthebasisofRegulation939oftheInternal Revenue Code? Should we serve free drinks at our departmentalChristmasparty?In traditional industries, where the management chain of command was

precisely defined, a person making a certain kind of decision was a personoccupying a particular position in the organization chart. As the sayingwent,authority (to make decisions) went with responsibility (position in themanagement hierarchy). However, in businesses that mostly deal withinformation and know-how, a manager has to cope with a new phenomenon.Herea rapiddivergencedevelopsbetweenpowerbasedonpositionandpowerbased on knowledge, which occurs because the base of knowledge thatconstitutesthefoundationofthebusinesschangesrapidly.What do I mean? When someone graduates from college with a technical

education,atthattimeandforthenextseveralyears,thatyoungpersonwillbefullyup-to-dateinthetechnologyofthetime.Hence,hepossessesagooddealofknowledge-basedpowerintheorganizationthathiredhim.Ifhedoeswell,hewill be promoted to higher and higher positions, and as the years pass, hisposition power will grow but his intimate familiarity with current technologywill fade. Put another way, even if today’s veteran manager was once anoutstandingengineer,heisnotnowthetechnicalexperthewaswhenhejoinedthe company.At Intel, anyway,wemanagers get a littlemore obsolete everyday.

Soabusinesslikeourshastoemployadecision-makingprocessunlikethoseusedinmoreconventionalindustries.IfIntelusedpeopleholdingold-fashionedposition power tomake all its decisions, decisions would bemade by peopleunfamiliarwiththetechnologyoftheday.Andingeneral,thefasterthechangein the know-how on which the business depends or the faster the change incustomer preferences, the greater the divergence between knowledge andposition power is likely to be. If your business depends on what it knows tosurviveandprosper,whatdecision-makingmechanismshouldyouuse?Thekeyto success isagain themiddlemanager,whonotonly isa link in thechainofcommandbutalsocanseetoitthattheholdersofthetwotypesofpowermeshsmoothly.

IdealModel

Illustrated on this page is an ideal model of decision-making in a know-howbusiness.The first stageshouldbe freediscussion, inwhichallpointsofviewand all aspects of an issue are openlywelcomed and debated.The greater thedisagreementandcontroversy,themoreimportantbecomesthewordfree.Thissounds obvious, but it’s not often the practice. Usually when a meeting getsheated, participants hang back, trying to sense the direction of things, sayingnothing until they see what view is likely to prevail. They then throw theirsupport behind that view to avoid being associated with a losing position.Bizarre as itmay seem, someorganizations actually encourage such behavior.Letmequote fromanewsaccount relating to thewoesof a certainAmericanautomobile company: “In the meeting in which I was informed that I wasreleased,Iwastold,‘Bill,ingeneral,peoplewhodowellinthiscompanywaituntiltheyheartheirsuperiorsexpresstheirviewandthencontributesomethinginsupportofthatview.’ ”Thisisaterriblewaytomanage.Allitproducesisbaddecisions, because if knowledgeable people withhold opinions, whatever isdecided will be based on information and insight less complete than it couldhavebeenotherwise.

Theidealdecision-makingprocess.

The next stage is reaching a clear decision. Again, the greater thedisagreement about the issue, themore important becomes theword clear. Infact, particular pains should be taken to frame the terms of the decisionwithutter clarity.Again, our tendency is to do just the opposite:whenwe know adecisioniscontroversialwewanttoobscurematterstoavoidanargument.Butthe argument is not avoided by our beingmealy-mouthed, merely postponed.Peoplewhodon’tlikeadecisionwillbealotmadderiftheydon’tgetapromptandstraightstoryaboutit.Finally, everyone involvedmustgive thedecision reachedby thegroup full

support.Thisdoesnotnecessarilymeanagreement: so longas theparticipantscommit toback thedecision, that isasatisfactoryoutcome.Manypeoplehavetroublesupportingadecisionwithwhichtheydonotagree,butthattheyneedtodosoissimplyinevitable.Evenwhenweallhavethesamefactsandweallhavethe interests of anorganization inmind,we tend tohavehonest, strongly felt,realdifferencesofopinion.Nomatterhowmuch timewemayspend trying toforge agreement, we just won’t be able to get it on many issues. But anorganizationdoesnotlivebyitsmembersagreeingwithoneanotheratalltimes

abouteverything.Itlivesinsteadbypeoplecommittingtosupportthedecisionsandthemovesofthebusiness.Allamanagercanexpectisthatthecommitmenttosupport ishonestlypresent,and this issomethinghecanandmustget fromeveryone.The ideal decision-makingmodel seems an easy one to follow. Yet I have

foundthatitcomeseasilytoonlytwoclassesofprofessionalemployees—seniormanagerswhohavebeeninthecompanyforalongtime,whofeelathomewiththeway thingsaredone,andwho identifywith thevaluesof theorganization;and the new graduates that we hire, because they used themodel as studentsdoingcollegework.Thisisthewayateamofstudentsworkingonalaboratoryexperimentwillresolveitsdifferences,sofortheyoungengineertheIntelmodelisacontinuationofwhathewasusedto.Butformiddlemanagers,thedecision-making model is easier to accept intellectually than it is to practice. Why?Because theyoftenhave trouble expressing theirviews forcefully, ahard timemakingunpleasantordifficultdecisions,andanevenhardertimewiththeideathattheyareexpectedtosupportadecisionwithwhichtheydon’tagree.Itmaytakeawhile,butthelogicoftheidealschemewilleventuallywineveryoneover.Anotherdesirableandimportantfeatureof themodel is thatanydecisionbe

workedoutandreachedatthelowestcompetentlevel.Thereasonisthatthisiswhere itwillbemadebypeoplewhoareclosest to thesituationandknowthemostaboutit.Andby“know”Idon’tjustmean“understandtechnically.”Thatkindofexpertisemustbetemperedwithjudgment,whichisdevelopedthroughexperience and learning from the many errors one has made in one’s career.Thus, ideally, decision-making should occur in the middle ground, betweenreliance on technical knowledge on the one hand, and on the bruises one hasreceivedfromhavingtriedtoimplementandapplysuchknowledgeontheother.Tomakeadecision,ifyoucan’tfindpeoplewithbothqualities,youshouldaimtogetthebestpossiblemixofparticipantsavailable.Forexperience,weatIntelare likely to ask a person inmanagement senior to the othermembers of thegrouptocometothemeeting.Butitisveryimportantthateverybodytherevoiceopinionsandbeliefsasequalsthroughoutthefreediscussionstage,forgettingorignoringstatusdifferentials.A journalist puzzled by ourmanagement style once askedme, “Mr.Grove,

isn’t your company’s emphasis on visible signs of egalitarianism such asinformal dress, partitions instead of offices, and the absence of other obviousperks like reservedparking spaces, just somuchaffectation?”Myanswerwasthat this isnotaffectationbutamatterof survival. Inourbusinesswehave to

mix knowledge-power people with position-power people daily, and togethertheymakedecisionsthatcouldaffectusforyearstocome.Ifwedon’tlinkourengineerswithourmanagers in suchawayas togetgooddecisions,wecan’tsucceedinourindustry.Now,statussymbolsmostcertainlydonotpromotetheflow of ideas, facts, and points of view.What appears to be amatter of stylereallyisamatterofnecessity.

ThePeer-GroupSyndrome

Themodel is also hard to implement because anybodywhomakes a businessdecision also possesses emotions such as pride, ambition, fear, and insecurity.These tend to come to the surface quickly when people who are not used toworkingwithoneanotherareaskedtomakeadecision.Thismeansweneedtothink about what keeps decision-making from happening smoothly along thelineswe’veadvocated.Themostcommonproblemissomethingwecallthepeer-groupsyndrome.A

numberofyearsago,atIntel’sveryfirstmanagementtrainingsession,wetriedsomerole-playingtoshowpeoplewhatcanoccurwhenagroupofpeersmeetstosolveaproblemormakeadecision.Wesatthepeoplearoundatabletotacklewhat was then a live issue for them in their real jobs. Everyone was anorganizationalequal.Thechairmanofthemeetingwasonelevelhigher,butwaspurposely sent out of the room so he couldn’t hear what was to happen.Observers in the audience couldn’t believe their eyes and ears as the mockmeetingproceeded.Themanagersworkingon theproblemdidnothingbutgoaround in circles for some fifteen minutes, and none of them noticed theyweren’tgettinganywhere.Whenthechairmanwasbroughtbackin,hesatdownandlistenedforawhileandcouldn’tbelievethingseither.Wewatchedhimleanforwardasifheweretryingtogleanmorefromtheconversation.Wethensawablack cloud form over his head; finally he slapped the table and exclaimed,“What’sgoingonhere?Youpeoplearetalkingincirclesandgettingnowhere.”Afterthechairmanintervened,theproblemwasresolvedinveryshortorder.Wenamed this the peer-plus-one approach, and have used it since then to aiddecision-makingwherewemust.Peerstendtolookforamoreseniormanager,evenifheisnotthemostcompetentorknowledgeablepersoninvolved,totakeoverandshapeameeting.Why?Becausemost people are afraid to stick their necks out. This is how

John,anIntelsoftwareengineer,seesthings:

Oneofthereasonswhypeoplearereluctanttocomeoutwithanopinioninthepresenceoftheirpeersisthe fear of going against the group by stating an opinion that is different from that of the group.Consequently, thegroup as awholewanders around for awhile, feeling eachother out,waiting for aconsensustodevelopbeforeanyoneriskstakingaposition.Ifandwhenagroupconsensusemerges,oneofthememberswillstateitasagroupopinion(“Ithinkourpositionseemstobe…”),notasapersonalposition. After a weak statement of the group position, if the rest of the mob buys in, the positionbecomesmoresolidandisrestatedmoreforcefully.

Note the difference between the situation described earlier by the autoexecutive and the one John describes. In the former instance, the peoplewereexpected to wait for their supervisor to state his opinion first. In the latter,membersof thegroupwerewaiting foraconsensus todevelop.Thedynamicsaredifferent,butthebottomlineinbothisthatpeopledidn’treallyspeaktheirminds freely. That certainlymakes it harder for a manager tomake the rightdecisions.Youcanovercomethepeer-groupsyndromeifeachofthemembershasself-

confidence, which stems in part from being familiar with the issue underconsiderationandfromexperience.Butintheendself-confidencemostlycomesfrom a gut-level realization that nobody has ever died frommaking a wrongbusiness decision, or taking inappropriate action, or being overruled. Andeveryoneinyouroperationshouldbemadetounderstandthis.If the peer-group syndromemanifests itself, and themeeting has no formal

chairman, the person who has the most at stake should take charge. If thatdoesn’twork,onecanalwaysask theseniorpersonpresent toassumecontrol.Heislikelytobenomoreexpertintheissuesathandthanothermembersofthegroup—perhapslessexpert—butheislikelytoactasagodfather,arepositoryofknowledge about how decisions should be made, and give the group theconfidenceneededtomakeadecision.Onethingthatparalyzesbothknowledgeandpositionpowerpossessorsisthe

fearofsimplysoundingdumb.Fortheseniorperson,thisislikelytokeephimfrom asking the questions he should ask. The same fear will make otherparticipantsmerelythinktheirthoughtsprivatelyratherthanarticulatethemforall tohear;atbest theywillwhisperwhattheyhavetosaytoaneighbor.Asamanager,youshouldremindyourselfthateachtimeaninsightorfactiswithheldand an appropriate question is suppressed, the decision-making process is lessgoodthanitmighthavebeen.Arelatedphenomenon influences lower-levelpeoplepresent in themeeting.

This group has to overcome the fear of being overruled, which might mean

embarrassment:iftherestofthegrouporasenior-levelmanagervetoedajuniorpersonoropposedapositionhewasadvocating,thejuniormanagermightlosefaceinfrontofhispeers.This,evenmorethanfearofsanctionsorevenofthelossofjob,makesjuniorpeoplehangbackandletthemoreseniorpeoplesetthelikelydirectionofdecision-making.But some issues are so complex that those called on to make a decision

honestlyaren’treallysurehowtheyfeel.Whenknowledgeandpositionpowerareseparated,thesenseofuncertaintycanbecomeespeciallyacute,becausetheknowledge people are often not comfortable with the purely business-relatedfactorsthatmightinfluenceadecision.Whatisoftenheardis,“Wedon’tknowwhatthecompany[ordivisionordepartment]wantsofus.”Similarly,managersholdingpositionpowerdon’tknowwhat todobecause they realize theydon’tknow enough about the technical details to arrive at the correct decision.Wemust strive not to be done in by such obstacles. We are all human beingsendowedwithintelligenceandblessedwithwillpower.Bothcanbedrawnupontohelpusovercomeourfearofsoundingdumborofbeingoverruled,andleadustoinitiatediscussionandcomeoutfrontwithastand.

StrivingfortheOutput

Sometimesnoamountofdiscussionwillproduceaconsensus,yetthetimeforadecision has clearly arrived.When this happens, the senior person (or “peer-plus-one”)whountilnowhasguided,coached,andproddedthegroupalonghasno choice but tomake a decision himself. If the decision-making process hasproceeded correctly up to this point, the senior manager will be making thedecisionhavinghadthefullbenefitoffreediscussionwhereinallpointsofview,facts,opinions,and judgmentswereairedwithoutposition-powerprejudice. Inother words, it is legitimate—in fact, sometimes unavoidable—for the seniorperson towield position-power authority if the clear decision stage is reachedandnoconsensushasdeveloped.Itisnotlegitimate—infact,itisdestructive—forhimtowieldthatauthorityanyearlier.Thisisoftennoteasy.WeAmericanstendtobereluctanttoexercisepositionpowerdeliberatelyandexplicitly—itisjust“notnice”togiveorders.Suchreluctanceonthepartoftheseniormanagercan prolong the first phase of the decision-making process—the time of freediscussion—pasttheoptimumpoint,andthedecisionwillbeputoff.Ifyoueitherenter thedecision-makingstage tooearlyorwait too long,you

won’tderivethefullbenefitofopendiscussion.Thecriteriontofollowisthis:

don’tpushforadecisionprematurely.Makesureyouhaveheardandconsideredtherealissuesratherthanthesuperficialcommentsthatoftendominatetheearlypartofameeting.Butifyoufeelthatyouhavealreadyheardeverything,thatallsides of the issue have been raised, it is time to push for a consensus—andfailingthat,tostepinandmakeadecision.Sometimesfreediscussiongoesoninan unending search for consensus.But, if that happens, people can drift awayfrom the near consensus when they are close to being right, diminishing thechancesofreachingthecorrectdecision.Somovingontomakethedecisionattherighttimeiscrucial.Basically, like other things managers do, decision-making has an output

associatedwithit,whichinthiscaseisthedecisionitself.Likeothermanagerialprocesses,decision-makingislikeliertogeneratehigh-qualityoutputinatimelyfashionifwesayclearlyattheoutsetthatweexpectexactlythat.Inotherwords,oneofthemanager’skeytasksistosettlesiximportantquestionsinadvance:

•Whatdecisionneedstobemade?•Whendoesithavetobemade?•Whowilldecide?•Whowillneedtobeconsultedpriortomakingthedecision?•Whowillratifyorvetothedecision?•Whowillneedtobeinformedofthedecision?

LetmeillustratehowthesesixquestionscameintoplayinarecentdecisionIwas involved in. Intel had already decided to expand its Philippinemanufacturing plant, roughly doubling its capacity. The next question waswhere.Only limited spacewas available next to the existing plant. But, otherthings being equal, building therewas themost desirable thing to do becauseoverheadandcommunicationscouldbeshared,transportationcostsbetweenthetwo plants would amount to virtually nothing, and our employees could betransferredfromoneplanttotheotherveryeasily.Thealternativeconsistedofbuyingalessexpensiveplotoflandquitesomedistanceaway.Thelandwouldbe not only cheaper but more plentiful, which would allow us to build arelativelyinexpensiveone-ortwo-storybuilding.Buyingthelotneartheexistingplantmeant thatwewould have had to build a high-rise to get the amount offloor space we needed, and a high-rise semiconductor manufacturing plantwouldnotbethemostefficient.Thatmadeushesitate.Butitwouldbenicetohaveasecondbuildingnexttotheonewealreadyown.Backandforthandsoon

andsoforthwentthediscussion.Let’s apply our six questions here. It is clear what decision needed to be

made: we either build a multistory building next to our existing plant, or webuildaone-ortwo-storybuildingatanewoutlyinglocation.Asforthequestionwhen:accordingtoourlong-rangeplans,weneededthenewplantintwototwoandahalfyears; ifweapply timeoffsets,wemustmakethedecisionwithinamonth.Thisanswersthewhen.Who will decide? Our facilities/construction people or the Intel group that

manages the manufacturing plants? The answer is not easy. The firstorganizationismoresensitivetomatterspertainingtothecostsanddifficultiesof construction, and will probably lean toward the new location. The plantmanagementgroup,knowingthatoperationalbenefitswillcomefromhavingthetwo plants side by side, will probably opt for the high-rise. So the decision-making body is composed of our construction manager for our Far Eastlocations; his supervisor, the construction manager for the corporation; themanager of the Far Eastmanufacturing plant network; and his supervisor, theseniormanufacturingmanager.Themeetinggaveusparallellevelsofmanagersfrom the twoorganizations.The sensitivities of two interest groups coming tobear on a single decision is quite common in real corporate life. In suchmeetings, it is important togive to the twosides roughlyequal representation,because only from such balancewill an even-handed decision emerge. All oftheseindividualshaveconsultedtheirstaffspriortothedecisionandgatheredallrelevantknowledgeandviewsonthesubject.Whowill ratifyorveto thedecision?Thefirstcommonperson towhomthe

senior managers of both organizations report is myself. Also, this was a bigenoughdealthatthepresidentofthecompanyshouldbeinvolved.Moreover,IwassomewhatfamiliarwiththelocationsinthePhilippinesandhowaplantliketheonewehavethereoperates.SoIwaschosenasthepersontovetoorratifythedecisionofthemeeting.Whowill need to be informedof this decision? I choseGordonMoore, our

chairmanoftheboard.He’snotdirectlyinvolvedwithmanufacturingplantsliketheonecontemplated,butwedon’tbuildanewoneintheFarEasteveryday,soheshouldknowaboutwhathappened.This ishowthedecisionwasmade.Afterstudyingmaps,constructionplans

and costs, land costs, and traffic patterns, and considering several timeseverything we thought was important, the group decided to build next to our

existing plant but to accept only as much manufacturing area as four storieswouldyield.Thecostwouldhaveescalatedhadweexceededthat.This,withallrelevant background, was presented to me at the meeting described on theagenda shown in the previous chapter. I listened to the presentation of thealternatives the group considered and to the reasons why they preferred theirchoice to these, and after asking a series of questions and probing both thegroup’s information and its thinking process, I ratified the decision.SubsequentlyIinformedGordonMooreoftheoutcome,andasyouarereadingthis,theplantiseitherunderconstructionoralreadyoperating.Employing consistent ways by which decisions are to be made has value

beyondsimplyexpeditingthedecision-makingitself.Peopleinvestagreatdealofenergyandemotionincomingupwithadecision.Thensomebodywhohasanimportantsay-soortherighttovetoitmaycomeacrossthedecisionlater.Ifhe does veto it, he can be regarded as a Johnny-come-lately who upsets thedecision-making applecart. This, of course, will frustrate and demoralize thepeoplewhomayhavebeenworkingonitforalongtime.Ifthevetocomesasasurprise, however legitimate itmay have been on itsmerits, an impression ofpolitical maneuvering is inevitably created. Politics andmanipulation or eventheirappearanceshouldbeavoidedatallcosts.AndIcanthinkofnobetterwaytomake the decision-making process straightforward than to apply before thefactthestructureimposedbyoursixquestions.One last thing. If the final word has to be dramatically different from the

expectationsofthepeoplewhoparticipatedinthedecision-makingprocess(hadI chosen, for example, to cancel thePhilippine plant project altogether),makeyourannouncementbutdon’tjustwalkawayfromtheissue.Peopleneedtimetoadjust, rationalize, and in general put their heads back together. Adjourn,reconvene themeeting after people have had a chance to recover, and solicittheir views of the decision at that time. This will help everybody accept andlearntolivewiththeunexpected.If good decision-making appears complicated, that’s because it is and has

been for a long time. Let me quote from Alfred Sloan, who spent a lifetimepreoccupiedwithdecision-making:“Groupdecisionsdonotalwayscomeeasily.There is a strong temptation for the leading officers to make decisionsthemselveswithout thesometimesonerousprocessofdiscussion.”Because theprocessisindeedonerous,peoplesometimestrytorunawayfromit.AmiddlemanagerIonceknewcamestraightfromoneofthebetterbusinessschoolsandpossessed what we might call a “John Wayne” mentality. Having become

frustrated with the way Intel made decisions, he quit. He joined a companywhere his employers assured him during the interview that people wereencouraged to make individual decisions which they were then free toimplement. Fourmonths later, he came back to Intel. He explained that if hecouldmakedecisionswithoutconsultinganybody,socouldeverybodyelse.

6Planning:Today’sActionsforTomorrow’sOutput

ThePlanningProcess

Most people think “planning” is one of the loftier responsibilities ofmanagement—we all learned somewhere that “a manager plans, organizes,controls.”Infact,planningisanordinaryeverydayactivity;it’ssomethingallofusdoall the timewithno fanfare, inbothourpersonalandprofessional lives.Forinstance,asyou’redrivingtoworkinthemorning,youarelikelytodecidewhetherornotyouneedgasoline.Youlookat thegaugetoseehowmuchgasyouhave in the tank,youestimatehow far it isyouneed togo, andyou thenmakearoughguessastohowmuchgasyouneedtogettoandfromyouroffice.Bycomparinginyourmindthegasyouneedwiththegasyouhave,youdecidewhetheryoushouldstopforgasornot.Thisisasimpleexampleofplanning.Thedynamicsofplanningcanbestbeunderstoodbygoingbacktoourbasic

productionprinciples.AswelearnedinChapter2,thekeymethodofcontrollingthe future output of a factory is through the use of a system of forecastingdemand and building to forecast.We operated our factory to fill existing andanticipated orders. Our jobwas tomatch the factory’s output at a given timewiththeordersforit.Iftheprojectedoutputdidnotmatchtheprojectedmarketdemand, either we made additional production starts or we reduced them toeliminate theexcess.Howoneplansat the factorycan thenbesummarizedasfollows:step1,determinethemarketdemandforproduct;step2,establishwhatthe factory will produce if no adjustment is made; and step 3, reconcile theprojected factory output with the projected market demand by adjusting theproductionschedule.Yourgeneralplanningprocessshouldconsistofanalogousthinking.Step1is

toestablishprojectedneedordemand:Whatwilltheenvironmentdemandfromyou, your business, or your organization? Step 2 is to establish your presentstatus: What are you producing now? What will you be producing as yourprojects in the pipeline are completed? Put another way, where will yourbusinessbeifyoudonothingdifferentfromwhatyouarenowdoing?Step3is

to compare and reconcile steps 1 and 2.Namely,whatmore (or less) do youneedtodotoproducewhatyourenvironmentwilldemand?Let’sconsidereachstepinmoredetail.

STEP1—ENVIRONMENTALDEMAND

Just what is your environment? If you look at your own group within anorganizationasifitwereastand-alonecompany,youseethatyourenvironmentis made up of other such groups that directly influence what you do. Forexample, if you were the manager of the company’s mailroom, yourenvironmentwouldconsistofcustomerswhoneedyourservices(therestofthecompany), vendors who are able to provide you with certain capabilities(postagemeters,mailcarts),andfinally,yourcompetitors.Youdon’t,ofcourse,have competitors internally—but you can compare your service to one likeUnitedParcelasawaytojudgeperformanceandsetstandards.Whatshouldyoulookforwhenyouexamineyourenvironment?Youshould

attempttodetermineyourcustomers’expectationsandtheirperceptionofyourperformance. You should keep abreast of technological developments likeelectronic mail and other alternative ways of doing your job. You shouldevaluate the performance of your vendors. You should also evaluate theperformance of other groups in the organization to which you belong. Doessomeothergroup(likethetrafficdepartment)affecthowwellyoucandoyourwork?Canthatgroupmeetyourneeds?Once you have establishedwhat constitutes your environment, you need to

examineitintwotimeframes—now,andsometimeinthefuture,let’ssayinayear.Thequestions thenbecome:Whatdomycustomerswant frommenow?AmIsatisfyingthem?Whatwilltheyexpectfrommeoneyearfromnow?Youneedtofocusonthedifferencebetweenwhatyourenvironmentdemandsfromyounowandwhatyouexpectit todemandfromyouayearfromnow.Suchadifferenceanalysisiscrucial,becauseifyourcurrentactivitiessatisfythecurrentdemandsplacedonyourbusiness,anythingmoreandnewshouldbeundertakento match this difference. How you react to this difference is in fact the keyoutcomeoftheplanningprocess.Shouldyouatthisstageconsiderwhatpracticalstepsyoucanactuallytaketo

handlematters?No, thatwill just confuse the issue.Whatwould happen to afactory,forinstance,ifthemarketingorganizationadjusteditsdemandforecastonthebasisofitsownassessmentofthemanufacturingunit’sabilitytodeliver?

If marketing knew they could sell 100 widgets per month but thought thatmanufacturingcouldonlydeliverten,andsosubmittedademandforecastoftenunits,manufacturingwouldnevertooluptosatisfytherealdemand.

STEP2—PRESENTSTATUS

Thesecondstepofplanningistodetermineyourpresentstatus.Youdothisbylistingyourpresentcapabilitiesandtheprojectsyouhaveintheworks.Asyouaccount for them, be sure to use the same terms, or “currency,” inwhich youstated demand. For instance, if your demand is listed in terms of completedproduct designs, your work-in-process should be listed as partially completedproduct designs. You also need to look at timing; namely, when will theseprojectscomeoutofyour“pipeline”?Youmustaskyourself,willeveryprojectnowmovingthroughbecompleted?Chancesare,no,somewillgetscrappedoraborted,andyouhavetofactorthisintoyourforecastedoutput.Statistically,insemiconductor manufacturing, only some 80 percent of the material startedactually gets finished. Similarly, while it is impossible to be precise in everycase,itisprudenttofactorinsomepercentageoflossformanagerialprojectsaswell.

STEP3—WHATTODOTOCLOSETHEGAP

The final stepofplanning consists of undertakingnew tasksormodifyingoldones to close the gap between your environmental demand and what yourpresent activitieswill yield. The first question is,What do you need to do toclosethegap?Thesecondis,Whatcanyoudotoclosethegap?Considereachquestionseparately,andthendecidewhatyouactuallywilldo,evaluatingwhateffectyouractionswillhaveonnarrowingthegap,andwhen.Thesetofactionsyoudecideuponisyourstrategy.Muchconfusionexistsbetweenwhatisstrategyandwhatistactics.Although

thedistinctionisrarelyofpracticalsignificance,here’sonethatmightbeuseful.Asyouformulate inwordswhatyouplan todo, themostabstractandgeneralsummaryofthoseactionsmeaningfultoyouisyourstrategy.Whatyou’lldotoimplementthestrategyisyourtactics.Frequently,astrategyatonemanageriallevel is the tactical concern of the next higher level. Let’s return to ourmailroom.Assumethatthemanagerofcorporatecommunicationshasdecidedtoinstall electronic mail service between all manufacturing plants. This is hisstrategy—aplanofactiontoimprovecommunicationscapabilitybetweenplants.

Themailroommanager then has to do certain things to provide servicewhenelectronicmail equipment isput inplace.For instance,his strategymaybe toinstall printers in themailroom and set up a service to deliver printed copiesthroughout the building. The mailroom manager’s strategy is thecommunicationsmanager’stactics.

SOMEEXAMPLES

As he defined his present environment and status, Bruce, an Intel marketingmanager, found that he had only three people in his department who couldpossiblyhandleahugeinventoryofprojects.Ashelookedathisdesiredfuturestatus,heconcluded thateverysingleoneof theprojectshad tobecompleted.Notfinishingeverythingwouldresultinsignificantextracostandfarmoreeffortlater.Brucewasfacedwithagenuinedilemma,especiallysincethebudgetkepthim from hiring more people. He realized that the best he could do was tonarrowthegapabit—gettingtheprojectsandhisgroup’scapacitytocompletethemtoconverge.Acompletematchwasimpossible.Brucedecidedtomoveasmanynoncriticaltasksaspossibletoothergroups

inthecompany—groupslessqualifiedforthemthanhisownbutalsolessbusy.He also made arrangements with his manager to bring a summer student onboard to help with some easily definable tasks, and then set himself up tomonitortheperformanceofhisgrouptightly.Healsoproceededtolookatotheravenues that could help him in the longer term, such as splitting the job ofcompleting some of the work with other similar marketing groups in thecompanyandeliminatinganyduplicationofeffortbetweenthem.FinallyBruceinitiated a request to increase the size of his organization. His plan—and theobviousrealitythatfullconvergencebetweenhistasksandhiscapabilitieswasnotpossibleevenaftergoingtheextramile—wouldlaythebasisforhisrequest.Let’s illustrate with another example. Our middle manager Cindy, the

manufacturing process engineer whom we’ve met before, is responsible formaintaining and improving the process by which complex microchips areproducedinaplant.Shedefinesherenvironmentasacollectionof“objects”and“influences.”The“objects”arenewprocessesandmanufacturingtoolsthathavenotyetbeentestedinproduction.The“influences”arethepeoplewhocanaffectherworkdirectlyorindirectly.Developmentengineers,forexample,wouldlikeher to require less experimentation and documentation before she chooses toimplement new processes they have developed. Meanwhile, production

engineerswouldlikehertoprovidemoreexperimentationanddocumentationonthesesamenewprocesses.Andfinally,therearetheproductengineerseagertogetchipsoutthedoorwhoneedherhelptodothat,alongwithothermembersoftheproductionteamwhoputpressureonhertoensurethatthenewprocessesaremanufacturableandthenewmanufacturingtoolsworkassoonastheyareputtouse.Cindyherselfworkslikeaconsultant,advisingeachgroupinfluencingheraboutwhethersomethingcangointoproduction—thechiefcoordinatorfor theset of events required to put a product, a process, or a tool on stream. Her“customer” is theproductionarea itself,andher“vendors”are theengineeringgroupsfromproduction,development,andproductengineering.Analyzingherpresentstatus,Cindyfound that thedataandexperimentation

she needed from the development group always came in incomplete. Lookinginto matters further, she found that providing complete data and adhering toscheduleswerenot reallyhighon thedevelopmentengineers’ setofpriorities.Determiningwheresheneededtogo,itwascleartoCindythatshemusthaveallfuturenewprocessesandproductionmachinerytested,debugged,demonstrated,and,mostimportant,accompaniedbythenecessarydataforthemtobeacceptedand used by the production engineers, who had become more demandingbecauseofpastproblems.Cindy then defined her strategy—her plan of action—to get there. She

specifiedexactlywhatstepshadtobecompletedbeforeanynewprocesswastobe implemented or tool deployed. Then she used time offsets (remember thebreakfast factory) todeterminewheneach stepneeded tobedone inorder forher entire plan to be completed on time. Next, she got the manager of thedevelopment engineers to agree to her detailed schedule. She negotiated withhimwhatshehadtodoandwhattheyhadtodo—andbywhatdate—inordertomeetwhat became themutually agreed-upongoals. Finally, to ensure that shestayedontrack,shedecidedtomonitorallofher“vendors”onaweeklybasis.Shewouldalsopublishtheirperformanceagainstthescheduletomotivatethemto meet key dates (an indicator) and to tell her about potential problems (awindowintheblackbox).

TheOutputofthePlanningProcess

ThekeytobothBruce’sandCindy’seffortsisthattheirplanningproducedtasksthathadtobeperformednowinordertoaffectfutureevents.Ihaveseenfartoomanypeoplewhouponrecognizingtoday’sgaptryveryhardtodeterminewhat

decision has to be made to close it. But today’s gap represents a failure ofplanningsometimeinthepast.Byanalogy,forcingourselvestoconcentrateonthe decisions needed to fix today’s problem is like scurrying after our car hasalreadyrunoutofgas.Clearlyweshouldhavefilledupearlier.Toavoidsuchafate,rememberthatasyouplanyoumustanswerthequestion:WhatdoIhavetodotodaytosolve—orbetter,avoid—tomorrow’sproblem?Thus,thetrueoutputoftheplanningprocessisthesetoftasksitcausestobe

implemented.TheoutputofIntel’sannualplan,forinstance,istheactionstakenand changes prompted as a result of the thinking process that took placethroughout the organization. I, for one, hardly ever look at the bound volumefinallycalledtheAnnualPlan.Inotherwords,theoutputoftheplanningprocessisthedecisionsmadeandtheactionstakenasaresultoftheprocess.Howfaraheadshouldtheplannerslook?AtIntel,weputourselvesthroughan

annualstrategiclong-rangeplanningeffortinwhichweexamineourfuturefiveyearsoff.Butwhatisreallybeinginfluencedhere?Itisthenextyear—andonlythenextyear.Wewillhaveanotherchancetoreplanthesecondofthefiveyearsinthenextyear’slong-rangeplanningmeeting,whenthatyearwillbecomethefirstyearofthefive.So,keepinmindthatyouimplementonlythatportionofaplan that lieswithin the timewindowbetween now and the next time you gothroughtheexercise.Everythingelseyoucanlookatagain.Weshouldalsobecarefulnottoplantoofrequently,allowingourselvestimetojudgetheimpactofthedecisionswemadeandtodeterminewhetherourdecisionswereontherighttrackornot.Inotherwords,weneedthefeedbackthatwillbeindispensabletoourplanningthenexttimearound.Whoshouldbeinvolvedintheplanningprocess?Theoperatingmanagement

of the organization.Why?Because the idea that planners can be people apartfrom those implementing the plan simply does not work. Planning cannot bemade a separate career but is instead a key managerial activity, one withenormous leverage through its impact on the future performance of anorganization.But this leverage canonlybe realized throughamarriage, and agoodcollaborativeoneatthat,betweenplanningandimplementation.Finally, remember that by saying “yes”—to projects, a course of action, or

whatever—you are implicitly saying “no” to something else. Each time youmakeacommitment,youforfeityourchancetocommittosomethingelse.This,of course, is an inevitable, inescapable consequence of allocating any finiteresource.Peoplewhoplanhavetohavetheguts,honesty,anddisciplinetodrop

projectsaswellastoinitiatethem,toshaketheirheads“no”aswellastosmile“yes.”

ManagementbyObjectives:ThePlanningProcessAppliedtoDailyWork

The system of management by objectives assumes that because our concernshereareshort-range,weshouldknowquitewellwhatourenvironmentdemandsfromus.Thus,managementbyobjectives—MBO—concentratesonsteps2and3of theplanningprocess and tries veryhard tomake them specific.The ideabehindMBOisextremelysimple: Ifyoudon’tknowwhereyou’regoing,youwillnotgetthere.Or,asanoldIndiansayingputsit,“Ifyoudon’tknowwhereyou’regoing,anyroadwillgetyouthere.”AsuccessfulMBOsystemneedsonlytoanswertwoquestions:

1.WheredoIwanttogo?(Theanswerprovidestheobjective.)2.HowwillIpacemyselftoseeifIamgettingthere?(Theanswergivesus

milestones,orkeyresults.)

Toillustrateanobjectiveandakeyresult,considerthefollowing:Iwanttogototheairporttocatchaplaneinanhour.Thatismyobjective.IknowthatImustdrive through towns A, B, and C on my way there. My key results becomereaching A, B, and C in 10, 20, and 30 minutes respectively. If I have beendrivingfor20minutesandhaven’tyetmadetownA,IknowI’mlost.UnlessIgetoffthehighwayandasksomeonefordirections,Iprobablywon’tmakemyflight.Upon what time period should an MBO system focus? MBO is largely

designedtoprovidefeedbackrelevanttothespecifictaskathand;itshouldtellushowwearedoingsowecanmakeadjustments inwhateverwearedoing ifneed be, such as getting off the highway and asking for directions. For thefeedback to be effective, itmust be received very soon after the activity it ismeasuring occurs. Accordingly, an MBO system should set objectives for arelatively short period. For example, if we plan on a yearly basis, thecorrespondingMBOsystem’stimeframeshouldbeatleastasoftenasquarterlyorperhapsevenmonthly.Theone thinganMBOsystemshouldprovidepar excellence is focus.This

canonlyhappen ifwekeep thenumberofobjectivessmall. Inpractice, this israre,andhere,aselsewhere,wefallvictimtoourinabilitytosay“no”—inthis

case,totoomanyobjectives.Wemustrealize—andactontherealization—thatifwe try to focuson everything,we focusonnothing.A fewextremelywell-chosenobjectivesimpartaclearmessageaboutwhatwesay“yes”toandwhatwesay“no”to—whichiswhatwemusthaveifanMBOsystemistowork.

TWOCASEHISTORIES

To familiarize ourselves with the MBO system, let’s look at a case history,Columbus’ discovery of the New World, though how I tell the story takesconsiderable libertieswith thegrammar-schoolversionof theevent.Thanks toitsannualplanningprocessof1491,thegovernmentofSpainconcludedthat itcould not continue a war everybody felt was utterly necessary unless moneybecameavailabletobuyweaponsandammunition.SincepushingtheMoorsoutof Spain was the supreme goal of Queen Isabella’s government, the Queenneededthefundstodoit.IsabelladecidedshewouldgetmoneybydramaticallyimprovingSpain’s foreign trade balance.She then talked to her subordinate—ChristopherColumbus—and toldhimaboutherobjective.Columbusagreed tothinkaboutvariousways todowhatshewantedandaftera timewentbacktoher with several suggestions, which included finding pirate-free passage toEnglandandperhapsfindinganewroute to theOrient. IsabellaandColumbusdiscussed the entirematter freely, eventually reaching a clear decision that hewouldlookforanewroutetotheEast.Oncethedecisionwasmade,Columbusbegantothinkofallthethingsthathe

wouldneedtodotoaccomplishhisintent.InMBOterms,theQueendefinedherownobjective(increaseSpain’swealth);Columbusand theQueen thenagreeduponhisobjective(findanewroute to theOrient).Columbus thenwenton toformulate the key results by which he would pace himself, which includedobtainingseveral ships, trainingcrews,conductinga shakedowncruise, settingsail,andsoforth,witheachpossessingaspecificdeadline.The relationship between Isabella’s and Columbus’ objectives is clear. The

Queenwantedtoincreasehernation’swealth,whileColumbuswantedtofindasafetraderoutetotheOrient.Andweseeanestinghierarchyofobjectives:ifthesubordinate’sobjectivesaremet,thesupervisor’swillbeaswell.Now,thekeyresultscancomeinlikeclockwork,buttheobjectivescanstill

bemissed.ForColumbus,thekeyresultswererelativelyeasytoachieve,buthemost certainlydidnot find anew trade route toChina, and therefore failed tomeethisobjective.

DidColumbusperformwelleventhoughhefailedbystrictMBOterms?Hediddiscover theNewWorld, and thatwas a sourceof incalculablewealth forSpain.So it isentirelypossible fora subordinate toperformwellandbe ratedwelleventhoughhemissedhisspecifiedobjective.TheMBOsystemismeanttopaceaperson—toputastopwatchinhisownhandsohecangaugehisownperformance. It is not a legal document upon which to base a performancereview,butshouldbejustoneinputusedtodeterminehowwellanindividualisdoing.IfthesupervisormechanicallyreliesontheMBOsystemtoevaluatehissubordinate’s performance, or if the subordinate uses it rigidly and forgoestaking advantage of an emerging opportunity because it was not a specifiedobjective or key result, then both are behaving in a petty and unprofessionalfashion.Let’s illustrate the workings of the MBO system using the decision about

Intel’s plant expansion in the Philippines. The Far East constructionmanagerhadanobjectivethatread“ObtaindecisiononPhilippineplantexpansion.”Thekeyresultssupportingtheobjectivewere:1)Doastudyoflandavailabilitynearthe present plant and at other acceptable locations by June. 2) Do financialanalyses showing the trade-offs between land costs and construction costs, aswell as the operating costs associated with the two locations. 3) Present theresultstotheplantlocationsteeringgroup,andobtainadecisionfromthem.4)HaveGroveratifythedecisionbyOctober.Eachkey resultwas accomplished and theobjectivewasmet.Note that the

objectiveisrelativelyshort-rangeandthekeyresultsaresospecificthatapersonknowswithoutquestionwhetherhehascompletedthemanddoneitontimeornot.Accordingly, tobeuseful akey resultmustcontainvery specificwordinganddates,sothatwhendeadlinetimearrives,thereisnoroomforambiguity.Asyoumighthaveguessed, theFarEast constructionmanager’s supervisor

had an objective that read “Ensure that all plant expansion projects stay onschedule.”Tosupport thisobjectivehe in turnhadakey result,much likehissubordinate’sobjective,thatsaid“ObtainPhilippineplantexpansiondecisionbyOctober.”You can now see, I hope, the parallels between how Isabella’s government

and Intelwork.Amanager’sobjectivesare supportedbyanappropriate setofkeyresults.Hisobjectivesinturnaretiedtohissupervisor’sobjectivessothatifthemanagermeets his objectives, his supervisorwillmeet his. But theMBOsystem cannot be run mechanically by a computer. The system requires

judgment and common sense to set the hierarchy of objectives and the keyresults that support them.Both judgment and common sense are also requiredwhenusingMBOtoguideyouinyourworkfromonedaytothenext.

III

TeamofTeams

7TheBreakfastFactoryGoesNational

Weleftthebreakfastfactoryasitwasenjoyinggreatsuccess—sogreat,infact,thatwehadtoinstallacontinuousegg-boilingunitatconsiderableexpense.Theequipment produced breakfasts of unprecedented uniformity. Moreover, ourvolume grew to the point where we could use the egg-boiling unit at fullcapacity;hence,ourcostofdeliveringoutstandingbreakfastssteadilydeclined.Wepassedonsomeofthesavingstoourcustomers,andsoonthereputationofourbreakfastsspread.Like good entrepreneurs, we knew we had a good thing going and started

another branch of theBreakfast Factory across town (we even named it that).Thistoobecamearemarkablesuccess.SoonthereafterNeighborhoodGourmet,amagazinewith a largenational circulation, ran a storyonouroperation.WedecidedtoseizetheopportunityandfranchisetheBreakfastFactorynationwide.Werapidlymoved intoneighborhoodswith the rightdemographicmix forourbreakfasts, and we were soon enough running a vast network of BreakfastFactories.Beforelongwefound,however, that thenetworkrequiredasetof tasksand

skills very different from those needed to run our one restaurant. The mostimportantofthesewastofigureouthowtousetheadvantagesmadepossiblebyhaving a local entrepreneur set up and run each franchise without losing theenormous economies of scale that became available to us. Because the localmanager knows his neighborhood, he can adapt his operation to it and so,wehope,operatethemostprofitablefranchisepossible.Atthesametime,withovera hundred Breakfast Factories, our purchasing power is immense. If wecentralizecertainactivities,weareinapositiontodomanythingsmuchbetterandmuch less expensively than each of our franchises could do individually.Andmost important, because the quality of our breakfasts has played amajorrole in our success so far,we have to be very concerned aboutmaintaining aperceptionoffirst-ratefoodandservice.Inotherwords,wecouldnotpermitanyoneBreakfastFactorybranchorthoseinanyregiontojeopardizetherealsecret

ofourbusiness.In fact, the centralization-decentralization dichotomy is so pervasive that it

has become one of the most important themes in the management of ournetwork. Dowe, for instance, want to advertise locally or nationally? Dowewant togive the localmanager thecontroloveradvertising inhiscommunity?Wedon’tknowwhoreadstheDailyBlatt,andheprobablydoes.Dowewanttogive him the right to hire and fire personnel?Shouldwe let him set hiswagescale,ordowewant to imposeonenationally?The latterhardlymakessense,since labormarketconditionsvaryconsiderably fromregion to region.Butwedowanttobuyoursophisticatedautomaticmachinerycentrally.Afterall,ithastaken us a long time to develop suitable vendors and the capacity to test theincomingmachinesgivenourdemandingrequirements.WenowhaveasizablegroupofpeopledoingonlythatinChicago,andwehardlywanteachbranchoreveneachregiontoduplicatetheeffort.ButIdon’tthinkweshouldbuyalloureggsinChicago.Wewantthemtobe

fresh,andwedon’twanttotruckthisdelicatecommodityalloverthecountry.Neitherdowewanttohaveeachbranchsetupitsownincomingegginspectionoperation. Here some kind of compromisemakes sense, such as regional eggpurchasingcenters,witheachonlya fewhoursby truck fromall the franchiselocationsintheregion.Wedowantuniform,high-qualitystandardseverywhereandwewillmonitoreachofthefranchisestomakesurethattheyareadheredto.Inotherwords,wedefinitelywanttoimposenationalqualitycontrolstandards.Whataboutitemsonthemenu?Byandlargewewanttokeepthesamecore

menu everywhere—people going into theBreakfast Factory should be able tocountonsomebasicchoices.Butweshouldalsorecognizeregionaldifferencesinculinarytaste,sosomediscretionoughttobelefttotheindividualfranchises.Whataboutrealestate?ShouldweallowourBreakfastFactoriestobehoused

inanylocallyavailablebuilding?Orshouldweprescribeauniformconstructionstyle andbuild eachof them from thegroundup?Perhapswe shouldgowithwhatever building is available provided it meets some standards we set inChicago.Whataboutfurniture?Doesithavetobeutterlyuniform?ShouldChicagobuy

furnitureforallbranches?Whatabouttableware?Sincepeopletendtoassociatewhattheyeatwithandfromwiththebreakfast,weshouldprobablyusethesametablewarealloverthecountry,whichmeanswemightaswellpurchaseallthatat one place too. But it is ridiculous for a local operation in Montana to

requisition Chicago for a few broken plates. So we should probably have acoupleofregionalwarehousesfromwhichtablewarecouldbedeliveredquickly.Howdowe choose the locationof new franchiseswithin eachmetropolitan

area?ShouldwemakethedecisionsinChicago?ShouldIdecideastheCEOofthe Breakfast Factory Corporation or should the corporate staff let the localbranchmanagerdecide?OrperhapsChicagoshoulddecideafterconsultingwiththeregionalmanagers,who,afterall,knowtheirownareasbetterthanmystaffandIdo.Thingshavebecomeverycomplicated.SometimesasIsitbehindmybigdesk

atcorporateheadquarters,IwishIcouldgobacktotheearlydayswhenIwasgettingtheeggsandtoastandpouringthecoffeemyself.Orifnotthat,atleastback to the dayswhen I was running a single Breakfast Factory, and I kneweverybodybynameandcouldmakeallthedecisionswithouthavingtostrugglewithamountainofprosandcons.Then therewasvirtuallynooverhead.Nowthere’s a corporate personnel manager. There’s also a traffic manager, whowantstobuyacomputertooptimizetheflowofeggsfromtheregionalcenterstotheindividualfranchises.Hesayshecanminimizetransportationcostswhileensuring same-day delivery. He also claims that if he had this computer, hecouldkeepthetablewareinventoryatthelowestpossiblelevel.Itwon’tbelongbefore we’ll have a corporate manager for real estate acquisitions. Verycomplicatedindeed.Earlier,weestablishedthefactthatthegameofmanagementisateamgame:

amanager’s output is the output of the organizations under his supervision orinfluence.We now discover thatmanagement is not just a team game, it is agameinwhichwehavetofashionateamofteams,wherethevariousindividualteams exist in some suitable and mutually supportive relationship with eachother.

8HybridOrganizations

WhathappenedtotheBreakfastFactoryhastohappen,orhasalreadyhappened,toeveryreasonablylargeorganization.Mostmiddlemanagersrundepartmentsthatareapartofalargerorganization.

The“blackboxes”theyoverseeareconnectedtootherblackboxesinmuchthesamewaythat theBreakfastFactoriesarelinkedtoeachotherandtothemainoffice. So let us look more carefully at what happens within an organizationcomposedofsmallerunits.Though most are mixed, organizations can come in two extreme forms: in

totally mission-oriented form or in totally functional form. The BreakfastFactory Corporation could be organized in one or the other extreme form, asshown on the next page. In the mission-oriented organization (a), which iscompletelydecentralized,eachindividualbusinessunitpursueswhatitdoes—itsmission—with little tie-in to other units. Here, each Breakfast Factory isresponsible for all elements of its operation: determining its own location andconstructing its own building, doing its own merchandising, acquiring andmaintaining its own personnel, and doing its own purchasing. In the end itsubmitsmonthlyfinancialstatementstothecorporateexecutiveoffice.

TheBreakfastFactorynetworkorganizedin(a)totallymission-orientedand(b)totallyfunctionalforms.

At the other extreme is the totally functional organization (b), which iscompletelycentralized.InaBreakfastFactoryCorporationsetupthisway, themerchandisingdepartmentisresponsibleformerchandisingatall locations;thestaff of the personnel organization hires, fires, and evaluates personnel at allbranches;andsoon.The desire to give the individual branch manager the power to respond to

local conditions moves us toward a mission-oriented organization. But asimilarlylegitimatedesire to takeadvantageof theobviouseconomiesofscaleand to increase the leverage of the expertisewehave in eachoperational areaacrosstheentirecorporationwouldpushustowardafunctionalorganization.Intherealworld,ofcourse,welookforacompromisebetweenthetwoextremes.Infact,thesearchfortheappropriatecompromisehaspreoccupiedmanagersfora long, long time.Alfred Sloan summed up decades of experience atGeneralMotorsbysaying,“Goodmanagementrestsonareconciliationofcentralizationand decentralization.” Or, we might say, on a balancing act to get the bestcombinationofresponsivenessandleverage.

Let’snowlookat Intel’sorganizationform,asshownon thenextpage.Weareahybridorganization.Ourhybridnaturecomesfromthefactthattheformoftheoverallcorporateorganization results fromamixof thebusinessdivisions,whicharemission-oriented,andthefunctionalgroups.ThisismuchlikethewayI imagine any army is organized. The business divisions are analogous toindividual fighting units, which are provided with blankets, paychecks, aerialsurveillance, intelligence, and so forth by the functional organizations, whichsupplysuchservicestoallfightingunits.Becauseeachsuchunitdoesnothavetomaintainitsownsupportgroups,itcanconcentrateonaspecificmission,liketakingahillinabattle.Andforthat,eachunithasallthenecessaryfreedomofactionandindependence.Thefunctionalgroupscanbeviewedasif theywereinternalsubcontractors.

Let’s take a sales organization as an example.Though a lot of companies useoutsidesalesrepresentatives,aninternalgrouppresumablyprovidestheserviceat less expense and with greater responsiveness. Likewise, manufacturing,finance,ordataprocessingcanallbe regardedas functionalgroups,which,asinternalsubcontractors,provideservicestoallthebusinessunits.

Intelisahybridorganization:balancingtogetthebestcombinationofresponsivenessandleverage.

SometwothirdsofIntel’semployeesworkinthefunctionalunits,indicatingtheir enormous importance.Whatare someof theadvantagesoforganizing somuchof thecompany in suchgroups?The first is theeconomiesof scale thatcan be achieved. Take the case of computerized information processing.Complex computer equipment is very expensive, and the capacity of largeelectronicmachinescanbebestusedifallthevariousbusinessunitsdrawfromthem. If each unit had its own computer, very expensive equipmentwould besittingidlemuchofthetime.Anotherimportantadvantageisthatresourcescanbeshiftedandreallocatedtorespondtochangesincorporate-widepriorities.Forinstance, becausemanufacturing is organized functionally, we can change themixofproductbeingmadetomatchneedasperceivedbytheentirecorporation.Ifeachbusinessunitdiditsownmanufacturing,shiftingcapacityawayfromoneunittoanotherwouldbeacumbersomeandstickyexercise.Andtheadvantagehere is that the expertise of specialists—know-how managers, such as theresearchengineerswhoworkintechnologydevelopment—canbeappliedacrossthebreadthoftheentirecorporation,givingtheirknowledgeandworkenormousleverage. Finally, Intel’s functional groups allow the business units toconcentrateonmasteringtheirspecifictradesratherthanhavingtoworryaboutcomputers,production,technology,andsoforth.Having so much of Intel organized in functional units also has its

disadvantages. The most important is the information overload hitting afunctionalgroupwhenitmustrespondtothedemandsmadeonitbydiverseandnumerous business units. Even conveying needs and demands often becomesverydifficult—abusinessunithastogothroughanumberofmanagementlayersto influence decision-making in a functional group. Nowhere is this moreevident than in thenegotiations thatgoon tosecureaportionofcentralized—and limited—resourcesof the corporation, be it production capacity, computertime, or space in a shared building. Indeed, things often move beyondnegotiation to intense and open competition among business units for theresourcescontrolledbythefunctionalgroups.Thebottomlinehereisthatboththe negotiation and competition waste time and energy because neithercontributestotheoutputorthegeneralgoodofthecompany.What are some of the advantages of organizing much of a company in a

mission-orientedform?Thereisonlyone.Itisthattheindividualunitscanstay

in touchwith theneedsof their businessorproduct areas and initiate changesrapidlywhen thoseneedschange.That is it.Allotherconsiderations favor thefunctional-typeoforganization.Butthebusinessofanybusinessistorespondtothedemandsandneedsof itsenvironment,andtheneedtoberesponsive issoimportant that it always leads to much of any organization being grouped inmission-orientedunits.Countlessmanagershavetriedtofindthebestmixofthetwoorganizational

forms. And it’s been no different at Intel, among senior management andthroughout the ranks of hundreds ofmiddlemanagers,who from time to timeattempttoimprovetheorganizationofthegroupstheysupervise.Butnomatterhow many times we have examined possible organizational forms, we havealwaysconcludedthatthereissimplynoalternativetothehybridorganizationalstructure.So that is how Intel is organized today. To further my case that hybrid

organizationsareinevitable,considerapressreleasethatIreadrecently.Oneofdozensthatshowupintheweeklytradenewspapers,itisreproducedherewithonlythenameschanged.

ABCTECHNOLOGIESREALIGNS(SANTACLARA,CA)Three-year-oldABCTechnologies,Inc.,hasreorganizedintothreeproductdivisions.TheSuperSystemDivisionVicePresidentandGeneralManagerisJohnDoe,formerlyVicePresidentand Engineering Director and a company founder. Vice President and GeneralManager of the UltraSystem Division is former Sales and Marketing Vice President William Smith. Vice President andGeneralManageroftheHyperSystemDivisionisRobertWorker,formerlyManagerofProductDesign.All threedivisionheads report toABCTechnologiesPresident andChiefExecutiveSamuelSimon.

The divisions will have product marketing and product development responsibilities, while sales andmanufacturingresponsibilitieswillremainatthecorporatelevelundernewlynamedSalesVicePresidentAlbertAbelandManufacturingVicePresidentWilliamWeary.

Note how the change follows the patternwe outlined and analyzed.As thecompany grew and its product line broadened, the number of things it had tokeeptrackofmultiplied.Itmademoreandmoresensetocreateanorganizationserving each product line; hence the three product divisions. But as the newsreleaseindicates,themajorfunctionalorganizationsofABCTechnologies,suchas sales and manufacturing, will remain centralized and will serve the threemission-orientedorganizations.Here I would like to propose Grove’s Law:All large organizations with a

commonbusinesspurposeendupinahybridorganizationalform.The Breakfast Factory, an army, Intel, and ABC Technologies provide

examples. But just about every large company or enterprise that I know isorganized inahybrid form.Takeaneducational institution inwhichone findsindividual mission-oriented departments such as mathematics, English,engineering, and so on, and also administration, composed of personnel,security, and library services, whose combined task is to supply the commonresourcesthateachoftheindividualdepartmentsneedstofunction.Another very different example of the hybrid form can be found in the

nationalJuniorAchievementorganization.Hereeachindividualchapterrunsitsownbusiness,with each decidingwhat product to sell, actually selling it, andotherwise maintaining all aspects of the business. Nevertheless, the nationalorganizationcontrolsthewaythechaptersaretogoabouttheirownpursuits:theform in which the individual businesses are to be structured, the paperworkrequirements,andtherewardsforsuccessfuloperation.Theuseof thehybridorganizational formdoesnotevennecessarilydepend

onhowlargeabusinessoractivityis.Afriendofmineisalawyerinamedium-size law firm. He told me how his firm tried to deal with the problems andconflictsheandhiscolleagueswerehavingoverresourcestheyallshared,suchas the steno pool and office space. They ended up forming an executivecommitteethatwouldnotinterferewiththelegal(mission-oriented)workoftheindividual attorneys but would address the acquisition and allocation ofcommon, shared resources. Here is a small operation finding itself with thehybridorganizationalform.Doanyexceptionsexisttotheuniversalityofhybridorganizations?Theonly

exceptions that come to my mind are conglomerates, which are typicallyorganizedinatotallymission-orientedform.Whyaretheyanexceptiontoourrule? Because they do not have a common business purpose. The variousdivisions(orcompanies)inthiscaseareallindependentandbearnorelationshipto one another beyond the conglomerate profit and loss statement. Butwithineachbusinessunitoftheconglomerate,theorganizationislikelytobestructuredalongthehybridline.Ofcourse,eachhybridorganization isuniquebecausea limitlessnumberof

points lie between the hypothetical extremes of the totally functional and thetotallymission-orientedforms.Infact,asingleorganizationmayverywellshiftbackandforthbetweenthetwopoles,movementthatshouldbebroughtonbypragmaticconsiderations.Forexample,acompanywithaninadequatecomputeracquires a large, powerful newone,makingpossible centralized economiesof

scale.Conversely,acompanyreplacesalargecomputerwithsmallinexpensiveonesthatcanbereadilyinstalledinvariousmission-orientedunitswithoutlossof the economies of scale. This is how a business can adapt. But the mostimportantconsiderationshouldbethis:theshiftbackandforthbetweenthetwotypesoforganizationscanandshouldbeinitiatedtomatchtheoperationalstylesandaptitudesofthemanagersrunningtheindividualunits.As I’ve said, sooneror later all reasonably large companiesmust copewith

the problems inherent in the workings of a hybrid organization. The mostimportanttaskbeforesuchanorganizationistheoptimumandtimelyallocationof its resources and the efficient resolution of conflicts arising over thatallocation.Thoughthisproblemmaybeverycomplex,“allocators”workingoutofsome

centralofficearecertainlynottheanswer.Infact, themostglaringexampleofinefficiencyI’veencounteredwentonsomeyearsagoinHungary,whereIoncelivedandwhereacentralplanningorganizationdecidedwhatgoodsweretobeproduced,when,andwhere.Therationaleforsuchplanningwasverysolid,butin practice it usually fell far, far short of meeting real consumer needs. InHungaryIwasanamateurphotographer.Duringthewinter,whenIneededhigh-contrastfilm,nonewastobefoundanywhere.Yetduringthesummer,everyonewasup tohiswaist in the stuff, even though regular filmwas in short supply.Year after year, decision-making in the central planning organization was soclumsythatitcouldnotevenrespondtototallypredictablechangesindemand.Inourbusinessculture,theallocationofsharedresourcesandthereconciliationof the conflicting needs and desires of the independent business units aretheoretically the function of corporate management. Practically, however, thetransactionloadisfartooheavytobehandledinoneplace.IfweatInteltriedtoresolve all conflicts and allocate all resources at the top, we would begin toresemblethegroupthatrantheHungarianeconomy.Instead,theanswerlieswithmiddlemanagers.Withinacompany,theyare,in

thefirstplace,numerousenoughtocovertheentirerangeofoperation;and,inthe second place, very close to the problem we’re talking about—namely,generating internal resources and consuming those resources. For middlemanagers tosucceedat thishigh-leverage task, two thingsarenecessary.First,theymustaccepttheinevitabilityofthehybridorganizationalformiftheyaretoserve itsworkings.Second, theymustdevelopandmaster thepractice throughwhichahybridorganizationcanbemanaged.Thisisdualreporting,thesubjectofournextchapter.

9DualReporting

Toput amanon themoon,NASAasked severalmajor contractors andmanysubcontractors towork together, each on a different aspect of the project. Anunintended consequence of the moon shot was the development of a neworganizationalapproach:matrixmanagement.Thisprovidedthemeansthroughwhichtheworkofvariouscontractorscouldbecoordinatedandmanagedsothatif problems developed in one place, they did not subvert the entire schedule.Resourcescouldbediverted,forexample,fromastrongorganizationtoonethatwasslippinginordertohelpthelattermakeuplosttime.Matrixmanagementisacomplicatedaffair.Bookshavebeenwrittenaboutit

and entire courses of instruction devoted to it. But the core idea was that aprojectmanager,somebodyoutsideanyofthecontractorsinvolved,couldwieldasmuch influence on thework of unitswithin a given company as could thecompany management itself. Thus, NASA elaborated the principle of dualreportingonagrandscale.Inreality,thebasicideahadbeenquietlyatworkformany years, enabling hybrid organizations of all types to function, from theneighborhoodhighschooltoAlfredSloan’sGeneralMotors—nottomentiontheBreakfast Factory franchises. Let’s re-create how Intel came to adopt a dualreportingsystem.

WhereShouldPlantSecurityReport?

When our company was young and small, we stumbled onto dual reportingalmost by accident.At a staffmeetingwewere trying to decide towhom thesecuritypersonnelatournewoutlyingplantsshouldreport.Wehadtwochoices.Onewouldhavetheemployeesreporttotheplantmanager.Butaplantmanager,bybackground, is typicallyanengineeroramanufacturingpersonwhoknowsverylittleaboutsecurityissuesandcaresevenless.Theotherchoicewouldhavethemreporttothesecuritymanageratthemainplant.Hehiredtheminthefirstplace,andhe is theexpertwhosets thestandards that thesecurityofficersaresupposed to adhere to throughout the company.And itwas clear that security

proceduresandpracticesattheoutlyingplantshadtoconformtosomekindofcorporatestandard.There was only one problem with the latter arrangement. The security

managerworksatcorporateheadquartersandnotat theoutlyingplant,sohowwouldheknowifthesecuritypersonneloutsidethemainplantevenshowedup,orcameinlate,orotherwiseperformedbadly?Hewouldn’t.Afterwewrestledwith thedilemmaforawhile, itoccurred tous thatperhapssecuritypersonnelshould report jointly to the corporate security manager and to the local plantmanager.Thefirstwouldspecifyhowthejoboughttobedone,andthesecondwouldmonitorhowitwasbeingperformeddaybyday.Whilethearrangementseemedtosolvebothproblems,thestaffcouldn’tquite

acceptit.Wefoundourselvesasking,“Apersonhastohaveaboss,sowhoisinchargehere?”Could an employee in fact have twobosses?The answerwas atentative “yes,” and the culture of joint reporting relationships, dual reporting,wasborn.Itwasaslow,laboriousbirth.Buttheneedfordualreportingisactuallyquitefundamental.Let’sthinkfora

minuteabouthowamanagercomestobe.Thefirststepinhiscareerisbeinganindividualcontributor—asalesman,forinstance.Ifheproveshimselfasuperiorsalesman,heispromotedtothepositionofsalesmanager,wherehesupervisespeople in his functional specialty, sales.When he has shown himself to be asuperstar sales manager he is promoted again, this time becoming a regionalsalesmanager.IfheworksatIntel,heisnownotonlysupervisingsalesmenbutalso so-called field application engineers, who obviously know more abouttechnical matters than he does but whom he still manages. The promotionscontinue until our superstar finds himself a general manager of a businessdivision.Amongotherthings,ournewgeneralmanagerhasnoexperiencewithmanufacturing. So while he is perfectly capable of supervising hismanufacturingmanagerinthemoregeneralaspectsofhisjob,thenewbosshasno choice but to leave the technical aspects to his subordinate, because as agraduate of sales, he has absolutely no background inmanufacturing. In otherdivisionsofthecorporation,manufacturingmanagersmaysimilarlybereportingtopeoplewhorosethroughtheranksofengineeringandfinance.We could handle the problem by designating one person the senior

manufacturing manager and having all the manufacturing managers report tohim insteadof to thegeneralmanager.But themorewedo this, themorewemovetowardatotallyfunctionalformoforganization.Ageneralmanagercould

no longer coordinate the activities of the finance,marketing, engineering, andmanufacturing groups toward a single business purpose responsive tomarketplaceneeds.Wewanttheimmediacyandtheoperatingprioritiescomingfrom the generalmanager aswell as a technical supervisory relationship. Thesolutionisdualreporting.But does the technical supervisor’s role have to be filled by a single

individual?No.Considerthefollowingscenario,whichcouldbetakenfromanordinary day at Intel. Our manufacturing manager is sitting in the cafeteriahaving a cup of coffee, and themanufacturingmanager from another division(whoseboss,thegeneralmanager,hasafinancebackground)comesover.Theystart chatting aboutwhat’s going on in their respective divisions and begin torealizethattheyhaveanumberoftechnicalproblemsincommon.Applyingthetheorythattwoheadsarebetterthanone,theydecidetomeetabitmoreoften.Eventually the meetings become regularly scheduled and manufacturingmanagers fromotherdivisions join the two toexchangeviewsaboutproblemsthey share.Pretty soona committeeor a councilmadeupof agroupofpeerscomesintoexistencetotackleissuescommontoall.Inshort,theyhavefoundawaytodealwiththosetechnical issuesthat theirbosses, thegeneralmanagers,can’t help them with. In effect, they now have supervision that a generalmanager competent in manufacturing could have given them, but thatsupervision is being exercised by a peer group. Themanufacturingmanagersreporttotwosupervisors:tothisgroupandtotheirrespectivegeneralmanagers,asthefigureoppositeshows.To make such a body work requires the voluntary surrender of individual

decision-making to the group. Being a member means you no longer havecomplete freedom of individual action, because you must go along with thedecisionsofyourpeersinmostinstances.Byanalogy,thinkofyourselfasoneofacouplewhodecidestotakeavacationwithanothercouple.Youknowthatifyougotogetheryouwillnotbefreetodoexactlywhatyouwanttodowhenyouwant todo it,butyougo togetheranywaybecauseyou’llhavemorefun,evenwhile you’ll have less freedom. At work, surrendering individual decision-making depends on trusting the soundness of actions taken by your group ofpeers.

Themanufacturingmanagersreporttotwosupervisors:totheirgeneralmanagersandtoagroupoftheirpeers.

Trustinnowayrelatestoanorganizationalprinciplebutisinsteadanaspectofthecorporateculture,somethingaboutwhichmuchhasbeenwritteninrecentyears.Putsimply,itisasetofvaluesandbeliefs,aswellasfamiliaritywiththewaythingsaredoneandshouldbedoneinacompany.Thepointisthatastrongand positive corporate culture is absolutely essential if dual reporting anddecision-makingbypeersaretowork.This systemmakes amanager’s life ambiguous, andmost people don’t like

ambiguity. Nevertheless, the system is needed to make hybrid organizationswork,andwhilepeoplewillstrivetofindsomethingsimpler,therealityisthatitdoesn’t exist. A strictly functional organization, which is clear conceptually,tends to remove engineering and manufacturing (or the equivalent groups inyour firm) from the marketplace, leaving them with no idea of what thecustomers want. A highly mission-oriented organization, in turn, may havedefinitecrispreportingrelationshipsandclearandunambiguousobjectivesatalltimes.However, the fragmented stateof affairs that results causes inefficiencyandpooroverallperformance.It’snotbecauseIntellovedambiguitythatwebecameahybridorganization.

We have tried everything else, and while other models may have been lessambiguous, they simply didn’t work. Hybrid organizations and theaccompanyingdualreportingprinciple,likeademocracy,arenotgreatinandofthemselves. They just happen to be the best way for any business to be

organized.

MakingHybridOrganizationsWork

Tomakehybridorganizationswork,youneedawaytocoordinatethemission-oriented units and the functional groups so that the resources of the latter areallocated and delivered to meet the needs of the former. Consider how thecontrollerworksatIntel.Hisprofessionalmethods,practices,andstandardsareset by the functional group to which he belongs, the finance organization.Consequently,thecontrollerforabusinessunitshouldreporttosomeoneinboththe functional and the mission-oriented organizations, with the type ofsupervision reflecting the varying needs of the two. The divisional generalmanagergives thecontrollermission-orientedprioritiesbyaskinghimtoworkon specific business problems. The finance manager makes sure that thecontrolleristrainedtodohisworkinatechnicallyproficientmanner,supervisesandmonitorshistechnicalperformance,andlooksafterhiscareerinsidefinance,promotinghim,perhaps,tothepositionofcontrollerofabigger,morecomplexdivisionifheperformswell.Again,asshownopposite,thisisdualreporting,themanagementprinciplethatenablesthehybridorganizationformtowork.The example has parallels throughout a corporation. Consider advertising.

Shouldeachbusinessdivisiondeviseandpursueitsownadvertisingcampaign,orshouldallofitbehandledthroughasinglecorporateentity?Asbefore,thereare pros and cons on both sides. Each division clearly understands its ownstrategy best, and therefore presumably best understands what its advertisingmessage should be and towhom it should be aimed. Thiswould suggest thatadvertisingstayinthehandsofthedivisions.Ontheotherhand,theproductsofvarious divisions often all serve the needs of a specific market, and takentogetherrepresentamuchmorecompletesolutiontothecustomers’needsthanwhatcanbeprovidedbyanindividualdivision.Herethecustomerandhencethemanufacturerclearlybenefit ifall theadvertisingstoriesare told inacoherent,coordinated fashion.Also, advertising sells not just a specific product but theentirecorporationaswell.Because theadsought toprojectaconsistent imagethatisrightforeverybody,weshouldattheveryleastnotletadivisiongooutandhireitsownadvertisingagency.

Thecontrollerforabusinessdivisionshouldbesupervisedbybothorganizations.

Aswithmuchelse inahybridorganization, theoptimumsolutionherecallsfortheuseofdualreporting.Thedivisionalmarketingmanagersshouldcontrolmost of their own advertising messages. But a coordinating body of peersconsistingofthevariousdivisionalmarketingmanagersandperhapschairedbythe corporatemerchandisingmanager should provide the necessary functionalsupervisionforallinvolved.Thisbodywouldchoosetheadvertisingagency,forinstance, and determine the graphic image to which all divisional ads shouldconform. It could also define theway the divisionmarketingmanagerswoulddealwith theagency,whichcould reduce thecostof space through theuseofvolumebuying.Yetthespecificsellingmessagecommunicatedbyanindividualadwouldbemainlylefttothedivisionalpeople.Dual reporting can certainly tax the patience of themarketingmanagers, as

they are now also required to understand the needs and thought processes oftheir peers. But no real alternative exists when you need to communicateindividualproductandmarketmessagesandmaintainacorporateidentityatthesametime.We have seen that all kinds of organizations evolve into a hybrid

organizational form. They must also develop a system of dual reporting.Consider the following story aboutOhioUniversity that appeared in theWall

StreetJournal(bracketedcommentsaremine):

Auniversityisanoddplacetomanage.ThepresidentoftheUniversitysaid,“There’sclearlyasharedresponsibilityfordecision-makingbetweenadministration[functionalorganization]andfaculty[mission-oriented organization].”AUniversity PlanningAdvisoryCouncil [a group of peers]was formedwithrepresentationfromthefacultyandadministrationtohelpallocatelimitedresources[amostdifficultandcommonproblem]inthefaceofseverebudgetcuts.“Wearebeingeducatedtothinkinstitutionally,”saidonecouncilmember.“I’mrepresentingstudentaffairs,whichhadsomeprojectsupforconsiderationthisyear.ButImadeabigpitchforbuyinganewbulldozer.”

Sotoputityetanotherway,thehybridorganizationalformistheinevitableconsequence of enjoying the benefits of being part of a large organization—acompanyoruniversityorwhatever.Tobesure,neither that formnor theneedfordualreportingisanexcuseforneedlessbusywork,andweshouldmercilesslyslashawayunnecessarybureaucratichindrance,applyworksimplificationtoallwedo,andcontinuallysubjectallestablishedrequirementsforcoordinationandconsultation to the test of common sense.Butwe shouldnot expect to escapefrom complexity by playing with reporting arrangements. Like it or not, thehybridorganizationisafundamentalphenomenonoforganizationallife.

AnotherWrinkle:TheTwo-PlaneOrganization

Whenever a person becomes involved in coordination—something not part ofhisregulardailywork—weencounterasubtlevariationofdualreporting.Remember Cindy, the know-how manager responsible for maintaining and

improving a specific manufacturing process? Cindy reports to a supervisingengineer, who in turn reports to the engineeringmanager of the plant. In herdaily work, Cindy keeps things going by manipulating the manufacturingequipment, watching the process monitors, and making adjustments whennecessary. But Cindy has another job, too. Shemeets formally once amonthwithhercounterparts fromtheotherproductionplants to identify,discuss,andsolveproblemsrelatedtotheprocessforwhichtheyareeachresponsibleintheirrespectiveplants.Thiscoordinatinggroupalsoworkstostandardizeproceduresusedatallplants.TheworkofCindy’sgroup,andotherslikeit,issupervisedbyanothermoreseniorgroup(calledtheEngineeringManagersCouncil),whichismadeupoftheengineeringmanagersfromalltheplants.Cindy’svariousreportingrelationshipscanbefoundin thefigurebelow.As

wecansee,asaprocessengineerintheproductionplant,whereshespends80percent of her time, Cindy has a clear, crisp reporting arrangement to hersupervisingengineer,andthroughhim,totheplantengineeringmanager.Butas

a member of the process coordinating group, she is also supervised by itschairman.SoweseethatCindy’snameappearsontwoorganizationchartsthatserve two separatepurposes—one tooperate theproductionplant, theother tocoordinate the efforts of various plants. Again we see dual reporting becauseCindyhastwosupervisors.

Cindy’snameappearsontwoorganizationcharts—coordinatinggroupsareameansforknow-howmanagerstoincreasetheirleverage.

Cindy’stworesponsibilitieswon’tfitonasingleorganizationchart.Instead,wehavetothinkofthecoordinatinggroupasexistingonadifferentchart,oronadifferentplane.Thissoundscomplicatedbutreallyisn’t.IfCindybelongedtoachurch,shewouldberegardedamemberofthatorganizationaswellasbeingpartofIntel.Hersupervisorthere,asitwere,wouldbethelocalpastor,whointurnisamemberofthechurchhierarchy.Noonewouldconfusethesetworoles:clearly operating on different planes, each has its own hierarchies, and thatCindyisamemberofbothgroupsatthesametimewouldhardlytroubleanyone.Cindy’sbeingpartofacoordinatinggroupislikeherchurchmembership.Our ability to use Cindy’s skill and know-how in two different capacities

makesitpossibleforhertoexertamuchlargerleverageatIntel.Inhermainjob,her knowledge affects the work that takes place in one plant; in her second,throughwhatshedoesintheprocesscoordinatinggroup,shecaninfluencethework of all plants. So we see that the existence of such groups is a way formanagers,especiallyknow-howmanagers,toincreasetheirleverage.Thetwo-planeconceptisapartofeverydayorganizationallife.Forinstance,

whilepeoplemostlyworkatanoperatingtask,theyalsoplan.Thehierarchyofthecorporation’splanningbodiesliesonaplaneseparatefromtheoneonwhichyou’ll find the operating groups. Moreover, if a person can operate in twoplanes, he can operate in three. Cindy could also be part of a task force toachieve a specific result inwhich her expertise is needed. This is as ifCindywere towork at Intel, to belong to a church, and to do advisorywork for thetown’sparksdepartment.Theseareseparaterolesanddonotconflictwithoneanother,thoughtheyalldovieforCindy’stime.It could also turn out that people who are in a subordinate/supervisory

relationship in one plane might find the relationship reversed in another. Forexample, I am president of Intel, but in another plane I am a member of astrategic planning group, where I report to its chairman, who is one of ourdivisioncontrollers.It’sasifIwereamemberoftheArmyReserve,andonourweekend exercises I foundmyself under the command of a regimental leaderwhohappenstobethisdivisioncontroller.Backattheoperatingranch,Imaybehissupervisororhissupervisor’ssupervisor,butintheArmyReserveheismycommandingofficer.The point is that the two- (or multi-) plane organization is very useful.

WithoutitIcouldonlyparticipateifIwereinchargeofeverythingIwaspartof.I don’t have that kind of time, and often I’m not the most qualified personaround to lead. The multi-plane organization enables me to serve as a footsoldier rather than as a general when appropriate and useful. This gives theorganizationimportantflexibility.Manyofthegroupsthatwearetalkingaboutherearetemporary.Some,like

task forces, are specifically formed for a purpose, while others aremerely aninformalcollectionofpeoplewhowork together tosolveaparticularproblem.Both cease towork as a grouponce the problemhas beenhandled.Themorevaried thenatureof theproblemswe face and themore rapidly things changearound us, the more we have to rely on such specially composed transitoryteams tocopewithmatters. In theelectronicsbusiness,wecan’tpossiblyshift

formal organization fast enough to keep up with the pace of advancingtechnology.Thetechniquesthatwehavetomastertomakehybridorganizationswork—dualormultiplereportingandalsodecision-makingbypeergroups—arebothnecessaryifsuchtransitoryteamsaretowork.Thekeyfactorcommontoall is the use of cultural values as amode of control,whichwewill considernext.

10ModesofControl

Let’slookatthewaysinwhichouractionscanbecontrolledorinfluenced.Sayyouneednewtiresforyourcar.Yougodownthestreettothedealerandtakealookatthevariouslineshehastooffer.Thenyou’llprobablygoupthestreettoseewhatthecompetitionhas.Maybelateryou’llturntoaconsumermagazinetohelpyou choose.Eventually, you’llmake adecisionbasedonone thing: yourownself-interest.Youwanttobuythetiresyouthinkwillmeetyourneedsatthelowestcosttoyou.Itisquiteunlikelythatanypersonalfeelingstowardthetiredealerwillcometomind.Youarenotconcernedabouthiswelfare—there’snotmuchchancethatyouwouldsaytohimthatheisn’tchargingyouenoughforthetires.Nowyouhavethetiresonyourcarandyoudriveoff.Afterawhile,youcome

toaredlight.Youstop.Doyouthinkaboutit?No.It’salawestablishedbythesociety at large that everybody stops at a red light and you unquestioninglyacceptandlivebyit.Vehicularchaoswouldreignifalldrivershadnotenteredintoacontract tostop.The trafficcopmonitorsadherenceandpenalizes thosewhobreakthelaw.After the light changes, you continueondown the road and comeupon the

sceneofamajoraccident.Quitelikely,you’llforgetaboutlawslikenotstoppingon a freeway and also forget about your own self-interest: you’ll probably doeverything you can to help the accident victims and, in themeantime, exposeyourselftoallkindsofdangersandrisks.Whatmotivatesyounowisnotatallwhatdidwhenyouwereshoppingfortiresorstoppingattheredlight:notself-interestorobeyingthelaw,butconcernaboutsomeoneelse’slife.Similarly, our behavior in a work environment can be controlled by three

invisibleandpervasivemeans.Theseare:

•free-marketforces•contractualobligations•culturalvalues

Free-MarketForces

Whenyouboughtyourtires,youractionsweregovernedbyfree-marketforces,whicharebasedonprice:goodsandservicesarebeingexchangedbetweentwoentities (individuals, organizational units, or corporations), with each seekingonlytoenrichhimselforitself.Thisisverysimple.Itisamatterof“Iwanttobuy the tire at the lowestprice I canget”versus“Iwant to sell the tire at thehighestpriceIcanget.”Neitherpartyherecaresiftheothergoesbankrupt,nordotheypretendto.Thisisaveryefficientwaytobuyandsell tires.Nooneisneeded to oversee the transaction because everyone is openly serving his ownself-interest.So why aren’t the forces of the marketplace used all the time in all

circumstances?Because towork, thegoodsandservicesboughtandsoldmustpossessaveryclearlydefineddollarvalue.Thefreemarketcaneasilyestablishapriceforsomethingassimpleastires.Butformuchelsethatchangeshandsinaworkorbusinessenvironment,valueishardtoestablish.

ContractualObligations

Transactionsbetweencompaniesareusuallygovernedbythefreemarket.Whenwebuyacommodityproductfromavendor,weare tryingtoget itat thebestpossibleprice,andviceversa.Butwhathappenswhenthevalueofsomethingisnoteasilydefined?Whathappens,forinstance,whenittakesagroupofpeopleto accomplish a certain task?Howmuch does each of them contribute to thevaluethebusinessaddstotheproduct?Thepointisthathowmuchanengineerisworthinagroupcannotbepinneddownbyappealingtothefreemarket.Infact, if we bought engineering work by the “bit,” I think we would end upspendingmoretimetryingtodecidethevalueofeachbitofcontributionthanthecontribution itself is worth. Here trying to use free-market concepts becomesquiteinefficient.Soyousaytotheengineers,“Okay,I’llretainyourservicesforayearforaset

amount ofmoney, and youwill agree to do a certain type ofwork in return.We’venowentered intoacontract. I’llgiveyouanofficeanda terminal, andyoupromisemetodothebestyoucantoperformyourtask.”Thenatureof control isnowbasedoncontractualobligations,whichdefine

the kind ofwork youwill do and the standards thatwill govern it.Because Ican’tspecifyinadvanceexactlywhatyouwilldofromdaytoday,Imusthavea

fairamountofgeneralizedauthorityoveryourwork.Soyoumustgivetomeaspartof thecontract the right tomonitorandevaluateand, ifnecessary,correctyourwork.Weagreeonotherguidelinesandworkout rules thatwewillbothobey.Inreturnforstoppingataredlight,wecountonotherdriverstodothesame

thing, and we can drive through green lights. But for lawbreakers we needpolicemen,andwiththem,aswithsupervisors,weintroduceoverhead.Whataresomeotherexamplesofcontractualobligation?Takethetaxsystem.

Wesurrender the right tosomeofwhatweearnandexpectcertainservices inreturn.Giantoverheadisnecessarytomonitorandauditourtaxreturns.Autilitycompanypresentsanotherexample.Itsrepresentativeswillgotosomebodywhoworks for the government and say, “I’ll build a three-hundred-million-dollargenerating plant and provide electricity for this portion of the state if youpromisemethatnooneelsewillbuildoneandtrytosellelectricityhere.”Thestatesays,“Well,that’sfine,butwe’renotgoingtoletyouchargewhateveryouwantforthepoweryougenerate.We’llestablishamonitoringagencycalledthePublic Utilities Commission and they’ll tell you how much you can chargeconsumers and how much profit you can make.” So, in exchange for amonopoly, the company is contractually obliged to accept the government’sdecisiononpricingandprofit.

CulturalValues

Whentheenvironmentchangesmorerapidlythanonecanchangerules,orwhenasetofcircumstancesissoambiguousandunclear thatacontractbetweentheparties that attempted to cover all possibilities would be prohibitivelycomplicated, we need another mode of control, which is based on culturalvalues.Itsmostimportantcharacteristicisthattheinterestofthelargergrouptowhichanindividualbelongstakesprecedenceovertheinterestoftheindividualhimself.When suchvalues are atwork, someemotionally loadedwords comeinto play—words like trust—because you are surrendering to the group yourabilitytoprotectyourself.Andforthistohappen,youmustbelievethatyouallshareacommonsetofvalues,acommonsetofobjectives,andacommonsetofmethods. These, in turn, can only be developed by a great deal of common,sharedexperience.

TheRoleofManagement

Youdon’tneedmanagementtosupervisetheworkingsoffree-marketforces;noone supervises sales made at a flea market. In a contractual obligation,managementhasaroleinsettingandmodifyingtherules,monitoringadherenceto them, and evaluating and improving performance. As for cultural values,managementhas todevelopandnurture thecommonsetofvalues,objectives,andmethodsessentialfortheexistenceoftrust.Howdowedothat?Onewayisbyarticulation,byspellingoutthesevalues,objectives,andmethods.Theother,even more important, way is by example. If our behavior at work will beregardedasinlinewiththevaluesweprofess,thatfostersthedevelopmentofagroupculture.

TheMostAppropriateModeofControl

There isa temptation to idealizewhat I’vecalledculturalvaluesasamodeofcontrol because it is so “nice,” even utopian, because everybody presumablycaresaboutthecommongoodandsubjugatesself-interesttothatcommongood.But this isnot themostefficientmodeofcontrolunderallconditions. It isnoguide to buying tires, nor could the tax system work this way. Accordingly,given a certain set of conditions, there is always amost appropriatemode ofcontrol,whichweasmanagersshouldfindanduse.How do we do that? There are two variables here: first, the nature of a

person’s motivation; and second, the nature of the environment in which heworks. An imaginary composite index can be applied to measure anenvironment’scomplexity,uncertainty,andambiguity,whichwe’llcalltheCUAfactor.Cindy, the process engineer, is surroundedby tricky technologies, newandnotfullyoperationalequipment,anddevelopmentengineersandproductionengineerspullingherinoppositedirections.Herworkingenvironment,inshort,iscomplex.Bruce,themarketingmanager,hasaskedforpermissiontohiremorepeople forhisgrosslyunderstaffedgroup;his supervisorwaffles, andBruce isleftwithno idea ifhe’llget thego-aheadorwhat todo ifhedoesn’t.Bruce’sworking environment is uncertain.Mike, whomwewill now introduce as anIntel transportationsupervisor,had todealwithsomanycommittees,councils,anddivisionalmanufacturingmanagers thathedidn’tknowwhich, if any,endwas up. He eventually quit, unable to tolerate the ambiguity of his workingenvironment.

Itisourtaskasmanagerstoidentifywhichmodeofcontrolismostappropriate.

Let’s now conceive a simple chart with four quadrants, shown above. Theindividualmotivationcanrunfromself-interesttogroup-interest,andtheCUAfactorofaworkingenvironmentcanvary from low tohigh.Nowlook for thebestmodeofcontrolforeachquadrant.Whenself-interestishighandtheCUAfactorislow,themostappropriateisthemarketmode,whichgovernedourtirepurchase.Asindividualmotivationmovestowardgroupinterest,thecontractualmodebecomesappropriate,whichgovernedourstoppingforaredlight.Whengroup-interestorientationandtheCUAfactorarebothhigh,theculturalvaluesmodebecomesthebestchoice,whichexplainstouswhywetriedtohelpatthesceneoftheaccident.Andfinally,whentheCUAfactorishighandindividualmotivation is based on self-interest, no mode of control will work well. Thissituation,likeeverymanforhimselfonasinkingship,canonlyproducechaos.Let’s apply our model to the work of a new employee. What is his

motivation? It is verymuch based on self-interest. So you should give him aclearlystructured jobwitha lowCUAfactor. Ifhedoeswell,hewillbegin tofeelmore at home,worry less about himself, and start to caremore about histeam.Helearnsthatifheisonaboatandwantstogetahead,itisbetterforhim

tohelprowthantoruntothebow.Theemployeecanthenbepromotedintoamore complex, uncertain, ambiguous job. (These tend to pay more.) As timepasses,hewillcontinuetogainanincreasingamountofsharedexperiencewithothermembersof theorganization andwill be ready to tacklemore andmorecomplex, ambiguous, and uncertain tasks. This iswhy promotion fromwithintendstobetheapproachfavoredbycorporationswithstrongcorporatecultures.Bringyoungpeopleinatrelativelylow-level,well-definedjobswithlowCUAfactors,andovertimetheywillshareexperienceswiththeirpeers,supervisors,and subordinates and will learn the values, objectives, and methods of theorganization.Theywillgraduallyaccept,evenflourishin,thecomplexworldofmultiplebossesandpeerdecision-making.Butwhatdowedowhenforsomereasonwehavetohireaseniorpersonfrom

outsidethecompany?Likeanyothernewhire,shetoowillcomeinhavinghighself-interest,butinevitablywewillgiveheranorganizationtomanagethatisintrouble;afterall,thatwasourreasonforgoingoutside.Sonotonlydoesournewmanagerhaveatoughjobfacingher,butherworkingenvironmentwillhaveaveryhighCUA.Meanwhile,shehasnobaseofcommonexperiencewiththerestoftheorganizationandnoknowledgeofthemethodsusedtohelpherwork.Allwecandoiscrossourfingersandhopeshequicklyforgetsself-interestandjustasquicklygetsontopofherjobtoreduceherCUAfactor.Shortofthat,she’sprobablyoutofluck.

ModesofControlatWork

At any one time, one of the threemodes of controlmay governwhatwe aredoing.Butfromonedaytothenext,wefindourselvesinfluencedbyall three.Let’strackBob’smodeofcontrolforabit.WhenBob,amarketingsupervisor,buyshislunchinthecafeteria,he’sinfluencedbymarketforces.Hischoicesarewelldefinedandbasedonwhathewantstobuyandwhathewantstopay.Bob’scoming to work in the first place represents a transaction governed bycontractualobligations.Heispaidasetsalaryfordoinghisbest,whichimpliesthathehas toshowup.Andhiswillingness toparticipate instrategicplanningactivitiesshowsculturalvaluesatwork.Thisisworkoutsideofhis“regular”jobasdefined contractually, and so represents extra effort for him.But hedoes itbecausehefeelsthecompanyneedswhathehastocontribute.Let’snowconsiderwhatgoesonduringthecourseofaworkproject.Aswe

know,Barbara’sdepartmentisresponsiblefortrainingtheIntelsalesforceinher

division’sproducts.Whenshebuysmaterialsusedinthetrainingprogram,free-marketforcesreignasbindersoftherequiredqualityarepurchasedatthelowestpossibleprice.Theexistenceofthetrainingprogramitself,however,presentsanexample of contractual obligations at work. The salespeople expect that eachdivision will provide training on a regular basis. While the program isn’t amandated requirement spelledout somewhere in a formalpolicy statement, itsbasisisnonethelesscontractual.Thepointis,expectationscanbeasbindingasalegaldocument.Whenanumberofdivisionsshareacommonsalesforce,eachofthemhasa

vested interest to train representatives to promote and sell its products.At thesame time,unless thedivisions arewilling to sacrifice self-interest in favorofthecommoninterest,thetrainingsessionscaneasilybecomedisjointedfree-for-allsandconfuseeverybody.Sotheneedtohavetheindividualdivisionspresentcoordinated messages is governed by corporate values. Thus, in field salestraining,wefindallthreemodesofcontrolatwork.Recentlyagroupoffactorymarketingmanagersclaimedthatoursalespeople

weregovernedonlyby self-interest.They said that theydevotedmostof theirattentiontosellingthoseitemsthatproducedthemostcommissionsandbonuses.Irritated and a bit self-righteous, the managers felt they were much moreconcernedaboutthecommongoodofthecompanythanweretheircolleaguesinthefield.But the marketing departments themselves created the monster. To get the

sales force to favor particular products, the divisions had for some time beenrunningcontests,withprizesrangingfromcashbonusestotripstoexoticplaces.Themarketingmanagerswere competing against one another for a finite andvaluableresource: thesalesmen’s time.And thesalesmenmerelyrespondedasonemightexpect.Butsalespeoplecanalsobehaveintheoppositefashion.Atonetime,oneof

ourdivisionshadseriousproblems,leavingthesalesengineerswithnoproducttosellfornearlyayear.TheycouldhaveleftIntelandimmediatelygottenotherjobs and quick commissions elsewhere, but by and large they stayedwith us.Theystayedbecausetheybelievedinthecompanyandhadfaiththateventuallythingswouldgetbetter.Beliefandfaitharenotaspectsofthemarketmode,butstemfromadherencetoculturalvalues.

IV

ThePlayers

11TheSportsAnalogy

EarlierIbuiltacasesummedupbythekeysentence:Amanager’soutputistheoutputoftheorganizationunderhissupervisionorinfluence.Putanotherway,thismeansthatmanagementisateamactivity.Butnomatter

howwellateamisputtogether,nomatterhowwellitisdirected,theteamwillperformonlyaswellastheindividualsonit.Inotherwords,everythingwe’veconsideredsofarisuselessunlessthemembersofourteamwillcontinuallytrytoofferthebesttheycando.Themeansamanagerhasathisdisposaltoelicitpeakindividualperformancearewhattherestofthisbookisabout.Whenapersonisnotdoinghisjob,therecanonlybetworeasonsforit.The

personeithercan’tdoitorwon’tdoit;heiseithernotcapableornotmotivated.To determinewhich,we can employ a simplemental test: if the person’s lifedependedondoingthework,couldhedoit?Iftheanswerisyes,thatpersonisnotmotivated; if the answer is no, he is not capable. Ifmy life depended onplayingtheviolinoncommand,Icouldnotdoit.ButifIhadtorunamileinsixminutes,Iprobablycould.NotthatIwouldwantto,butifmylifedependedonit,Iprobablycould.The single most important task of a manager is to elicit peak performance

from his subordinates. So if two things limit high output, amanager has twowaystotackletheissue:throughtrainingandmotivation.Each,asweseeinthenextfigure,canimproveaperson’sperformance.Inthischapter,ourconcernismotivation.

Amanagerhastwowaystoimproveperformance:trainingandmotivation.

How does a manager motivate his subordinates? For most of us, the wordimpliesdoing something to anotherperson.But I don’t think that canhappen,because motivation has to come from within somebody. Accordingly, all amanager can do is create an environment in which motivated people canflourish.Becausebettermotivationmeansbetterperformance,notachangeofattitude

or feeling, a subordinate’s saying “I feel motivated” means nothing. Whatmattersis ifheperformsbetterorworsebecausehisenvironmentchanged.Anattitudemay constitute an indicator, a “window into the black box” of humanmotivation, but it is not the desired result or output. Better performance at agivenskilllevelis.For most of Western history, including the early days of the Industrial

Revolution, motivation was basedmostly on fear of punishment. In Dickens’time,thethreatoflossoflifegotpeopletowork,becauseifpeopledidnotwork,theywerenotpaidandcouldnotbuyfood,andiftheystolefoodandgotcaught,they were hanged. The fear of punishment indirectly caused them to producemorethantheymighthaveotherwise.

Over thepast thirtyyearsorso,anumberofnewapproacheshavebeguntoreplace older practices keyed to fear. Perhaps the emergence of the new,humanisticapproachestomotivationcanbetracedtothedeclineintherelativeimportanceofmanuallaborandthecorrespondingriseintheimportanceofso-calledknowledgeworkers.Theoutputofamanuallaborerisreadilymeasurable,anddeparturesfromtheexpectedcanbespottedanddealtwithimmediately.Butforaknowledgeworker,suchdeparturestakelongertodeterminebecauseeventheexpectationsthemselvesareverydifficulttostateprecisely.Inotherwords,fearwon’tworkaswellwithcomputerarchitectsaswithgalleyslaves;hence,newapproachestomotivationareneeded.My description of what makes people perform relies heavily on Abraham

Maslow’s theory of motivation, simply because my own observations ofworkinglifeconfirmMaslow’sconcepts.ForMaslow,motivationiscloselytiedtotheideaofneeds,whichcausepeopletohavedrives,whichinturnresultinmotivation.Aneedoncesatisfiedstopsbeinganeedandthereforestopsbeingasourceofmotivation.Simplyput,ifwearetocreateandmaintainahighdegreeofmotivation,wemustkeepsomeneedsunsatisfiedatalltimes.People,ofcourse,tendtohaveavarietyofconcurrentneeds,butoneamong

them is always stronger than the others.And that need is the one that largelydetermines an individual’s motivation and therefore his level of performance.Maslowdefinedasetofneeds,asshownbelow,thattendtolieinahierarchy:whenalowerneedissatisfied,onehigherislikelytotakeover.

Maslowdefinedasetofneedsthattendtolieinahierarchy:whenalowerneedissatisfied,onehigherislikelytotakeover.

PhysiologicalNeeds

Theseneedsconsistofthingsthatmoneycanbuy,likefood,clothing,andotherbasic necessities of life. Fear is hitched to such needs: one fears the possibledeprivationoffood,clothing,andsoon.

Security/SafetyNeeds

These come from a desire to protect oneself from slipping back to a state ofbeingdeprivedof thebasicnecessities.Safetyand securityneedsare fulfilled,forexample,whenmedicalinsuranceprovidesemployeesprotectionagainstthefearofgoingbankrupt trying topaydoctorandhospital fees.Theexistenceofbenefitsisrarelyadominantsourceofemployeemotivation,butifbenefitswereabsentandemployeeshadtoworryaboutsuchconcerns,performancewouldnodoubtbebadlyaffected.

Social/AffiliationNeeds

The social needs stem from the inherent desire of human beings to belong tosomegroupor other.But people don’twant to belong to just anygroup; theyneed to belong to one whose members possess something in common withthemselves. For example, when people are excited, confident, or happy, theywanttobearoundpeoplewhoarealsoexcited,confident,orhappy.Conversely,miserylovesnotjustanycompany,butthecompanyofothermiserablepeople.Nobodywhoismiserablewantstobearoundsomeonehappy.Socialneedsarequitepowerful.Afriendofminedecidedtogobacktowork

aftermanyyears ofmindingher home.She took a low-paying job,whichdidlittleforherfamily’sstandardofliving.Foralongtime,Ididn’tunderstandwhyshedidwhatshedid,butfinallyitdawnedonme:sheneededthecompanionshipher work offered. Going to work meant being around a group of people sheliked.Another example of the power of social needs is provided by Jim, a young

engineer.His first job after he graduated from collegewaswith a very large,long-established company, while his two college roommates came to work atIntel. Because Jim continued to room with them, he was exposed to whatworkingwithinIntelwas like.Moreover,mostofhis roommates’ friendsfromworkwerealsoyoung,unmarried,and justayearor twooutofcollege,whilemostofthepeoplewhereJimworkedweremarriedandatleasttenyearsolder.Jimfeltleftout,andhisneedforagroupinwhichhefeltcomfortablepromptedhimtocometoworkatIntel,thoughheverymuchenjoyedhisworkattheothercompany.As one’s environment or condition in life changes, one’s desire to satisfy a

particularsetofneedsisreplacedbyadesiretosatisfyanotherset.There’sthestoryofayoungIntelmanager,Chuck,whenhewasafirst-yearstudentat theHarvard Business School. Initially, he was engulfed by a fear of the classmaterial,ofhisprofessors,offailure,offlunkingout.Afterawhilehisfeargaveway to the realization that everyone else was in the same boat, also afraid.Studentsbegantoformstudygroupswhoseostensiblepurposewastoconsiderclassmaterial together, but whose real purpose was to strengthen confidence.Chuckmoved from being governed largely by his need for sheer survival—a“physiological”need—tooneforsecurityandsafety.Astimewenton,thestudygroupsdissolvedandthestudentsstartedtoassociatewithothermembersoftheclass.The entire class, or “section,” as itwas called, developed adefinite andrecognizablesetofcharacteristics;itbecame,inshort,ateam.Membersenjoyedbelonging, associating, and identifying with it, and worked to sustain the

section’s image among the professors and other students. Chuck was nowsatisfyinghisneedforaffiliation.Of course, regressive movement is also possible. Recently, a highly

motivated, smoothlyworking team ofmanufacturing employees in one of ourCaliforniaplantswassuddenly jolted—all too literally—fromsatisfyingaveryhigh level of human needs to abandoning an inventory of silicon wafers,expensive manufacturing equipment, even friends. An earthquake shook theirfactory.Peoplefearedfortheirlives,droppedeverything,andrantothenearestexitas theyfoundthemselvestotallyconsumedbythemostfundamentalofallphysiologicalneeds—survival.The physiological, safety/security, and social needs all can motivate us to

show up for work, but other needs—esteem and self-actualization—make usperformoncewearethere.

Esteem/RecognitionNeeds

Theneedforesteemorrecognitionisreadilyapparentinthecliché“keepingupwiththeJoneses.”Suchstrivingiscommonlyfrownedupon,butifanathlete’s“Jones”islastyear’sOlympicgoldmedalist,orifanactor’s“Jones”isLaurenceOlivier, theneed tokeepupwithor emulate someone is apowerful sourceofpositive motivation. The person or group whose recognition you desire maymeannothingtosomeoneelse—esteemexistsintheeyesofthebeholder.Ifyouareanaspiringhighschoolathleteandoneofthetopplayerspassesyouinthehall and says hello, you’ll feel terrific. Yet if you try to tell your family orfriendshowpleasedyouwereabouttheencounter,youarelikelytobemetwithblankstares,becausethe“hello”meansnothingtopeoplewhoarenotaspiringathletesinyourhighschool.All of the sourcesofmotivationwe’ve talked about so far are self-limiting.

That is, when a need is gratified, it can no longermotivate a person.Once apredeterminedgoalorlevelofachievementisreached,theneedtogoanyfurtherloses urgency. A friend of mine was thrust into a premature “mid-life crisis”when,inrecognitionoftheexcellentworkhehadbeendoing,hewasnamedavice president of the corporation. Such a position had been a life-long goal.Whenhehadsuddenlyattainedit,hefoundhimselflookingforsomeotherwaytomotivatehimself.

Self-ActualizationNeeds

ForMaslow,self-actualizationstemsfromapersonalrealizationthat“whatIcanbe,Imustbe.”Thetitleofamovieaboutathletes,PersonalBest,captureswhatself-actualization means: the need to achieve one’s utter personal best in achosen field of endeavor. Once someone’s source of motivation is self-actualization, his drive to perform has no limit. Thus, its most importantcharacteristic is that unlike other sources of motivation, which extinguishthemselvesafter theneedsarefulfilled,self-actualizationcontinuestomotivatepeopletoeverhigherlevelsofperformance.Twoinnerforcescandriveaperson touseallofhiscapabilities.Hecanbe

competence-driven orachievement-driven.The former concerns itselfwith jobor taskmastery.Avirtuosoviolinistwhocontinues topracticedayafterday isobviouslymovedbysomethingotherthananeedforesteemandrecognition.Heworkstosharpenhisownskill, tryingtodoalittlebitbetterthistimethanthetimebefore,justasateenageronaskateboardpracticesthesametrickoverandoveragain.Thesameteenagermaynotsitstillfortenminutestodohomework,butonaskateboardheisrelentless,drivenbytheself-actualizationneed,aneedtogetbetterthathasnolimit.Theachievement-drivenpathtoself-actualizationisnotquitelikethis.Some

people—not themajority—aremovedbyanabstractneedtoachieveinallthatthey do.Apsychology lab experiment illustrated the behavior of such people.Somevolunteerswereputintoaroominwhichpegsweresetatvariousplaceson the floor.Each personwas given some rings but not instructedwhat to dowith them. People eventually started to toss the rings onto the pegs. Somecasuallytossedtheringsatfarawaypegs;othersstoodoverthepegsanddroppedthe rings down onto them. Still otherswalked just far enough away from thepegs so that to toss a ring onto a peg constituted a challenge. These peopleworkedattheboundaryoftheircapability.Researchers classified the three types of behavior. The first group, termed

gamblers,tookhighrisksbutexertednoinfluenceontheoutcomeofevents.Thesecondgroup,termedconservatives,werepeoplewhotookverylittlerisk.Thethirdgroup,termedachievers,hadtotest thelimitsofwhattheycoulddo,andwithnopromptingdemonstratedthepointoftheexperiment:namely,thatsomepeople simplymust test themselves. By challenging themselves, these peoplewere likely tomiss a peg several times, butwhen they began to ring the pegconsistently, theygained satisfaction anda senseof achievement.Thepoint isthatbothcompetence-andachievement-orientedpeoplespontaneouslytrytotesttheouterlimitsoftheirabilities.

Whentheneedtostretchisnotspontaneous,managementneedstocreateanenvironmenttofosterit.InanMBOsystem,forexample,objectivesshouldbesetatapointhighenoughsothateveniftheindividual(ororganization)pusheshimselfhard,hewillstillonlyhaveafifty-fiftychanceofmakingthem.Outputwilltendtobegreaterwheneverybodystrivesforalevelofachievementbeyondhisimmediategrasp,eventhoughtryingmeansfailurehalfthetime.Suchgoal-setting is extremely important if what you want is peak performance fromyourselfandyoursubordinates.Moreover,ifwewanttocultivateachievement-drivenmotivation,weneedto

createanenvironmentthatvaluesandemphasizesoutput.Myfirstjobwaswitharesearchanddevelopmentlaboratory,wherealotofpeoplewereveryhighlymotivated but tended to be knowledge-centered. They were driven to knowmore, but not necessarily to knowmore in order to produce concrete results.Consequently,relativelylittlewasactuallyachieved.ThevaluesystematInteliscompletelythereverse.ThePh.D.incomputersciencewhoknowsananswerinthe abstract, yet does not apply it to create some tangible output, gets littlerecognition, but a junior engineer who produces results is highly valued andesteemed.Andthatishowitshouldbe.

MoneyandTask-RelevantFeedback

We now come to the question of howmoneymotivates people. At the lowerlevelsofthemotivationhierarchy,moneyisobviouslyimportant,neededtobuythe necessities of life.Once there is enoughmoney to bring a person up to alevelheexpectsofhimself,moremoneywillnotmotivate.ConsiderpeoplewhoworkatourassemblyplantintheCaribbean.Thestandardoflivingthereisquitelow,andpeoplewhoworkforusenjoyonesubstantiallyhigherthanmostofthepopulation.Yet, in the early years of operation,many employeesworked justlongenoughtoaccumulatesomesmallsumofmoneyandthenquit.Forthem,money’smotivationwasclearlylimited;havingreachedapredeterminednotionof howmuchmoney theywanted,moremoney and a steady job provided nomoremotivation.Nowconsideraventurecapitalistwhoaftermakingtenmilliondollarsisstill

very hard at work trying tomake another ten. Physiological, safety, or socialneeds hardly apply here. Moreover, because venture capitalists usually don’tpublicize their successes, they are not driven by a need for esteem orrecognition.Soitappearsthatattheupperleveloftheneedhierarchy,whenone

isself-actualized,moneyinitselfisnolongerasourceofmotivationbutratherameasureofachievement.Moneyinthephysiological-andsecurity-drivenmodesonly motivates until the need is satisfied, but money as a measure ofachievementwillmotivatewithoutlimit.Thusthesecondtenmillioncanbejustas important to the venture capitalist as the first, since it is not the utilitarianneedforthemoneythatdriveshimbuttheachievementthatitimplies,andtheneedforachievementisboundless.Asimpletestcanbeusedtodeterminewheresomeoneisinthemotivational

hierarchy. If the absolute sum of a raise in salary an individual receives isimportanttohim,heisworkingmostlywithinthephysiologicalorsafetymodes.If,however,whatmatters tohim ishowhis raisestacksupagainstwhatotherpeoplegot,heismotivatedbyesteem/recognitionorself-actualization,becauseinthiscasemoneyisclearlyameasure.Once in the self-actualizationmode, a person needs measures to gauge his

progressandachievement.Themost important typeofmeasure is feedbackonhis performance. For the self-actualized person driven to improve hiscompetence, the feedback mechanism lies within that individual himself. Ourvirtuosoviolinistknowshowthemusicshouldsound,knowswhenitisnotright,and will strive tirelessly to get it right. Accordingly, if the possibility forimprovement does not exist, the desire to keeppracticingvanishes. I knewanOlympicfencingchampion,aHungarianwhoimmigratedtothiscountry.WhenIranintohimrecently,hetoldmethathehadquitfencingshortlyafterarrivingin theU.S.He said that the level of competitionherewasnot high enough toproduce someonewho could give him a contest, and that he couldn’t bear tofenceanylongerbecauseeverytimehedid,hefelthisskillwasdiminishing.What are some of the feedbackmechanisms ormeasures in theworkplace?

Themostappropriatemeasures tieanemployee’sperformanceto theworkingsoftheorganization.Ifperformanceindicatorsandmilestonesinamanagement-by-objectives systemare linked to theperformanceof the individual, theywillgaugehisdegreeofsuccessandwillenhancehisprogress.Anobviousandveryimportantresponsibilityofamanageristosteerhispeopleawayfromirrelevantandmeaninglessrewards,suchasofficesizeordecor,andtowardrelevantandsignificantones.Themostimportantformofsuchtask-relevantfeedbackistheperformancerevieweverysubordinateshouldreceivefromhissupervisor.Moreaboutthislater.

Fear

In physiological and security/safety need-dominated motivation, one fears thelossoflifeorlimborlossofjoborliberty.Doesfearhaveaplaceintheesteemorself-actualizedmodes?Itdoes,buthereitbecomesthefearoffailure.Butisthat a positive or negative source of motivation? It can be either. Given aspecific task, fear of failure can spur a person on, but if it becomes apreoccupation, a person driven by a need to achieve will simply becomeconservative.Let’s think back to the ring tossers. If a person got an electricalshockeachtimehethrewaringandmissed,soonenoughhewouldwalkovertothepeganddrop theringfromdirectlyover it toeliminate thepainassociatedwithfailure.In general, in the upper levels ofmotivation, fear is not something coming

fromtheoutside. It is insteadfearofnotsatisfyingyourself thatcausesyou tobackoff.Youcannotstayintheself-actualizedmodeifyou’realwaysworriedaboutfailure.

TheSportsAnalogy

We’vestudiedmotivationtotrytounderstandwhatmakespeoplewanttoworksothatasmanagerswecanelicitpeakperformancefromoursubordinates—their“personal best.”Of course,whatwe are really after is the performance of theorganization as a whole, but that depends on how skilled and motivated thepeoplewithinitare.Thus,ourroleasmanagersis,first,totraintheindividuals(tomovethemalongthehorizontalaxisshownintheillustrationonthispage),and,second,tobringthemtothepointwhereself-actualizationmotivatesthem,becauseoncethere,theirmotivationwillbeself-sustainingandlimitless.Isthereasystematicwaytoleadpeopletoself-actualization?Forananswer,

let’saskanotherquestion.Whydoesapersonwho isnot terribly interested inhis work at the office stretch himself to the limit running a marathon?Whatmakes him run?He is trying to beat other people or the stopwatch. This is asimple model of self-actualization, wherein people will exert themselves topreviouslyundreamedheights,forcingthemselvestorunfartherorfaster,whiletheireffortsfillbarrelswithsweat.Theywilldothisnotformoney,butjusttobeat the distance, the clock, or other people. Considerwhatmade Joe Frazierbox:

ItastoundsJoeFrazierthatanyonehastoaskwhyhefights.“ThisiswhatIdo.Iamafighter,”hesays.

“It’smyjob.I’mjustdoingmyjob.”Joedoesn’tdenytheattractivenessofmoney.“Whowantstoworkfornothing?”Buttherearethingsmoreimportantthanmoney.“Idon’tneedtobeastar,becauseIdon’tneedtoshine.ButIdoneedtobeaboxer,becausethat’swhatIam.It’sassimpleasthat.”

Imaginehowproductiveourcountrywouldbecomeifmanagerscouldendowallworkwiththecharacteristicsofcompetitivesports.To try to do this, we must first overcome cultural prejudice. Our society

respectssomeone’s throwinghimself intosports,butanybodywhoworksverylonghours is regardedassick,aworkaholic.So theprejudicesof themajoritysaythatsportsaregoodandfun,butworkisdrudgery,anecessaryevil,andinnowayasourceofpleasure.Thatmakestheclichéapply:ifyoucan’tbeatthem,jointhem—endowwork

withthecharacteristicsofcompetitivesports.Andthebestwaytogetthatspiritinto the workplace is to establish some rules of the game and ways foremployees tomeasure themselves.Elicitingpeakperformancemeansgoingupagainst somethingor somebody.Letmegiveyoua simpleexample.Foryearstheperformanceof the Intel facilitiesmaintenancegroup,which is responsiblefor keeping our buildings clean and neat, was mediocre, and no amount ofpressureorinducementseemedtodoanygood.Wetheninitiatedaprograminwhich each building’s upkeep was periodically scored by a resident seniormanager, dubbed a “building czar.” The score was then comparedwith thosegiven the other buildings. The condition ofall of them dramatically improvedalmostimmediately.Nothingelsewasdone;peopledidnotgetmoremoneyorother rewards.What they did getwas a racetrack, an arena of competition. Ifyourworkisfacilitiesmaintenance,havingyourbuildingreceivethetopscoreisa powerful source of motivation. This is key to the manager’s approach andinvolvement:hehastoseetheworkasitisseenbythepeoplewhodothatworkevery day and then create indicators so that his subordinates can watch their“racetrack”takeshape.Conversely, of course, when the competition is removed, motivation

associated with it vanishes. Consider the example of a newspaper columnistreflectingonhispast.Thisjournalist“thrivedonbeatingthecompetitioninthecolumn, and his pleasure in his work began to wane after [his paper and thecompetitive paper] merged. ‘I’ll never forget that day of the merger,’ thecolumnist said. ‘I walked out to get the train, and I just thought: There isn’tanyoneelsetobeat.’ ”Comparingourworktosportsmayalsoteachushowtocopewithfailure.As

noted,oneof thebigimpediments toafullycommitted,highlymotivatedstateof mind is preoccupation with failure. Yet we know that in any competitivesport,atleast50percentofallmatchesarelost.Allparticipantsknowthatfromtheoutset,andyetrarelydotheygiveupatanystageofacontest.Theroleofthemanagerhereisalsoclear:itisthatofthecoach.First,anideal

coachtakesnopersonalcreditforthesuccessofhisteam,andbecauseofthathisplayerstrusthim.Second,heistoughonhisteam.Bybeingcritical,hetriestoget the best performance his teammembers can provide. Third, a good coachwaslikelyagoodplayerhimselfatonetime.Andhavingplayedthegamewell,healsounderstandsitwell.Turning the workplace into a playing field can turn our subordinates into

“athletes”dedicated toperformingat the limitof their capabilities—thekey tomakingourteamconsistentwinners.

12Task-RelevantMaturity

I’ll say again that a manager’s most important responsibility is to elicit topperformancefromhissubordinates.Assumingweunderstandwhatmotivatesanemployee, the questionbecomes: Is there a single bestmanagement style, oneapproachthatwillworkbetterthanallothers?Many have looked for that optimum.Considering the issue historically, the

managementstylemostinfavorseemstohavechangedtoparallelthetheoryofmotivation espoused at the time.At the turn of the century, ideas aboutworkweresimple.Peopleweretoldwhattodo,andiftheydidit, theywerepaid;iftheydidnot,theywerefired.Thecorrespondingleadershipstylewascrispandhierarchical: therewere thosewhogaveordersand thosewho tookordersandexecuted them without question. In the 1950’s, management theory shiftedtoward a humanistic set of beliefs that held that therewas a nicerway to getpeople towork. The favored leadership style changed accordingly. Finally, asuniversity behavioral science departments developed and grew, the theories ofmotivationand leadershipbecamesubjectsofcarefullycontrolledexperiments.Surprisingly, none of the early intuitive presumptions could be borne out: thehardfindingssimplywouldnotshowthatonestyleofleadershipwasbetterthananother.Itwashardtoescapetheconclusionthatnooptimalmanagementstyleexisted.My own observations bear this out. At Intel we frequently rotate middle

managersfromonegrouptoanotherinordertobroadentheirexperience.Thesegroups tend tobe similar inbackgroundand in the typeofwork that theydo,although their output tends to vary greatly. Some managers and their groupsdemonstrate themselves to be higher producers; others do not. The result ofmoving the managers about is often surprising. Neither the managers nor thegroups maintain the characteristic of being either high-producing or low-producingasthemanagersareswitchedaround.Theinevitableconclusionisthathighoutput isassociatedwithparticularcombinationsofcertainmanagersandcertaingroupsofworkers.Thisalsosuggeststhatagivenmanagerialapproachis

notequallyeffectiveunderallconditions.Someresearchers in thisfieldargue that there isafundamentalvariable that

tells you what the best management style is in a particular situation. Thatvariable is the task-relevant maturity (TRM) of the subordinates, which is acombinationofthedegreeoftheirachievementorientationandreadinesstotakeresponsibility,aswellastheireducation,training,andexperience.Moreover,allthisisveryspecifictothetaskathand,anditisentirelypossibleforapersonoragroupofpeopletohaveaTRMthatishighinonejobbutlowinanother.LetmegiveyouanexampleofwhatImean.Werecentlymovedanextremely

productivesalesmanagerfromthefield into theplant,wherehewasplaced inchargeofafactoryunit.Thesizeandscopeofthetwojobswerecomparable,yetthe performanceof the seasonedmanager deteriorated, andhe started to showthesignsofsomeoneoverwhelmedbyhiswork.Whathappenedwasthatwhilethepersonalmaturityofthemanagerobviouslydidnotchange,histask-relevantmaturityinthenewjobwasextremelylow,sinceitsenvironment,content,andtaskswereallnew tohim. In timehe learned tocope,andhisTRMgraduallyincreased.With that,hisperformancebegan toapproach theoutstanding levelshe had exhibited earlier, which was whywe promoted him in the first place.Whathappenedhereshouldhavebeentotallypredictable,yetweweresurprised:we confused the manager’s general competence and maturity with his task-relevantmaturity.Similarly, a person’s TRM can be very high given a certain level of

complexity,uncertainty,andambiguity,butifthepaceofthejobacceleratesorifthejobitselfabruptlychanges,theTRMofthatindividualwilldrop.It’sabitlikeapersonwithmanyyears’experiencedrivingonsmallcountryroadsbeingsuddenlyasked todriveonacrowdedmetropolitan freeway.HisTRMdrivinghisowncarwilldropprecipitously.Theconclusionisthatvaryingmanagementstylesareneededastask-relevant

maturityvaries.Specifically,whentheTRMislow,themosteffectiveapproachisone thatoffersverypreciseanddetailed instructions,wherein thesupervisortells thesubordinatewhatneeds tobedone,when,andhow: inotherwords,ahighly structured approach. As the TRM of the subordinate grows, the mosteffectivestylemovesfromthestructuredtoonemoregiventocommunication,emotional support, and encouragement, in which the manager pays moreattentiontothesubordinateasanindividualthantothetaskathand.AstheTRMbecomesevengreater, theeffectivemanagement stylechangesagain.Here the

manager’s involvement should be kept to a minimum, and should primarilyconsist of making sure that the objectives toward which the subordinate isworkingaremutuallyagreedupon.ButregardlessofwhattheTRMmaybe,themanager should alwaysmonitor a subordinate’swork closely enough to avoidsurprises.The presence or absence ofmonitoring, aswe’ve said before, is thedifference between a supervisor’s delegating a task and abdicating it. Thecharacteristics of the effective management style for the supervisor given thevaryingdegreesofTRMaresummarizedinthetablebelow.Awordofcautionisinorder:donotmakeavaluejudgmentandconsidera

structured management style less worthy than a communication-oriented one.Whatis“nice”or“notnice”shouldhavenoplaceinhowyouthinkorwhatyoudo.Remember,weareafterwhatismosteffective.Thetheoryhereparallelsthedevelopmentoftherelationshipbetweenaparent

and child. As the child matures, the most effective parental style changes,varyingwiththe“life-relevantmaturity”—orage—ofthechild.Aparentneedstotellatoddlernottotouchthingsthathemightbreakorthatmighthurthim.The child cannot understand that the vase he wants to play with is anirreplaceable heirloom, but he can understand “no.” As he grows older, hebegins to do things on his own initiative, something the parent wants toencouragewhilestill trying tokeephimfrominjuringhimself.Aparentmightsuggest,forexample,thathischildgiveuphistricycleforhisfirsttwo-wheeler.Theparentwillnotsimplysendhimoutonhisown,butwillaccompanyhimtokeep the bicycle from tipping over while talking to him about safety on thestreets.As the child’smaturity continues to grow, the parent can cut back onspecific instruction.When the childgoesout to ridehisbicycle, theparentnolonger has to recite the litany of safety rules. Finally, when the life-relevantmaturityofthechildishighenough,heleaveshomeandperhapsgoesawaytocollege.Atthispointtherelationshipbetweenparentandchildwillchangeagainastheparentmerelymonitorsthechild’sprogress.

TASK-RELEVANTMATURITYOFSUBORDINATE CHARACTERISTICSOFTHEEFFECTIVEMANAGEMENTSTYLE

low Structured;task-oriented;tell“what,”“when,”“how”

medium Individual-oriented;emphasisontwo-waycommunication,support,mutualreasoning

high Involvementbymanagerminimal:establishingobjectivesandmonitoring

Thefundamentalvariablethatdeterminestheeffectivemanagementstyleisthetask-relevantmaturityofthesubordinate.

Shouldthechild’senvironmentsuddenlychangetoonewherehislife-relevantmaturityisinadequate(forexample,ifherunsintosevereacademictrouble),theparentmayhavetoreverttoastyleusedearlier.As parental (or managerial) supervision moves from structured to

communicatingtomonitoring,thedegreeofstructuregoverningthebehaviorofthechild(orthesubordinate)doesnotreallychange.Ateenagerknowsitisnotsafetocrossabusyinterstatehighwayonhisbicycle,andtheparentnolongerhas to tell himnot todo it.Structuremoves frombeingexternally imposed tobeinginternallygiven.Iftheparent(orsupervisor)impartedearlyontothechild(orsubordinate)the

rightwaytodothings(thecorrectoperationalvalues),laterthechildwouldbelikely to make decisions the way the parent would. In fact, commonality ofoperational values, priorities, and preferences—how an organization workstogether—isamustiftheprogressioninmanagerialstyleistooccur.Without that commonality, anorganization canbecome easily confused and

lose its sense of purpose. Accordingly, the responsibility for transmittingcommonvalues restssquarelywith thesupervisor.He is,afterall,accountablefortheoutputofthepeoplewhoreporttohim;then,too,withoutasharedsetofvaluesa supervisorcannoteffectivelydelegate.Anassociateofminewhohadalwaysdoneanoutstandingjobhiredajuniorpersontohandlesomeoldtasks,whilehehimself tookonsomenewones.Thesubordinatedidpoorwork.Myassociate’sreaction:“Hehastomakehisownmistakes.That’showhelearns!”Theproblemwiththisisthatthesubordinate’stuitionispaidbyhiscustomers.And that is absolutely wrong. The responsibility for teaching the subordinatemust be assumed by his supervisor, and not paid for by the customers of hisorganization,internalorexternal.

ManagementStyleandManagerialLeverage

As supervisors, we should try to raise the task-relevant maturity of oursubordinates as rapidly as possible for obvious pragmatic reasons. Theappropriatemanagementstyle foranemployeewithhighTRMtakes less timethan detailed, structured, supervision requires. Moreover, once operationalvaluesarelearnedandTRMishighenough,thesupervisorcandelegatetaskstothesubordinate,thusincreasinghismanagerialleverage.Finally,atthehighest

levels of TRM, the subordinate’s training is presumably complete, andmotivation is likely to come fromwithin, from self-actualization,which is themostpowerfulsourceofenergyandeffortamanagercanharness.As we’ve learned, a person’s TRM depends on a specific working

environment.Whenthatchanges,sowillhisTRM,aswillhissupervisor’smosteffectivemanagementstyle.Let’sconsideranarmyencampmentwherenothingeverhappens.Thesergeantincommandhascometoknoweachofhissoldiersverywell, andbyand largemaintainsan informal relationshipwith them.Theroutines are so well established that he rarely has to tell anyone what to do;appropriate to the highTRMof the group, the sergeant contents himselfwithmerely monitoring their activity. One day a jeepload of the enemy suddenlyappears, coming over the hill and shooting at the camp. Instantly the sergeantrevertstoastructured,task-orientedleadershipstyle,barkingordersateveryone,tellingeachofhissoldierswhattodo,when,andhow….Afterawhile,iftheseskirmishescontinueandthegroupkeepsonfightingfromthesameplaceforacoupleofmonths, thistoowilleventuallybecomeroutine.Withthat, theTRMof the group for the new task—fighting—will increase.The sergeant can thengraduallyeaseofftellingeverybodywhattodo.Put another way, a manager’s ability to operate in a style based on

communicationandmutualunderstandingdependson therebeingenough timeforit.Thoughmonitoringisonpaperamanager’smostproductiveapproach,wehavetoworkourwayuptoitintherealworld.Evenifweachieveit,ifthingssuddenlychangewehavetorevertquicklytothewhat-when-howmode.Thatmodeisonethatwedon’tthinkanenlightenedmanagershoulduse.Asa

result,weoftendon’ttakeitupuntilitistoolateandeventsoverwhelmus.Wemanagersmustlearntofightsuchprejudicesandregardanymanagementmodenotaseithergoodorbadbutratheraseffectiveornoteffective,giventheTRMof our subordinates within a specific working environment. This is whyresearchers cannot find the single bestway for amanager towork. It changesdaybydayandsometimeshourbyhour.

It’sNotEasytoBeaGoodManager

Deciding the TRM of your subordinates is not easy. Moreover, even if amanagerknowswhattheTRMis,hispersonalpreferencestendtooverridethelogicalandproperchoiceofmanagementstyle.Forinstance,evenifamanagerseesthathissubordinate’sTRMis“medium”(seethetableonthispage),inthe

realworld themanagerwill likelyopt foreither the“structured”or“minimal”style. In otherwords,wewant either to be fully immersed in theworkof oursubordinates, making their decisions, or to leave them completely alone, notwantingtobebothered.Anotherproblemhere is amanager’sperceptionofhimself.We tend to see

ourselvesmore as communicators and delegators thanwe really are, certainlymuchmorethandooursubordinates.Itestedthisconclusionbyaskingagroupofmanagers to assess themanagement style of their supervisors, and then byaskingthosesupervisorswhattheythoughttheirstylewas.Some90percentofthesupervisorssawtheirstyleasmorecommunicatingordelegating than theirsubordinates’view.Whataccountsforthelargediscrepancy?Itispartlybecausemanagers think of themselves as perfect delegators. But also, sometimes amanagerthrowsoutsuggestionstoasubordinatewhoreceivesthemasmarchingorders—furtheringthedifferenceinperceptions.Amanageronce toldme thathis supervisordefinitelypracticedaneffective

communicating stylewithhimbecause they skied anddrank together.Hewaswrong. There is a huge distinction between a social relationship and acommunicatingmanagementstyle,whichisacaringinvolvementintheworkofthesubordinate.Closerelationshipsoffthejobmayhelptocreateanequivalentrelationshiponthejob,buttheyshouldnotbeconfused.TwopeopleIknewhada supervisor-subordinate relationship. They spent one week each year bythemselves,fishinginaremotearea.Whenfishing,theynevertalkedaboutwork—it being tacitly understood that work was off conversational limits. Oddlyenough, their work relationship remained distant, their personal friendshiphavingnoeffectonit.This brings us to the age-old question of whether friendship between

supervisorandsubordinateisagoodthing.Somemanagersunhesitatinglyassertthat they never permit social relationships to develop with people they workwith.Infact,thereareplusesandminuseshere.Ifthesubordinateisapersonalfriend, the supervisor canmove intoa communicatingmanagement stylequiteeasily, but the what-when-how mode becomes harder to revert to whennecessary.It’sunpleasanttogiveorderstoafriend.I’veseenseveralinstanceswhereasupervisorhad tomakeasubordinate-friend toeadisciplinary line. Inonecase,afriendshipwasdestroyed;inanother,thesupervisor’sactionworkedoutbecausethesubordinatefelt,thankstothestrengthofthesocialrelationship,that the supervisor was looking out for his (the subordinate’s) professionalinterests.

Everyonemustdecideforhimselfwhatisprofessionalandappropriatehere.Atestmightbetoimagineyourselfdeliveringatoughperformancereviewtoyourfriend.Doyoucringeatthethought?Ifso,don’tmakefriendsatwork.Ifyourstomach remains unaffected, you are likely to be someone whose personalrelationshipswillstrengthenworkrelationships.

13PerformanceAppraisal:ManagerasJudgeandJury

WhyBother?

Why are performance reviews a part of the management system of mostorganizations?Andwhydowe review theperformanceofour subordinates? Iposed both questions to a group of middle managers and got the followingresponses:

toassessthesubordinate’sworktoimproveperformancetomotivatetoprovidefeedbacktoasubordinatetojustifyraisestorewardperformancetoprovidedisciplinetoprovideworkdirectiontoreinforcethecompanyculture

Next, I asked the group to imagine themselves to be a supervisor giving areviewtoasubordinate,andaskedthemwhat theirfeelingswere.Someof theanswers:

prideangeranxietydiscomfortguiltempathy/concernembarrassmentfrustration

Finally, I asked the same group to think back to some of the performancereviews theyhad receivedandaskedwhat, if anything,waswrongwith them.Theiranswerswerequickandmany:

reviewcommentstoogeneralmixedmessages(inconsistentwithratingordollarraise)noindicationofhowtoimprovenegativesavoidedsupervisordidn’tknowmyworkonlyrecentperformanceconsideredsurprises

This should tell you that giving performance reviews is a very complicatedanddifficultbusinessandthatwe,managers,don’tdoanespeciallygoodjobatit.Thefactisthatgivingsuchreviewsisthesinglemostimportantformoftask-

relevant feedback we as supervisors can provide. It is how we assess oursubordinates’levelofperformanceandhowwedeliverthatassessmenttothemindividually.Itisalsohowweallocatetherewards—promotions,dollars,stockoptions,orwhateverwemayuse.Aswesawearlier,thereviewwillinfluenceasubordinate’s performance—positively or negatively—for a long time, whichmakes the appraisal one of themanager’s highest-leverage activities. In short,the review is an extremely powerful mechanism, and it is little wonder thatopinionsandfeelingsaboutitarestronganddiverse.Butwhatisitsfundamentalpurpose?Thoughalloftheresponsesgiventomy

questionsarecorrect,thereisonethatismoreimportantthananyoftheothers:itistoimprovethesubordinate’sperformance.Thereviewisusuallydedicatedtotwo things: first, theskill levelof thesubordinate, todeterminewhat skillsaremissing and to find ways to remedy that lack; and second, to intensify thesubordinate’smotivation inordertogethimonahigherperformancecurveforthesameskilllevel(seetheillustrationonthispage).The reviewprocess also represents themost formal typeof institutionalized

leadership.Itistheonlytimeamanagerismandatedtoactasjudgeandjury:wemanagersarerequiredbytheorganizationthatemploysustomakeajudgmentregardingafellowworkerandthentodeliverthatjudgmenttohim,facetoface.A supervisor’s responsibility here is obviously very significant. What

preparationhavewehadtodothejobproperly?AbouttheonlythingIcanthink

of is that as subordinateswe’vebeenon the receivingend.But ingeneraloursociety values avoiding confrontation. Even the word “argument” is frownedupon,somethingIlearnedmanyyearsagowhenIfirstcametothiscountryfromHungary. InHungarian, theword “argument” is frequently used to describe adifference of opinion. When I began to learn English and used the word“argument,”Iwouldbecorrected,aspeoplewouldsay,“Ohno,youdon’tmean‘argument,’ youmean ‘debate,’ ” or “youmean ‘discussion.’ ”Among friendsand peers you are not supposed to discuss politics, religion, or anything thatmightpossiblyproduceadifferenceofopinionandaconflict.Football scores,gardening,andtheweatherareokay.Wearetaughtthatthewell-manneredskirtpotentiallyemotionalissues.Thepointis,deliveringagoodperformancereviewis really aunique act givenbothour cultural backgroundandourprofessionaltraining.Don’t think for a moment that performance reviews should be confined to

largeorganizations.Theyshouldbepartofmanagerialpracticeinorganizationsof any size and kind, from the insurance agent with two office assistants toadministratorsineducation,government,andnonprofitorganizations.Thelongandshortof it: ifperformancematters inyouroperation,performance reviewsareabsolutelynecessary.Two aspects of the review—assessing performance and delivering the

assessment—areequallydifficult.Let’slookateachinalittlemoredetail.

AssessingPerformance

Determining the performance of professional employees in a strictly objectivemanner is very difficult because there is obviously no cut-and-dried way tomeasureandcharacterizeaprofessionalemployee’sworkcompletely.Mostjobsinvolveactivitiesthatarenotreflectedbyoutputinthetimeperiodcoveredbythereview.Yetwehavetogivesuchactivitiesappropriateweightasweassessaperson’sperformance,eventhoughweknowwewon’tnecessarilybeobjective,since only output can be measured with true objectivity. Anybody whosupervisesprofessionals, therefore,walksa tightrope:heneeds tobeobjective,but must not be afraid of using his judgment, even though judgment is bydefinitionsubjective.Tomakeanassessment lessdifficult, a supervisor shouldclarify inhisown

mindinadvancewhatitisthatheexpectsfromasubordinateandthenattempttojudge whether he performed to expectations. The biggest problem with most

reviewsisthatwedon’tusuallydefinewhatitiswewantfromoursubordinates,and,asnotedearlier,ifwedon’tknowwhatwewant,wearesurelynotgoingtogetit.Let’s thinkback toourconceptof themanagerial“blackbox.”Using it,we

can characterize performance by output measures and internal measures. Thefirstrepresenttheoutputoftheblackbox,andincludesuchthingsascompletingdesigns,meetingsalesquotas,orincreasingtheyieldinaproductionprocess—thingswecanandshouldplotoncharts.Theinternalmeasurestakeintoaccountactivitiesthatgooninsidetheblackbox:whateverisbeingdonetocreateoutputfortheperiodunderreviewandalsothatwhichsetsthestagefortheoutputoffutureperiods.Arewereachingourcurrentproductiongoalsinsuchawaythattwomonths fromnowweare likely to face a groupof disgruntledproductionemployees? Are we positioning and developing people in the organization insuchawaythatourbusinesscanhandleitstasksinthefuture?Arewedoingallofthethingsthatadduptoawell-rundepartment?Thereisnostrictformulabywhichwecancomparetherelativesignificanceofoutputmeasuresandinternalmeasures. In agiven situation, theproperweighting couldbe50/50, 90/10, or10/90andcouldevenshift fromonemonthto thenext.Butat leastweshouldknowwhichtwovariablesarebeingtradedoffagainsteachother.A similar kind of trade-off also has to be considered here: weighing long-

term-orientedagainstshort-term-orientedperformance.Anengineerworkingonthe design of a product needs to complete the project on a strict schedule togeneraterevenue.Hemayalsobeworkingonadesignmethodthatwillmakeiteasierforotherstodesignsimilarproductsinthefuture.Theengineerobviouslyneedsbothactivitiesevaluatedandreviewed.Whichismoresignificant?Awaytohelpweighquestionslikethisistheideaof“presentvalue”usedinfinance:howmuchwillthefuture-orientedactivitypaybackovertime?Andhowmuchisthatworthtoday?There is also a time factor to consider.The subordinate’s output during the

reviewperiodmayhaveall,some,ornothingtodowithhisactivitiesduringthesameperiod.Accordingly,thesupervisorshouldlookatthetimeoffsetbetweentheactivityofthesubordinateandtheoutputthatresultsfromthatactivity.LetmeexplainwhatImean,becausethisisonelessonIlearnedthehardway.Theorganizationofoneofthemanagersreportingtomehadhadasuperbyear.Alloutputmeasureswereexcellent, sales increased,profitmarginsweregood, theproductsworked—youcouldhardlyeventhinkofgivinganythingbutasuperiorreview to the person in charge. Yet I had some misgivings. Turnover in his

groupwashigher thanitshouldhavebeen,andhispeopleweregrumblingtoomuch.Therewereothersuchstrawsinthewind,butwhocouldgivecredencetoelusivesignswhentangible,measurableperformancewassooutstanding?Sothemanagergotaverypositivereview.The next year his organization took a nose dive. Sales growth disappeared,

profitabilitydeclined,productdevelopmentwasdelayed,andtheturmoilamonghis subordinates deepened. As I prepared the next review of this manager, Istruggled to sort out what had happened. Did the manager’s performancedeteriorate as suddenly as his organization’s outputmeasures indicated?Whatwas going on? I concluded that in fact the manager’s performance wasimprovinginthesecondyear,evenasthingsseemedtogotohell.Theproblemwasthathisperformancehadnotbeengoodayearearlier.Theoutputindicatorsmerelyrepresentedworkdoneyearsago—thelightfromdistantstars,asitwere—whichwasstillholdingup.Thetimeoffsetbetweenthemanager’sworkandthe output of his organization was just about a year. Greatly embarrassed, IregretfullyconcludedthatthesuperiorratingIhadgivenhimwastotallywrong.Trusting the internalmeasures, Ishouldhavehad the judgmentandcourage togivethemanageramuchlowerratingthanIdidinspiteoftheexcellentoutputindicatorsthatdidnotreflecttheyearunderreview.The time offset between activity and output can also work the other way

around.IntheearlyyearsofIntel,Iwascalledupontoreviewtheperformanceofasubordinatewhowassettingupaproductionfacilityfromscratch.Ithadnotmanufactured anything as yet, but of course the review could not wait fortangibleoutput.Ihadhadnopriorexperiencesupervisingsomeonewhodidnothavea recordof concreteoutput.Here Igavemy subordinate credit fordoingwell,eventhoughoutputremaineduncertain.Asmanagers,wearereallycalledupontojudgeperformance,notjusttoseeandrecorditwhenit’sinplainsight.Finally,asyoureviewamanager,shouldyoubejudginghisperformanceor

theperformanceofthegroupunderhissupervision?Youshouldbedoingboth.Ultimatelywhatyouareafteristheperformanceofthegroup,butthemanageristheretoaddvalueinsomeway.Youneedtodeterminewhatthatis.Youmustask:Ishedoinganythingwithhisgroup?Ishehiringnewpeople?Ishetrainingthepeoplehehas,anddoingotherthingsthatarelikelytoimprovetheoutputoftheteaminthefuture?Themostdifficultissuesindeterminingaprofessional’sperformancewill be based on asking questions andmaking judgments of thissort.

Onebigpitfall tobeavoided is the“potential trap.”Atall timesyoushouldforceyourselftoassessperformance,notpotential.By“potential”Imeanformratherthansubstance.Iwasonceaskedtoapprovetheperformancereviewofageneralmanagerwhosesupervisor ratedhimhighly for theyear.Themanagerwasresponsibleforabusinessunitthatlostmoney,misseditsrevenueforecastmonthaftermonth,slippedengineeringschedules,and ingeneralshowedpooroutputandinternalmeasuresovertheyear.Accordingly,Icouldnotapprovethereview. Whereupon his supervisor said, “But he is an outstanding generalmanager.He is knowledgeable and handles himselfwell. It’s his organizationthatdidnotdowell,notthemanagerhimself!”Thiscutnoicewithmebecausetheperformanceratingofamanagercannotbehigher than theonewewouldaccordtohisorganization!Itisveryimportanttoassessactualperformance,notappearances; real output, not good form. Had themanager been given a highrating,Intelwouldhavesignaledtoallatthecompanythattodowell,youmust“act”likeagoodmanager,talklikeone,andemulateone—butyoudon’tneedtoperformlikeone.A decision to promote is often linked, as it should be, to the performance

review.Wemustrecognizethatnoactioncommunicatesamanager’svaluestoanorganizationmoreclearlyandloudlythanhischoiceofwhomhepromotes.Byelevatingsomeone,weare, ineffect,creatingrolemodels forothers inourorganization.Theoldsayinghasitthatwhenwepromoteourbestsalesmanandmakehimamanager,weruinagoodsalesmanandgetabadmanager.Butifwethink about it, we see we have no choice but to promote the good salesman.Should our worst salesman get the job? When we promote our best, we aresayingtooursubordinatesthatperformanceiswhatcounts.Itishardenoughforustoassessoursubordinates’performance,butwemust

alsotrytoimproveit.Nomatterhowwellasubordinatehasdonehisjob,wecanalways findways to suggest improvement, something aboutwhich amanagerneednotfeelembarrassed.Blessedwith20/20hindsight,wecancomparewhatthesubordinatedidagainstwhathemighthavedone,andthevariancecan tellbothofushowtodothingsbetterinthefuture.

DeliveringtheAssessment

TherearethreeL’stokeepinmindwhendeliveringareview:Level,listen,andleaveyourselfout.You must level with your subordinate—the credibility and integrity of the

entiresystemdependonyourbeingtotallyfrank.Anddon’tbesurprisedtofindthat praising someone in a straightforward fashion can be just as hard ascriticizinghimwithoutembarrassment.Theword“listen”hasspecialmeaninghere.Theaimofcommunicationisto

transmitthoughtsfromthebrainofpersonAtothebrainofpersonB.Thoughtsin the head ofA are first converted intowords,which are enunciated and viasoundwavesreachtheearofB;asnerveimpulsestheytraveltohisbrain,wheretheyaretransformedbackintothoughtsandpresumablykept.ShouldpersonAusea taperecorder toconfirmthewordsusedin thereview?Theanswer isanemphatic no. Words themselves are nothing but a means; getting the rightthought communicated is theend.PerhapsBhasbecomesoemotional thathecan’tunderstandsomethingthatwouldbeperfectlycleartoanyoneelse.PerhapsB has become so preoccupied trying to formulate answers that he can’t reallylistenandgetA’smessage.PerhapsBhastunedoutandasadefenseisthinkingofgoingfishing.Allofthesepossibilitiescananddooccur,andallthemoresowhenA’smessageisladenwithconflict.Howthencanyoubesureyouarebeingtrulyheard?Whattechniquescanyou

employ?Is itenough tohaveyoursubordinateparaphraseyourwords?Idon’tthinkso.Whatyoumustdoisemployallofyoursensorycapabilities.Tomakesure you’re being heard, you should watch the person you are talking to.Remember, themore complex the issue, themore prone communication is tobeing lost. Does your subordinate give appropriate responses towhat you aresaying? Does he allow himself to receive your message? If his responses—verbalandnonverbal—donotcompletelyassureyouthatwhatyou’vesaidhasgotten through, it isyourresponsibility tokeepat ituntilyouaresatisfied thatyouhavebeenheardandunderstood.This iswhat Imean by listening: employing your entire arsenal of sensory

capabilities tomakecertainyourpoints arebeingproperly interpretedbyyoursubordinate’s brain. All the intelligence and good faith used to prepare yourreviewwillproducenothingunlessthisoccurs.Yourtool,tosayitagain,istotallistening.Everygoodclassroomteacherworksinthesameway.Heknowswhenwhat

he is saying is being understoodby his students. If it isn’t, he takes heed andexplains thingsagainorexplains things inadifferentway.Allofushavehadprofessors who lectured by looking at the blackboard, mumbling to it, andcarefully avoiding direct eye contactwith the class. The reason: knowing that

theirpresentationwasmurkyandincomprehensible,theseteacherslookedawayfrom their audience to avoid confirming visually what they already knew. Sodon’timitateyourworstprofessorswhiledeliveringperformancereviews.Listenwith all yourmight tomake sure your subordinate is receiving yourmessage,anddon’tstopdeliveringituntilyouaresatisfiedthatheis.ThethirdLis“leaveyourselfout.”Itisveryimportantforyoutounderstand

that the performance review is about and for your subordinate. So your owninsecurities,anxieties,guilt,orwhatevershouldbekeptoutofit.Atissuearethesubordinate’sproblems,not the supervisor’s, and it is the subordinate’sday incourt.Anyonecalledupontoassesstheperformanceofanotherpersonislikelytohavestrongemotionsbeforeandduringthereview,justasactorshavestagefright.Youshouldworktocontroltheseemotionssothattheydon’taffectyourtask,thoughtheywillwellupnomatterhowmanyreviewsyou’vegiven.Letusnowconsiderthreetypesofperformancereviews.

“OntheOneHand…OntheOtherHand…”

Most reviews probably fall into this category, containing both positive andnegative assessments. Common problems here include superficiality, clichés,and laundry lists of unrelated observations. All of these will leave yoursubordinate bewildered and will hardly improve his future performance, thereview’sbasicpurpose.Letmesuggestsomewaystohelpyoudeliverthistypeofreview.The key is to recognize that your subordinate, likemost people, has only a

finite capacity to deal with facts, issues, and suggestions. You may possessseventruthsabouthisperformance,butifhiscapacityisonlyfour,atbestyou’llwasteyourbreathontheotherthree.Atworstyouwillhavelefthimwithacaseof sensory overload, and hewill go awaywithout getting anything out of thereview.Thefactisthatapersoncanonlyabsorbsomanymessagesatonetime,especiallywhentheydealwithhisownperformance.Thepurposeofthereviewisnottocleanseyoursystemofallthetruthsyoumayhaveobservedaboutyoursubordinate, but to improve his performance. So here less may very well bemore.Howcanyoutargetafewkeyareas?First,considerasmanyaspectsofyour

subordinate’s performance as possible. You should scan material such asprogress reports, performance against quarterly objectives, and one-on-onemeetingnotes.Thensitdownwithablankpieceofpaper.Asyouconsideryour

subordinate’sperformance,writeeverythingdownonthepaper.Donotedit inyourhead.Geteverythingdown,knowingthatdoingsodoesn’tcommityoutodo anything.Thingsmajor,minor, and trivial can be included in no particularorder. When you have run out of items, you can put all of your supportingdocumentationaway.Now,fromyourworksheet, lookforrelationshipsbetweenthevarious items

listed. You will probably begin to notice that certain items are differentmanifestationsofthesamephenomenon,andthattheremaybesomeindicationswhyacertainstrengthorweaknessexists.Whenyoufindsuchrelationships,youcan start calling them “messages” for the subordinate. At this point, yourworksheetmight looksomethinglikethatshownonthenextpage.Now,againfrom your worksheet, begin to draw conclusions and specific examples tosupport them. Once your list of messages has been compiled, ask yourself ifyoursubordinatewillbeabletorememberallofthemessagesyouhavechosentodeliver.Ifnot,youmustdeletethelessimportantones.Remember,whatyoucouldn’tincludeinthisreview,youcanprobablytakeupinthenextone.

POSITIVES NEGATIVES

—planningprocessmuchbetter!(quickstart)

—specprocess:zero!

—goodreporttoMaterialsCouncil —debatingsocietymeetings—allmushy

—helpedonPurchasingcostanalysisproject

—poorkick-offforspectraining

Messages

1.Goodresultsonplanningsystem(analytical/financialbackgrounduseful)2.Hardtimesettingclear,crispgoals—satisfiedwithactivitiesinsteadofdrivingresults!3. (No-let’sjustconcentrateon#2!)

Worksheetforperformancereview.

Let’s talk about surprises. If you have discharged your supervisoryresponsibilities adequately throughout the year, holding regular one-on-onemeetings and providing guidance when needed, there should never be anysurprises at a performance review, right? Wrong. When you are using theworksheet,sometimesyoucomeupwithamessagethatwillstartleyou.Sowhatdo you do?You’re facedwith either delivering themessage or not, but if thepurposeof thereviewis to improveyoursubordinate’sperformance,youmustdeliver it. Preferably, a review should not contain any surprises, but if youuncoverone,swallowhardandbringitup.On this page you’ll find an “on the one hand, on the other hand” kind of

performance review. It was written to correspond to the worksheet shownopposite.Ihaveannotatedittocallattentiontosomeofthethingswediscussedinthischapter.

TheBlast

With a little soul-searching, you may come to realize that you have a majorperformanceproblemonyourhands.Youhaveasubordinatewho,unlessturnedaround,couldgetfired.Todealwiththeproblem,youandyoursubordinatewilllikelygothroughstagescommonlyexperiencedinproblem-solvingofallkindsandparticularlyinconflictresolution.Theseareshownonthispage.You’llfindtheseoccurringdefinitelyduringandpossiblyafterthe“blast”review,whichisbasicallyanexerciseofresolvingconflictaboutabigperformanceproblem.A poor performer has a strong tendency to ignore his problem. Here a

manager needs facts and examples so that he can demonstrate its reality.Progressofsomesortismadewhenthesubordinateactivelydeniestheexistenceofaproblemratherthanignoringitpassively,asbefore.Evidencecanovercomeresistancehereaswell,andweenterthethirdstage,whenthesubordinateadmitsthatthereisaproblem,butmaintainsitisnothisproblem.Insteadhewillblameothers, a standard defensemechanism.Using this defense, he can continue toavoidtheresponsibilityandburdenofremedyingthesituation.Thesethreestepsusuallyfollowoneanotherinfairlyrapidsuccession.Butthingstendtogetstuckat theblame-others stage. Ifyour subordinatedoeshaveaproblem, there’snoway of resolving it if he continues to blame it on others. He has to take thebiggeststep:namelyassumingresponsibility.Hehastosaynotonlythatthereisaproblembutthatitishisproblem.Thisisfateful,becauseitmeanswork:“Ifitismyproblem,Ihavetodosomethingaboutit.IfIhavetodosomething,itis

likelytobeunpleasantandwilldefinitelymeanalotofworkonmypart.”Onceresponsibilityhasbeenassumed,however,findingthesolutionisrelativelyeasy.This is because the move from blaming others to assuming responsibilityconstitutes an emotional step,while themove from assuming responsibility tofindingthesolutionisanintellectualone,andthelatteriseasier.

Thestagesofproblem-solving:Thetransitionfromblamingotherstoassumingresponsibilityisanemotionalstep.

Itisthereviewer’sjobtogetthesubordinatetomovethroughallofthestagestothatofassumingresponsibility,thoughfindingthesolutionshouldbeasharedtask. The supervisor should keep track of what stage things are in. If thesupervisor wants to go on to find the solution when the subordinate is stilldenying or blaming others, nothing can happen. Knowingwhere you arewillhelpyoubothmovethroughthestagestogether.In the end, there are three possible outcomes. One, the subordinate accepts

your assessment and your recommended cure, and commits himself to take it.Two, hemay disagree completelywith your assessment but still accepts yourcure. Three, the subordinate disagrees with your assessment and does not

commithimself todowhatyou’ve recommended.As the supervisor,whichofthesethreeshouldyouconsideracceptableresolutionstotheproblem?Ifeelverystronglythatanyoutcomethatincludesacommitmenttoactionis

acceptable. Complex issues do not lend themselves easily to universalagreement.Ifyoursubordinatesayshe’scommittedtochangethings,youhaveto assume he’s sincere. The keyword here isacceptable. It is certainlymoredesirable for you and your subordinate to agree about the problem and thesolution, because that will make you feel that he will enthusiastically worktowardremedyingit.Souptoapointyoushouldtrytogetyoursubordinatetoagreewithyou.But ifyoucan’t,accepthiscommitment tochangeandgoon.Don’tconfuseemotionalcomfortwithoperationalneed.Tomake thingswork,peopledonotneedtosidewithyou;youonlyneedthemtocommitthemselvesto pursue a course of action that has been decided upon. There seems to besomething not quite nice about expecting a person towalk down a path he’drather not be on.But on the jobwe are after a person’s performance, not ourpsychologicalcomfort.IlearnedthedistinctionbetweenthetwoduringoneofthefirstreviewsIhad

togive.Iwastryingveryhardtopersuademysubordinatetoseethingsmyway.Hesimplywouldnotgoalongwithmeandfinallysaidtome,“Andy,youwillnever convince me, but why do you insist on wanting to convince me? I’vealreadysaidIwilldowhatyousay.”Ishutup,embarrassed,notknowingwhy.IttookmealongtimebeforeIrealizedIwasembarrassedbecausemyinsistencehadalottodowithmakingmefeelbetterandlittletodowiththerunningofthebusiness.If it becomes clear that you are not going to get your subordinate past the

blame-others stage,youwill have to assume the formal roleof the supervisor,endowed with position power, and say, “This is what I, as your boss, aminstructingyoutodo.Iunderstandthatyoudonotseeitmyway.Youmayberight or I may be right. But I am not only empowered, I am required by theorganizationforwhichwebothworktogiveyouinstructions,andthisiswhatIwantyoutodo…”Andproceedtosecureyoursubordinate’scommitmenttothecourseof actionyouwant and thereaftermonitor hisperformance against thatcommitment.RecentlyoneofmysubordinateswroteareviewthatIconsideredsuperficial,

lackinganalysisanddepth.Mysubordinate,aftersomediscussion,agreedwithmyassessment,butheconsideredtheissuenotimportantenough,asheputit,to

spend time rewriting the review. Aftermore spirited discussion, we remaineddeadlocked.Finally, I tookadeepbreathandsaid tohim,“Look, Iunderstandthatyoudon’tconsider itworthyour time todo it.But Iwantyou todo it.” Iadded that“Iguess there isabasicdifferencebetweenus.The integrityof theperformancereviewsystemisjustmoreimportanttomethanitistoyou.Thatiswhy I have to insist.”He lookedback atme and after amoment simply said,“Okay.”HethoughtIwasoutinleftfieldandresentedthefactthatImadehimspendtimeonsomethinghethoughtwasunimportant,buthecommittedhimselftoredothereview,and,infact,hediditwell.Hissubordinateendedupgettingthereworked,muchmorethoroughandthoughtfulreview,andthefactthathisreview was rewritten without the agreement of my subordinate made nodifferencetohim.

ReviewingtheAce

Aftertryingtoestablishtheprinciplesofperformanceappraisalwithagroupofabout twentymiddle managers, I asked them to take a review they had oncereceived and to analyze it according to our new criteria. The resultswere notwhatIexpected,butIdidlearnfromthem.This group consisted of achievers, and their ratingsweremostly very high.

The reviewswere exceptionallywell written,much better than the average atIntel. However, for content, they tended to be retrospective assessments,analysesofwhatthesubordinatehaddoneinthecourseoftheprioryear.Eventhoughtheirkeypurposewastoimprovethesubordinate’sfutureperformance,amajorityofthereviewsmadelittleornoattempttodefinewhatthesubordinateneededtodotoimprovehisperformanceoreventomaintainhiscurrentlevel.Itseems that for an achiever the supervisor’s effort goes into determining andjustifyingthejudgmentofthesuperiorperformance,whilegivinglittleattentiontohowhecoulddoevenbetter.Butforapoorperformer,thesupervisortendstoconcentrate heavily on ways he can improve performance, providing detailedandelaborate“correctiveactionprograms,”step-by-stepaffairsmeanttoensurethatthemarginalemployeecanpullhimselfuptomeetminimumrequirements.Ithinkwehaveourprioritiesreversed.Shouldn’twespendmoretimetrying

to improve the performanceof our stars?After all, these people account for adisproportionatelylargeshareoftheworkinanyorganization.Putanotherway,concentrating on the stars is a high-leverage activity: if they get better, theimpactongroupoutputisverygreatindeed.

Weallhaveahardtimesayingthingsthatarecritical,whetherwe’retalkingtoasuperioremployeeoramarginalone.Wemustkeepinmind,however,thatnomatterhowstellaraperson’sperformancelevelis, thereisalwaysroomforimprovement.Don’thesitatetousethe20/20hindsightprovidedbythereviewtoshowanyone,evenanace,howhemighthavedonebetter.

OtherThoughtsandPractices

Is it a good idea to ask the subordinate to prepare some kind of a self-reviewbeforebeingreviewedbyhissupervisor?Letmeanswerthequestionthisway.Yourown review isobviously important toyou, andyou reallywant toknowhowyoursupervisorseesyouryear’swork.Ifyouprepareareviewandgiveitto your supervisor, and he simply changes the format, retypes it, gives you asuperior rating, and then hands it back to you, how will you feel? Probablycheated. If you have to tell your supervisor about your accomplishments, heobviously doesn’t pay much attention to what you are doing. Reviewing theperformanceofsubordinatesisaformalactofleadership.Ifsupervisorspermitthemselves to be prompted in one way or another, their leadership and theircapacity for it will begin to appear false. So the integrity of the supervisors’judgmentheremustbepreservedatallcosts,andtheymustcommitthemselvesthrough an up-front judgment of their subordinates’ performance if the healthandvitalityofthereviewprocessaretobemaintained.What about asking your subordinate to evaluate your performance as his

supervisor? I think thismightbeagood idea.Butyoushouldmake it clear toyour subordinate that it’s your job to assess his performance, while hisassessmentofyouhasonlyadvisorystatus.Thepointis,heisnotyourleader;youarehis.Andundernocircumstancesshouldyoupretendthatyouandyoursubordinatesareequalduringperformancereviews.Shouldyoudeliverthewrittenreviewbefore,during,oraftertheface-to-face

discussion? I have tried it all threeways.Let’s consider someof thepros andcons of each.What happens if you have the review first and then give yoursubordinatewhatyou’vewrittenlater?Uponreadingit,thesubordinatemayfindaphrasethathedidn’t“hear”earlierandblowupoverit.Whataboutdeliveringthewrittenreviewduringthediscussion?Onemanagertoldmethathegivesthesubordinate a copy of the review and tells him to read the first severalparagraphs,which they then discuss.Grouping the paragraphs, supervisor andsubordinateworktheirwaythroughtheappraisal.Icanseeaproblemwiththis.

Howcanasupervisoraskasubordinatetostopatparagraphthreewhenheissoeagertoreadtherestofwhathe’sgot?Anothermanagertoldmethathereadsthewrittenreviewtohissubordinatetotrytocontrolthesession.Buthere,too,the subordinate is left eager to know what comes next and might not payattention towhat is really being said.Also,when your subordinate is given awrittenreviewduringthediscussion,hewon’thavethetimetothinkaboutwhatitsaysandislikelytowalkawaymutteringtohimself,“Ishouldhavesaidthisin response, and I should have said that.” For a goodmeeting ofminds, yoursubordinateshouldhavetimetoworkouthisreactionstowhat’sinthereview.Inmyexperience,thebestthingtodoistogiveyoursubordinatethewritten

reviewsometimebeforetheface-to-facediscussion.Hecanthenreadthewholething privately and digest it. He can react or overreact and then look at the“messages” again. By the time the two of you get together, he will bemuchmoreprepared,bothemotionallyandrationally.Preparinganddeliveringaperformanceassessmentisoneofthehardesttasks

you’llhavetoperformasamanager.Thebestwaytolearnhowtodooneistothinkcriticallyaboutthereviewsyouyourselfhavereceived.Andifyou’vebeenlucky, the tradition of good performance reviews has been handed fromsupervisor to subordinate, which has helped to maintain the integrity of thesystem in your company. Nevertheless, people constantly need to be proddedintodoingagoodjobofreviewing.EachyearIreadsomethinglikeahundredevaluations,allofthosewrittenbymyownsubordinatesandarandomselectionfrom throughout Intel. I commenton themandsend themback for rewritesorwithacomplimentarynote.IdothiswithasmuchnoiseandvisibilityasIcan,because I want to reiterate and reaffirm the significance the system has andshouldhave for every Intel employee.Anything lesswouldnotbe appropriatefor the most important kind of task-relevant feedback we can give oursubordinates.

14TwoDifficultTasks

There are twoother emotionally charged tasks amanagermust perform.Theyareinterviewingapotentialemployeeandtryingtotalkavaluedemployeeoutofquitting.

Interviewing

Thepurposeoftheinterviewisto:

•selectagoodperformer•educatehimastowhoyouandthecompanyare•determineifamutualmatchexists•sellhimonthejob

Themeansatyourdisposal typicallyconsistof anhouror twoof interviewtimeandacheckofthecandidate’sreferences.Weknowhowharditistoassesstheactualpastperformanceofourownsubordinateseventhoughwespentmuchtimeworkingcloselywiththem.Herewesitsomebodydownandtrytofindoutinanhourhowwellhe is likely toperforminanentirelynewenvironment. Ifperformanceappraisalisdifficult,interviewingisjustaboutimpossible.Thefactis,wemanagershavenochoicebuttoperformtheinterview,nomatterhowharditis.Butwemustrealizethattherisksoffailurearehigh.Theothertoolwehaveforassessingpotentialperformanceistoresearchpast

performance by checking references. But you’ll often be talking to a totalstranger,soevenifhecommentsfreelyaboutthecandidate,whathesayswon’thavemuchmeaningtoyouwithoutsomeknowledgeofhowhiscompanydoesbusinessandwhatvaluesitworksby.Moreover,whilefewreferenceswillout-and-out lie, they tend not to volunteer specific critical remarks. So reference-checking hardly exempts you from getting as much as you can out of theinterview.

CONDUCTINGTHEINTERVIEW

Theapplicantshoulddo80percentofthetalkingduringtheinterview,andwhathetalksaboutshouldbeyourmainconcern.Butyouhaveagreatdealofcontrolherebybeinganactive listener.Keep inmindyouonlyhaveanhourorso tolisten.Whenyouaskaquestion,agarrulousornervouspersonmightgoonandonwithhisanswerlongafteryou’velostinterest.Mostofuswillsitandlistenuntiltheendoutofcourtesy.Instead,youshouldinterruptandstophim,becauseifyoudon’t,youarewastingyouronlyasset—theinterviewtime,inwhichyouhavetogetasmuchinformationandinsightaspossible.Sowhenthingsgooffthetrack,getthembackonquickly.Apologizeifyoulike,andsay,“IwouldliketochangethesubjecttoX,Y,orZ.”Theinterviewisyourstocontrol,andifyoudon’t,youhaveonlyyourselftoblame.An interview produces the most insight if you steer the discussion toward

subjects familiar to both you and the candidate. The person should talk abouthimself,his experience,whathehasdoneandwhy,whathewouldhavedonedifferentlyifhehadittodoover,andsoforth,butthisshouldbedoneintermsfamiliartoyou,sothatyoucanevaluateitssignificance.Inshort,makesurethewordsusedmeanthesamethingtobothofyou.Whatarethesubjectsthatyoushouldbringupduringaninterview?Agroup

ofmanagersprovidedmewithwhattheythoughtwerethebestquestions.Theywere:

— Describe someprojects thatwerehighly regardedbyyourmanagement,especiallybythelevelsaboveyourimmediatesupervisor.

—Whatareyourweaknesses?Howareyouworkingtoeliminatethem?—Convincemewhymycompanyshouldhireyou.— What are some of the problems you are encountering in your current

position?Howareyougoing about solving them?What couldyouhavedonetopreventthemfromcroppingup?

—Whydoyouthinkyou’rereadyforthisnewjob?— What do you consider your most significant achievements?Why were

theyimportanttoyou?—Whatdoyouconsideryourmostsignificantfailures?Whatdidyoulearn

fromthem?—Whydoyouthinkanengineershouldbechosenforamarketingposition?

(Varythisoneaccordingtothesituation.)— Whatwas themost important course or project you completed in your

collegecareer?Whywasitsoimportant?

The information tobegainedhere tends to fall into fourdistinct categories.First,you’reafteranunderstandingofthecandidate’stechnicalknowledge:notengineering or scientific knowledge, butwhat he knows about performing thejob he wants—his skill level. For an accountant, technical skill means anunderstandingofaccounting;forataxlawyer,taxlaws;foranactuary,statisticsand theuseofactuarial tables;andsoon.Second,you’re trying toassesshowthispersonperformedinanearlierjobusinghisskillsandtechnicalknowledge;inshort,not justwhat thecandidateknows,butalsowhathedidwithwhatheknows. Third, you are after the reasons why there may be any discrepancybetween what he knew and what he did, between his capabilities and hisperformance.Andfinally,youaretryingtogetafeelforhissetofoperationalvalues,thosethatwouldguidehimonthejob.Let’slookathowthequestionsabovefitintothefourcategories.

Technical/Skillsdescribesomeprojectswhatareyourweaknesses

WhatHeDidWithKnowledgepastachievementspastfailures

Discrepancieswhatdidyoulearnfromfailuresproblemsincurrentposition

OperationalValueswhyareyoureadyfornewjobwhyshouldmycompanyhireyouwhyshouldengineerbechosenformarketingmostimportantcollegecourse/project

The ultimate purpose of interviewing is tomake a judgment about how thecandidatewouldperforminyourcompany’senvironment.Thisisatoddswithaprinciple we stressed about performance reviews: the need to avoid the“potential”trap.Butwhenyou’rehiring,youmustjudgepotentialcontribution.

Withinthehourorsoatyourdisposal,youmustmovebetweentheworldofthepastemployerandyourown,andprojectthecandidate’sfutureperformanceinanew environment based on his own description of past performance. Thismanagerialtaskisclearlytrickyandhigh-risk,butunfortunatelyunavoidable.Youcan’tgetawayfromrelyingonacandidate’sself-assessment.Butthat’s

notabadwaytogetdirectanswerstodirectquestions.If,forexample,youwereto ask, “Howgood are you technically?” the intervieweemight be taken backmomentarilybutthenclearhisthroatandsaytimidly,“Well,IthinkI’mprettygood…”Asyoulisten,you’llprobablygetadecentfixonhowcapablehereallyis.Don’tworryaboutbeingblunt;directquestionstendtobringdirectanswers,andwhentheydon’t,theyproduceotherformsofinsightintothecandidate.Askingacandidatetohandleahypotheticalsituationcanalsoenlightenyou.I

onceinterviewedsomeoneforthepositionofcostaccountantatIntel.HehadaHarvardMBAandcamefromthefoodserviceindustry.Heknewnothingaboutthe semiconductor business and I knew nothing about finance, so we reallycouldn’ttalkinmuchdetailabouthistechnicalabilitytodothejob.Idecidedtotakehim through the semiconductormanufacturingprocess stepby step.Aftersaying I would answer any specific questions he had, I asked him what thefinished cost of a wafer would be. He asked some questions and ponderedmatters for a while. He then started to think his way through the basicsemiconductorcostaccountingprinciples,discoveringsomeofthemashewentalong, andultimatelycameupwith thecorrect answer.Hewashired,becausethis exercise demonstrated (as it turns out, correctly) that his problem-solvingcapacitywasfirst-rate.Anotherapproachfollowsthatyoumaywanttousewhileinterviewing.The

candidate can tell you agreat deal about his capabilities, skills, andvaluesbyaskingyouquestions.Askthecandidatewhathewouldliketoknowaboutyou,the company, or the job. The questions he askswill tell youwhat he alreadyknows about the company,what hewould like to knowmore about, and howwell prepared he is for the interview. There’s nothing foolproof about this,however. Once a prospectivemanager came tomy office with a copy of ourannualreport,whichhehadreadverycarefullyandmarkedupwithpenetratingquestions. In fact, I couldn’t answermanyof them. Iwasvery impressed.Wehiredhimandhefailedbadlyon the job.AsIsaid, interviewingisahigh-riskproposition…A final point about references:when you are talking to them, you’re really

after the same information that you tried to get directly from the candidate. Ifyou know the reference personally, you have amuch better chance of getting“real”information.Ifyoudon’t,trytokeephimonthephonelongenoughtoletsome sort of personal bond develop. If you can uncover some commonexperienceor association, the referencewill probablybecomemoreopenwithyou.Inmyexperience,thelasttenminutesofahalf-hourconversationaremuchmorevaluablethanthefirsttenminutes,thankstothatbond.Ifpossible,youshould talkwith theapplicantagainafteryouhavechecked

his references, because you may have gotten some new perspectives. Such afollow-upinterviewcanbequiteafocusedaffair.What about “tricks”? The best ones I’ve heard about come to me from

somebody who had tried to get into the Navy’s nuclear submarine program.Admiral Rickover apparently personally interviewed each candidate andemployedtechniqueslikehavingthecandidatesitonathree-leggedchair.Whenit tipped over, the poor man would be left sprawling on the floor. Rickoverevidently thought the trick tested strength of character in the face ofembarrassment.ButIthinktheinterviewshouldbecompletelystraightforward.Remember, a candidate is apotential employee.Hewillgoaway fromhavingtalkedtoyouwithastrongsetoffirstimpressions.Ifthosearewrongandyouhiretheperson,itwilltakealongtimebeforetheychange.Soshowyourselfandyourenvironmentastheyreallyare.Arethereanyguaranteesofsuccess?SeveralyearsagoIinterviewedaperson

forahigh-levelpositionatIntel.IdidtheworkascarefullyandthoroughlyasIcould. I had a very good feel, I thought, of the whys and wherefores of theperson’sskills,pastperformance,andvalues,andwehiredhim.Fromdayonehe was a disaster. Much humbled, I’ve since gone over my notes from theinterviews and the conversationswith references.To this day I haven’t a clueaboutwhyIdidn’tspotthecandidate’sconsiderableflaws.Sointheendcarefulinterviewing doesn’t guarantee you anything, itmerely increases your odds ofgettinglucky.

“IQuit!”

This is what I most dread as a manager: a subordinate, highly valued andesteemed, decides to quit. I am talking not about someonewhosemotives aremoremoneyandbetterperksatanothercompany,butaboutanemployeewhoisdedicatedandloyalyetfeelshisworkisnotappreciated.Youandthecompany

don’twanttolosehim,andhisdecisiontoleavereflectsonyou.Ifhefeelshiseffortshavegoneunrecognized,youhavenotdoneyourjobandhavefailedashismanager.The opening shot usually occurswhen you are on the run.On yourway to

whatyouconsideranimportantmeeting,yoursubordinatetimidlystopsyouandmuttersunderhisbreath,“Doyouhaveaminute?”Hethenmuttersfurtherthathehasdecided to leave thecompany.You lookathimwide-eyed.Your initialreaction to his announcement is absolutely crucial. If you’re human, you’llprobably want to escape to your meeting, and you mumble something backabout talking things over later. But in almost all such cases, the employee isquittingbecausehefeelsheisnotimportanttoyou.Ifyoudonotdealwiththesituationrightatthefirstmention,you’llconfirmhisfeelingsandtheoutcomeisinevitable.Dropwhatyouaredoing.Sithimdownandaskhimwhyhe isquitting.Let

himtalk—don’targueaboutanythingwithhim.Believeme,he’srehearsedhisspeechcountlesstimesduringmorethanonesleeplessnight.Afterhe’sfinishedgoingthroughallhisreasonsforwantingtoleave(theywon’tbegoodones),askhim more questions. Make him talk, because after the prepared points aredelivered, the real issuesmay comeout.Don’t argue, don’t lecture, and don’tpanic.Remember,thisisonlytheopeningskirmish,notthewar.Andyoucannotwinthewarhere—butyoucanloseit!Youhavetoconveytohimbywhatyoudothatheisimportanttoyou,andyouhavetofindoutwhatisreallytroublinghim.Don’ttrytochangehismindatthispoint,butbuytime.Afterhe’ssaidallhehastosay,askforwhatevertimeyoufeelisnecessarytoprepareyourselfforthe next round. But know that you must follow through on whatever you’vecommittedyourselftodo.What’syournextmove?Becauseyouhaveamajorproblem,yougotoyour

supervisorforhelpandadvice.Henodoubtisalsoonhiswaytoanimportantmeeting…He,likeyou,willtrytoputthingsoff,andmostprobablynotbecausehedoesn’tcare,butbecausethesituationaffectsyoumorethanyoursupervisor—afterall,itisyoursubordinatewhohasdecidedtoquit.Itisuptoyoutomakeit your supervisor’s problem andmake him participate in the solution to yourproblem.Corporate citizenshipwill probably play a substantial role in what happens

next.Yoursubordinate isavaluedemployee—ofthecompany.Younowmustvigorouslypursueeveryavenueavailabletoyoutokeephimwiththefirm,even

if itmeanstransferringhimtoanotherdepartment.If itseemsthat is thelikelysolution,youmustbecometheprojectmanagerofthatsolutionuntilthewholething is settled. You may ask why you should put yourself out to keep anemployeewhomyouaregoingtolose.Thereisabasicprincipleinvolved:youoweittoyouremployertosaveanemployeeforthecompany.Beyondthis,thegoldenrulecanbecomemorethananiceidealinsituationsofthissort.Todayyou save a valued contributor for the company by virtually giving him to afellowmanager.Tomorrowitwillbehis turn todoyouthesamefavor. In thelongrun,ifallmanagerstakethisposition,theywillallwin.Nowyoumaybe ready togoback toyour subordinatewitha solution,one

that addresses his real reasons for wanting to quit and one that in turn willbenefitthecompany.Bynowheshouldknowthatheisimportanttoyou,buthemightsaythatyoushouldhaveofferedhimthenewpositionlongago.Hemightgo on to say that you’re only doing it now because he forced you into it, hisfeelingbeingthat“IfIstay,you’llthinkofmeastheblackmailerforever!”Younowhavetomakehimfeelcomfortablewiththenewarrangement.You

might say something like, “You did not blackmail us into doing anythingweshouldn’thavedoneanyway.Whenyoualmostquit,youshookusupandmadeusawareoftheerrorofourways.Wearejustdoingwhatweshouldhavedonewithoutanyofthishappening.”Thenyoursubordinatemaysayhe’sacceptedajobsomewhereelseandcan’t

back out. You have to make him quit again. You say he’s really made twocommitments:first toapotentialemployerheonlyvaguelyknows,andsecondtoyou,hispresentemployer.Andcommitmentshehasmade to thepeoplehehas been working with daily are far stronger than onemade to a casual newacquaintance.As I said, the whole thing is not easy, either for the subordinate or the

supervisor. But you must give it your best shot, because the good of thecompany is involved and the issue is even more important than keeping onevalued employee. This subordinate is valuable and important because he hasattributes thatmakehimso.Otheremployees respecthim;and if theyare likehim, they identify with him. So other superior performers like himwill trackwhat happens to him, and their morale and commitment to the companywillhingeontheoutcomeofthisperson’sfate.

15CompensationasTask-RelevantFeedback

MoneyhassignificanceatalllevelsofMaslow’smotivationhierarchy.Asnotedearlier, a person needs money to buy food, housing, and insurance policies,whicharepartofhisphysiologicalandsafety/securityneeds.Asonemovesuptheneedhierarchy,moneybeginstomeansomethingelse—ameasureofone’sworthinacompetitiveenvironment.EarlierIdescribedasimpletestthatcanbeappliedtodeterminetherolemoneyplaysforsomeone.Iftheabsoluteamountof a raise in salary is important, that person is probably motivated byphysiologicalorsafety/securityneeds.Iftherelativeamountofaraise—whathegot compared to others—is the important issue, that person is likely to bemotivated by self-actualization, because money here is a measure, not anecessity.Athigherlevelsofcompensation,anincrementalamountofmoneygradually

will have less and less material utility to the person who gets it. In myexperience,middlemanagersareusuallypaidwellenoughthatmoneydoesnothavecrucialmaterialsignificancetothem,butnotwellenoughthatitiswithoutany material significance. Of course, one middle manager’s needs can differgreatly from another’s, depending on individual circumstances—number ofchildren, aworking spouseornot, and soon.Asa supervisor,youhave tobevery sensitive toward thevariousmoneyneedsofyour subordinatesand showempathy toward them.Youmustbeespeciallycarefulnot toprojectyourowncircumstancesontoothers.As managers, our concern is to get a high level of performance from our

subordinates. So we want to dispense, allocate, and use money as a way todeliver task-relevant feedback. To do this, compensation should obviously betied to performance, but that, as we’ve seen, is very hard to assess precisely.Because amiddlemanager cannot be paid by the piece, his job can never bedefined by simple output. And because his performance is woven into theperformanceofateam,itishardtodesignacompensationschemetieddirectlytotheindividualperformanceofamiddlemanager.

Butcompromisescanbesetup.Wecanbaseaportionofamiddlemanager’scompensation on his performance. Let’s call this a performance bonus. Thepercentage thebonus representsof amanager’s total compensation should risewithhistotalcompensation.Thus,forahighlypaidseniormanager,forwhomthe absolute dollars make relatively little difference, the performance bonusshouldbeashighas50percent,whileamiddlemanagershouldreceivemoreinthe rangeof10 to25percentofhis totalcompensation thisway.Even thoughwhathemakesistypicallyatalevelwheresubstantialfluctuationscouldcausepersonalhardship,wecanatleastgiveatasteoftask-relevantfeedback.Todesignagoodperformancebonusscheme,wemustdealwithavarietyof

issues.Weneed to figureout if theperformance is linked to a teamor if it ismostlyrelatedtoindividualwork.Ifitistheformer,whomakesuptheteam?Isitaprojectteam,adivision,ortheentirecorporation?Wealsoneedtofigureoutwhatperiodtheperformancebonusshouldcover,realizingagainthatcauseandeffecttendtobeoffsetfromeachother,oftenbyalongtime,butabonusneedstobepaidcloseenoughtothetimetheworkwasdonethatthesubordinatecanrememberwhy itwasawarded.Furthermore,wemust thinkaboutwhether thebonus should be based strictly on countable items (financial performance, forexample),onachievingmeasurableobjectives,oronsomesubjectiveelementsthatmightgetusdrawnintoabeautycontest.Finally,ofcourse,wedon’twantto devise something that pays out lavishly even as the company is goingbankrupt.If you take all of this into account, you are likely to come up with some

complex arrangements. For example, you might have a scheme in which amanager’sperformancebonusisbasedonthreefactors.Thefirstwouldincludehisindividualperformanceonly,asjudgedbyhissupervisor.Thesecondwouldaccount for his immediate team’s objective performance, his departmentperhaps.Thethirdfactorwouldbelinkedtotheoverallfinancialperformanceofthe corporation. When you take, let’s say, 20 percent of a manager’scompensationandsplititintothreeparts,anyonewillhaveonlyasmallimpacton total compensation, yet attentionwill still be called to its significance.Nomatterwhatwayyouchoosetodeterminebonuses,nonegivesyouexactlywhatyouwant,butmostofthemwillspotlightperformanceanddelivertask-relevantfeedback.Let’snowlookattheadministrationofbasesalaries.Intheabstract,thereare

twowaystodoit.Atoneextreme,thedollarlevelisdeterminedbyexperienceonly; at the other, by merit alone. In the experience-only approach, an

employee’ssalaryincreaseswiththetimehehasspentinaparticularposition.Akey point here is that any job has amaximum value; nomatter how long anindividualhasbeenin it,hissalaryultimatelyhas to leveloff,asshownin thefigureonthenextpage.Inthemerit-onlyapproach,salaryisindependentofthetimespentinthejob.Herethetheorysays,“Idon’tcareifyouareoneyearoutofcollegeorhavespenttwentyyearsintheworkforce.Ionlycaretoseehowyou perform in this job.” But here too, of course, a given job still has amaximumvalue.Socialnormscanforceusintosomeunfortunatecompensationpractices. For instance, even though we say that every job has a finite valuewhere compensation should level off, we often let an individual become toohighlypaidbecausewe,management,keepgivingroutineraises.

Therearetwopureformsofsalaryadministration;mostcompaniesuseacompromise.

Many organizations practice a pure experience-only form of salaryadministration.LargeJapanesecompaniestendtoplacenodistinctionbasedonperformanceduringthefirsttenorsoyearsofemployment—whichareprobablythemost productive years of a professional’s life. Likewise, unions andmostgovernment jobs lean toward pure experience-only salary scales. Apart fromwhether this is fair or not, themessage frommanagement is that performancedoesn’tmattermuch. Consider teachers inmany school systems. A good onegets paid the same salary as a bad one if they both have been around for thesame length of time. How a teacher is evaluated is not usually tied even

symbolicallytocompensation,whichmakesmewonderifthepass/failsystemofgradingdidnothaveitsorigininthewaythetypicalteacherispaid.At the same time,merit-only salary administration is impractical in its pure

form. It is very hard to ignore a person’s experience as you try to pay a fairsalary.Thus,mostcompanieschooseacoursebetweenthetwoextremes,whichisacompromiseschemethattakestheshapeofafamilyofcurvesshownintheprevious figure.The shapes of all of themapproximate the curve representingtheexperience-onlyapproach,butasyoucansee,whilepeoplestartatthesamesalary level, they move up at different speeds and arrive at different places,dependinguponindividualperformance.Of the three schemes, the one based on experience only is obviously the

easiesttoadminister.Ifyoursubordinatedoesnotliketheraisehe’sbeengiven,allyouhavetodoisshowhimthebookwhereitsaysthatforXamountoftimeon the job he deserves and gets Y amount of pay. A supervisor trying toadministersometypeofmerit-basedorcompromiseschemehastodealwiththeallocationofafiniteresource—money—andthisrequiresthoughtandeffort.Ifwewant to use such schemes,wehave to come to termswith the principle—troubling to many managers—that any merit-based system requires acompetitive,comparativeevaluationofindividuals.Merit-based compensation simply cannotwork unlesswe understand that if

someone is going tobe first, somebodyelsehas tobe last.AsAmericans,wehave no problem accepting a competitive ranking in a sports event. Even thepersonwhocomesinlastinaracefeelscomfortableaboutthesystemthatsayssomeone has to finish last. But at work, unfortunately, competitive rankingfrequentlybecomesahighlychargedissue,difficulttoacceptandtoadminister—yetitisamustifwewanttousesalaryasawaytoencourageperformance.Promotions, defined as a substantial change in a person’s job, are very

importanttothehealthofanyorganizationandshouldbeconsideredwithgreatcare.Obviously, for the individual concerned, promotions often produce a bigraise.Aswehaveseen,promotionsarealso readily seenbyothermembersofthe organization, and so take on a vitally important role in communicating avalue system to the rest of the company. Promotions must be based onperformance, because that is the only way to keep the idea of performancehighlighted,maintained,andperpetuated.If we are going to consider promotions, we have to consider the Peter

Principle,whichsaysthatwhensomeoneisgoodathisjob,heispromoted;he

keepsgettingpromoteduntilhereacheshislevelofincompetenceandthenstaysthere.Like all good caricatures, this one captures at least someofwhat reallyhappensinamerit-basedpromotionsystem.Take a look at the illustration opposite, where we track someone’s

promotions. At point A the demands of Job 1 so tax him that he can onlyperform in an average fashion. In the jargon of performance assessment, he“meets therequirements”of the job.As timepasses,hereceivesmore trainingand becomes more motivated, and improves his performance to an above-average level, or, again in the jargon, to a point where he “exceeds therequirements”of theposition.At this timeweconsider thepersonpromotable,and in fact do promote him to Job 2,where hewill at first performonly at a“meets requirements” level.Withmore experience, he again will “exceed therequirements”of the job.Eventually,heprobablygetspromotedagainand thecycle repeats itself. Thus, an achiever will alternate between the “meetsrequirements” and the “exceeds requirements” ratings throughout his career,untilheeventuallysettlesata“meetsrequirements”level,atwhichtimehewillno longer bepromoted.This, perhaps, is a better descriptionof how thePeterPrincipleworks.

Anachieverwillalternatebetween“meetsrequirements”and“exceedsrequirements”ratingsthroughouthiscareer.

Now,isthereanalternativetothis?Isaythereisnot.IfwetakeapersonatpointBanddon’tofferhimmoreworkandgreaterchallengeseven thoughhe“exceedstherequirements”ofJob1,wearenotfullyutilizingahumanresourceof thecompany. In time,hewill atrophy,andhisperformancewill return toa“meetsrequirements”levelandstaythere.Thus, you’ll find two basic types of “meets” performers. One has no

motivation to do more or faces no challenge to do more. This is thenoncompetitor,whohasbecomesettledandsatisfiedinhis job.Theother typeof“meets”performeristhecompetitor.Eachtimehereachesalevelof“exceedsrequirements,”hebecomesacandidateforpromotion.Uponbeingpromoted,hevery likely becomes a “meets” performer again. This is the person Dr. Peterwrote about. But we really have no choice but to promote until a level of“incompetence”isreached.Atleast thiswaywedriveoursubordinatestowardhigher performance, and while they may perform at a “meets” level half thetime,theywilldothatatanincreasinglymorechallenginganddifficultjoblevel.Therearetimeswhenapersonispromotedintoapositionsomuchoverhis

head that he performs in a below-average fashion for too long a time. Thesolutionistorecyclehim:toputhimbackintothejobhedidwellbeforehewaspromoted.Unfortunately,thisisaverydifficultthingtodoinoursociety.Peopletend to view it as a personal failure. In fact, management was at fault formisjudgingtheemployee’sreadinessformoreresponsibility.Usuallythepersonwhowaspromotedbeyondhiscapabilityisforcedtoleavethecompanyratherthanencouragedtotakeastepback.Thisisoftenrationalizedbythenotionthat“It is better thatwe let himgo, forhisown sake.” I think it is deadwrong toforce someone in such circumstances out of the company. Instead, I thinkmanagementought to faceup to its ownerror in judgment and take forthrightanddeliberatestepstoplacethepersonintoajobhecando.Managementshouldalsosupport theemployee in thefaceof theembarrassment thathe is likely tofeel.Ifrecyclingisdoneopenly,allwillbepleasantlysurprisedhowshort-livedthatembarrassmentwillbe.Andtheresultwillbeapersondoingworkweknowfrompastexperiencehecanperformwell.Inmyexperience,suchpeople,oncetheyregaintheirconfidence,willbeexcellentcandidatesforanotherpromotionatalatertime—andthesecondtimetheyarelikelytosucceed.Insum,wemanagersmustberesponsibleandprovideoursubordinateswith

honestperformanceratingsandhonestmerit-basedcompensation.Ifwedo,theeventual result will be performance valued for its own sake throughout ourorganization.

16WhyTrainingIstheBoss’sJob

Recently my wife and I decided to go out to dinner. The woman who tookreservationsoverthephoneseemedflusteredandthenvolunteeredthatshewasnew and didn’t know all the ropes. No matter, we were booked. When weshowedupfordinner,wequickly learnedthat therestauranthad lost its liquorlicenseandthatitspatronswereexpectedtobringtheirownwineiftheywantedany.Asthemaitred’rubbedhishands,heasked,“Weren’tyoutoldthisonthephone when you made your reservations?” As we went through our dinnerwithoutwine,Ilistenedtohimgothroughthesameroutinewitheverypartyheseated. I don’t know for sure, but it’s probably fair to assume that nobodyinstructedthewomantakingcallstotellpotentialguestswhatthesituationwas.Instead, themaitred’had togo throughan ineptapology timeand timeagain,andnobodyhadwine—allbecauseoneemployeewasnotproperlytrained.The consequences of an employee being insufficiently trained can bemuch

moreserious.InaninstanceatIntel,forexample,oneofoursophisticatedpiecesofproductionmachineryinasiliconfabricationplant—amachinecalledanionimplanter—driftedslightlyoutoftune.Themachineoperator,likethewomanatthe restaurant, was relatively new. While she was trained in the basic skillsneededtooperatethemachine,shehadn’tbeentaughttorecognizethesignsofan out-of-tune condition. So she continued to operate themachine, subjectingnearlyaday’sworthofalmostcompletelyprocessedsiliconwaferstothewrongmachine conditions. By the time the situation was discovered, material worthmore thanonemilliondollarshadpassed through themachine—andhad tobescrapped.Because it takes over twoweeks tomake up such a losswith freshmaterial,deliveriestoourcustomersslipped,compoundingtheproblem.Situations like these occur all too frequently in business life. Insufficiently

trainedemployees,inspiteoftheirbestintentions,produceinefficiencies,excesscosts, unhappy customers, and sometimes even dangerous situations. Theimportanceof training rapidlybecomesobvious to themanagerwho runs intotheseproblems.

Forthealreadyoverscheduledmanager,thetrickierissuemaybewhoshoulddothetraining.Mostmanagersseemtofeelthattrainingemployeesisajobthatshould be left to others, perhaps to training specialists. I, on the other hand,stronglybelievethatthemanagershoulddoithimself.Letmeexplainwhy,beginningwithwhatIbelieveisthemostbasicdefinition

ofwhatmanagersaresupposedtoproduce.Inmyviewamanager’soutputistheoutput of his organization—no more, no less. A manager’s own productivitythusdependsonelicitingmoreoutputfromhisteam.A manager generally has two ways to raise the level of individual

performance of his subordinates: by increasingmotivation, the desire of eachpersontodohisjobwell,andbyincreasingindividualcapability,whichiswheretraining comes in. It is generally accepted thatmotivating employees is a keytaskofallmanagers,onethatcan’tbedelegatedtosomeoneelse.Whyshouldn’tthe same be true for the other principal means at a manager’s disposal forincreasingoutput?Trainingis,quitesimply,oneofthehighest-leverageactivitiesamanagercan

perform.Consider for amoment the possibility of your putting on a series offour lectures for members of your department. Let’s count on three hours ofpreparationforeachhourofcoursetime—twelvehoursofworkintotal.Saythatyouhave ten students inyour class.Nextyear theywillworka total of abouttwentythousandhoursforyourorganization.Ifyourtrainingeffortsresultina1percent improvement in your subordinates’ performance, your company willgaintheequivalentoftwohundredhoursofworkastheresultoftheexpenditureofyourtwelvehours.This assumes, of course, that the training will accurately address what

studentsneedtoknowtodotheirjobsbetter.Thisisn’talwaysso—particularlywithrespect to“cannedcourses”taughtbysomeonefromoutside.For trainingtobeeffective,ithastobecloselytiedtohowthingsareactuallydoneinyourorganization.Recently,someoutsideconsultantstaughtacourseoncareerdevelopmentat

Intel. Their approachwas highly structured and academic—and very differentfromanythingpracticedatthecompany.Whiletheyadvocatedcareerplansthatlooked ahead several years, together with carefully coordinated job rotationsbasedonthem,ourtraditionhasbeenmorelikeafreemarket:ouremployeesareinformedofjobopportunitieswithinthecompanyandareexpectedtoapplyfordesirable openings on their own initiative. Troubled by the disparity between

whatwastaughtinthecourseandwhatwaspracticed,theparticipantsgotabitdemoralized.For training to be effective, it also has to maintain a reliable, consistent

presence. Employees should be able to count on something systematic andscheduled,notarescueeffortsummonedtosolvetheproblemofthemoment.Inotherwords,trainingshouldbeaprocess,notanevent.Ifyouaccept that training,alongwithmotivation, is thewayto improvethe

performanceofyoursubordinates,andthatwhatyouteachmustbecloselytiedtowhatyoupractice, and that trainingneeds tobea continuingprocess ratherthanaone-timeevent,itisclearthatthewhoofthetrainingisyou,themanager.Youyourselfshouldinstructyourdirectsubordinatesandperhapsthenextfewranks below them. Your subordinates should do the same thing, and thesupervisorsateverylevelbelowthemaswell.Thereisanotherreasonthatyouandonlyyoucanfilltheroleoftheteacherto

yoursubordinates.Trainingmustbedonebyapersonwhorepresentsasuitablerole model. Proxies, no matter how well versed they might be in the subjectmatter,cannotassumethatrole.Thepersonstandinginfrontoftheclassshouldbeseenasabelievable,practicingauthorityonthesubjecttaught.We at Intel believe that conducting training is a worthwhile activity for

everyone from the first-line supervisor to the chief executive officer. Some 2percentto4percentofouremployees’timeisspentinclassroomlearning,andmuchoftheinstructionisgivenbyourownmanagerialstaff.We have a “university catalogue” that lists over fifty different classes. The

courses range fromproper telephonemanners to quite complicated productioncourses—like one on how to operate the ion implanter,which requires nearlytwohundredhours of on-the-job training to learn to use correctly, almost fivetimesthehoursof trainingneededtogetaprivatepilot’s license.Wetrainourmanagers in disciplines such as strategic planning as well as in the art ofconstructiveconfrontation,aproblem-solvingapproachwefavoratIntel.My own training repertoire includes a course on preparing and delivering

performancereviews,onconductingproductivemeetings,andathree-hour-longintroduction to Intel, inwhich I describe our history, objectives, organization,andmanagement practices.Over the years I have given the latter to a sizableproportionofourprofessionalemployees.Ihavealsobeenrecruitedtopinch-hitinothermanagementcourses.(Tomyregret,Ihavebecomefartooobsoletetoteachtechnicalmaterial.)

AtIntelwedistinguishbetweentwodifferenttrainingtasks.Thefirst taskisteaching newmembers of our organization the skills needed to perform theirjobs.Thesecond task is teachingnew ideas,principles,orskills to thepresentmembersofourorganization.The distinction between new-employee and new-skill training is important

because the magnitude of the tasks is very different. The size of the job ofdelivering anew-employee course is set by thenumberof newpeople joiningtheorganization.Forinstance,adepartmentthathas10percentannualturnoverandgrows10percentperyearhas to teach20percentof itsstaff thebasicsoftheirworkeachyear.Trainingeven20percentofyouremployeescanbeahugeundertaking.Teaching new principles or skills to an entire department is an even bigger

job.Ifwewanttotrainallofourstaffwithinayear,thetaskwillbefivetimesaslargeastheannualtaskoftrainingthe20percentwhorepresentnewmembers.RecentlyIlookedatthecostofdeliveringanewone-daycoursetoourmiddle-management staff. The cost of the students’ time alonewas over onemilliondollars.Obviouslysuchataskshouldnotbeenteredintolightly.Sowhat should you do if you embrace the gospel of training? For starters,

make a list of the things you feel your subordinates or the members of yourdepartmentshouldbetrainedin.Don’tlimitthescopeofyourlist.Itemsshouldrange from what seems simple (training the person who takes calls at therestaurant) to loftier and more general things like the objectives and valuesystems of your department, your plant, and your company. Ask the peopleworking for youwhat they feel they need. They are likely to surprise you bytellingyouofneedsyouneverknewexisted.Havingdonethis,takeaninventoryofthemanager-teachersandinstructional

materials available to help deliver training on items on your list. Then assignprioritiesamongtheseitems.Especially if you haven’t done this sort of thing before, start very

unambitiously—likedevelopingoneshortcourse(threetofourlectures)onthemosturgentsubject.Youwillfindthatskillsthatyouhavehadforyears—thingsthatyoucoulddoinyoursleep,asitwere—aremuchhardertoexplainthantopractice.Youmayfindthatinyourattempttoexplainthings,you’llbetemptedtogo intomore andmorebackgrounduntil thisbegins toobscure theoriginalobjectiveofyourcourse.Toavoidlettingyourselfbogdowninthedifficulttaskofcoursepreparation,

setascheduleforyourcourse,withdeadlines,andcommityourselftoit.Createanoutlineforthewholecourse,developjustthefirstlecture,andgo.Developthesecondlectureafteryouhavegiventhefirst.Regardthefirsttime

youteachthecourseasathrowaway—itwon’tbegreat,becausenomatterhowhardyou try,you’llhave togo throughoneversion thatwon’tbe.Rather thanagonizeoverit,accepttheinevitabilityofthefirsttimebeingunsatisfactoryandconsideritthepathtoamoresatisfactorysecondround.Tomakesurethatyourfirstattemptcausesnodamage,teachthiscoursetothemoreknowledgeableofyour subordinates,whowon’tbeconfusedby itbutwhowillhelpyouperfectthecoursethroughinteractionandcritique.Withyoursecondattemptintheoffing,askyourselfonefinalquestion:Will

youbeabletoteachallmembersofyourorganizationyourself?Willyoubeableto cover everybody in one or two courses, orwill it require ten or twenty? Ifyour organization is large enough to require many repetitions of your coursebefore different audiences, then set yourself up to train a few instructorswithyourfirstsetoflectures.After you’ve given the course, ask for anonymous critiques from the

employees in your class. Prompt them with a form that asks for numericalratingsbut that alsoposes someopen-endedquestions.Studyandconsider theresponses,butunderstand thatyouwillneverbeable topleaseallmembersofyour class: typical feedback will be that the course was too detailed, toosuperficial,andjustright,inaboutequalbalance.Yourultimateaimshouldbetosatisfyyourselfthatyouareaccomplishingwhatyousetouttodo.Ifthisisyourfirsttimeteaching,you’lldiscoverafewinterestingthings:

• Training is hard work. Preparing lectures and getting yourself ready tohandleall thequestions thrownatyou isdifficult.Even ifyouhavebeendoing your job for a long time and even if you have done yoursubordinates’jobsingreatdetailbefore,you’llbeamazedathowmuchyoudon’tknow.Don’tbediscouraged—thisistypical.Muchdeeperknowledgeof a task is required to teach that task than simply to do it. If you don’tbelieveme,tryexplainingtosomeoneoverthephonehowtodriveastick-shiftcar.

•Guesswhowillhavelearnedmostfromthecourse?You.Thecrispnessthatdeveloping it gave to your understanding of your own work is likely initselftohavemadetheeffortextremelyworthwhile.

•Youwillfindthatwhenthetrainingprocessgoeswell,itisnothingshortof

exhilarating. And even this exhilaration is dwarfed by the warm feelingyou’llgetwhenyouseeasubordinatepracticesomethingyouhavetaughthim.Relishtheexhilarationandwarmth—it’llhelpyoutoarmyourselffortacklingthesecondcourse.

OneMoreThing…

Please! You invested the price of this book plus perhaps eight hours of yourtime.Attheriskofsoundingliketheauthorofadietbook,Iwouldaskyoutodosomethingspecific,andIleaveyouwithasetofassignments.Choosewhatyoulike—butchoosesome—andperformthemhonestly.Youhave trustedmeenoughtobuymybookandread it.Nowletmesaya

finalthing:ifyoudoatleast100pointsworthofwhatyoufindhere,you’llbeadistinctlybettermanagerforit.

Production Points

Identify the operations in yourworkmost like process, assembly,andtestproduction.

10

Foraprojectyouareworkingon,identifythelimitingstepandmapouttheflowofworkaroundit.

10

Definetheproperplacesfortheequivalentsofreceivinginspection,in-process inspection, and final inspection in your work. Decidewhether these inspectionsshouldbemonitoringstepsorgate-like.Identifytheconditionsunderwhichyoucanrelaxthingsandmovetoavariableinspectionscheme.

10

Identifyhalfadozennewindicatorsforyourgroup’soutput.Theyshouldmeasureboththequantityandqualityoftheoutput.

10

Install these new indicators as a routine in your work area, andestablishtheirregularreviewinyourstaffmeetings.

20

Whatisthemostimportantstrategy(planofaction)youarepursingnow? Describe the environmental demand that prompted it andyourcurrentstatusormomentum.Isyourstrategylikelytoresultina satisfactory state of affairs for you or your organization ifsuccessfullyimplemented?

20

Leverage

Conductworksimplificationonyourmosttedious,time-consumingtask. Eliminate at least 30 percent of the total number of stepsinvolved.

10

Define your output: What are the output elements of theorganizationyoumanageandtheorganizationsyoucaninfluence?Listtheminorderofimportance.

10

Analyze your information-and knowledge-gathering system. Is itproperly balanced among “headlines,” “newspaper articles,” and“weeklynewsmagazines”?Isredundancybuiltin?

10

Takea“tour.”Afterward, list the transactionsyougot involvedinduringitscourse.

10

Createaonce-a-month“excuse”foratour. 10

Describehowyouwillmonitor thenextprojectyoudelegate toasubordinate.Whatwillyoulookfor?How?Howfrequently?

10

Generate an inventory of projects on which you can work atdiscretionarytimes.

10

Hold a scheduled one-on-one with each of your subordinates.(Explaintotheminadvancewhataone-on-oneisabout.Havethemprepareforit.)

20

Lookatyourcalendarforthelastweek.Classifyyouractivitiesaslow-/medium-/high-leverage.Generateaplanofactiontodomoreofthehigh-leveragecategory.(Whatactivitieswillyoureduce?)

10

Forecastthedemandonyourtimeforthenextweek.Whatportionofyour timeis likelytobespent inmeetings?Whichof theseareprocess-orientedmeetings?Mission-orientedmeetings?Ifthelatterareover25percentofyourtotaltime,whatshouldyoudotoreducethem?

10

Definethethreemostimportantobjectivesforyourorganizationforthenextthreemonths.Supportthemwithkeyresults.

20

Have your subordinates do the same for themselves, after athoroughdiscussionofthesetgeneratedabove.

20

Generateaninventoryofpendingdecisionsyouareresponsiblefor.Take three and structure the decision-making process for them,usingthesix-questionapproach.

10

Evaluate your own motivational state in terms of the Maslowhierarchy.Dothesameforeachofyoursubordinates.

10

Give your subordinates a racetrack: define a set of performanceindicatorsforeach.

20

List thevariousformsoftask-relevantfeedbackyoursubordinatesreceive.Howwellcantheygaugetheirprogressthroughthem?

10

Classifythetask-relevantmaturityofeachofyoursubordinatesaslow,medium, or high.Evaluate themanagement style thatwouldbemostappropriateforeach.Comparewhatyourownstyleiswithwhatitshouldbe.

10

Evaluatethelastperformancereviewyoureceivedandalsothelastset of reviews you gave to your subordinates as a means ofdelivering task-relevant feedback.Howwelldid thereviewsdo toimproveperformance?Whatwasthenatureof thecommunicationprocessduringthedeliveryofeach?

20

Redooneofthesereviewsasitshouldhavebeendone. 10

Notes

1 The group of middle managers I referred to in the AcknowledgmentsconsistedofC.Bickerstaff,J.Crawford,R.Hamrick,B.Kraft,B.Kubicka,D.Lenehan,D.Ludington,B.Maxey,B.McCormick,C.McMinn,B.Michael,S.Overcashier,B.Patterson,J.Rizzo,R.Schell,J.Vidal,J.Weisenstein,andD.Yaniec.

PARTI2MANAGINGTHEBREAKFASTFACTORY

1Ilearnedthemetaphorof“cuttingwindowsinablackbox,”alongwithmanyotherthingsaboutproduction,fromalong-timeassociate,GeneFlath.

2adelugeofvisaapplications:“LaDolceVisa,”Time,June22,1981,pp.16,19.

PARTII3MANAGERIALLEVERAGE

1amanager’sworkisneverdone:LestyouthinkIamunique,Ihastentopointout that I am not; I discovered, with great relief, in a study by HenryMintzberg (“The Manager’s Job: Folklore and Fact,” Harvard BusinessReview, vol. 53, no. 4, July-August 1975, pp. 49–61) that other managers’daysarealtogethersimilartomine.

2 The idea of “nudging” as an important element of the decision-makingprocesswaspointedoutbymycolleagueLesVadasz.

4MEETINGS—THEMEDIUMOFMANAGERIALWORK

1 “The good time users…”: Peter Drucker, People and Performance: PeterDruckeronManagement(NewYork:Harper’sCollegePress,1977),p.57.

5DECISIONS,DECISIONS

1 “In the meeting…”: Robert L. Simison, “Ford Fires an Economist,”WallStreetJournal,July30,1980,p.20.

2 This role-playingexperiment,aswellas thepeer-groupsyndrome,was firstsuggestedbyGerryParker,aseniortechnologistatIntel.

3 The six-question approach to expedite the decision-making process wassuggestedbyLesVadaszofIntel.

4 “Group decisions…”: Alfred P. Sloan, Jr.,My Years with GeneralMotors(NewYork:Doubleday,1964),p.512.

6PLANNING1Columbus:Tospreadmyguiltintinkeringwithhistory,IhastentocreditmycolleaguesHarryChapmanandRosemaryRemadeforthisadaptation.

PARTIII8HYBRIDORGANIZATIONS

1“Goodmanagement…”:Sloan,op.cit.,p.505.

9DUALREPORTING1 Books have beenwritten aboutmatrixmanagement:An example is JayR.Galbraith, Designing Complex Organizations (Reading, Mass.: Addison-Wesley,1973).

2 “A university is an odd place to manage….”: John A. Prestbo, “PinchingPennies:OhioUniversityFindsParticipatoryPlanningEndsFinancialChaos,”WallStreetJournal,May27,1981,pp.1,20.

10MODESOFCONTROL

1 thethreemeansofcontrol:OliverE.Williamson,MarketsandHierarchies:AnalysisandAntitrustImplications(NewYork:FreePress,1975);RaymondL. Price andWilliamG.Ouchi, “Hierarchies, Clans and Theory Z:ANewPerspective on Organization Development,” Organizational Dynamics,Autumn1978,pp.35–44.

PARTIV11THESPORTSANALOGY

1 Maslow’s theory: Abraham H. Maslow,Motivation and Personality (New

York:Harper&Row,1954).2 “ItastoundsJoeFrazier…”:“FightOneMoreRound,”Time,December14,1981,p.90.

3 “thrived on beating the competition…”: Bundsen, syndicated column,PeninsulaTimesTribune(PaloAlto,Calif.),September18,1982,p.B-3C.

12TASK-RELEVANTMATURITY

1 For a compilation of work on task-relevantmaturity, see Paul Hersey andKenneth H. Blanchard,Management of Organizational Behavior, 2nd. ed.(NewYork:Prentice-Hall,1972).

16WHYTRAININGISTHEBOSS’SJOB1ThematerialinthischapteroriginallyappearedintheJanuary23,1984,issueofFortune.

What’snextonyourreadinglist?

Discoveryournextgreatread!

Getpersonalizedbookpicksandup-to-datenewsaboutthisauthor.

Signupnow.