best practices study

Upload: abidyounas

Post on 07-Apr-2018

215 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/4/2019 Best Practices Study

    1/123

    PPAAKKIISSTTAANNSSSSOOFFTTWWAARREEIINNDDUUSSTTRRYY

    BBEESSTTPPRRAACCTTIICCEESS&&SSTTRRAATTEEGGIICCCCHHAALLLLEENNGGEESS

    AANNEEXXPPLLOORRAATTOORRYYAANNAALLYYSSIISS

    MMIINNIISSTTRRYYOOFFIINNFFOORRMMAATTIIOONNTTEECCHHNNOOLLOOGGYY

    GGOOVVEERRNNMMEENNTTOOFFPPAAKKIISSTTAANN

    IISSLLAAMMAABBAADD

    FEBRUARY2005

  • 8/4/2019 Best Practices Study

    2/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 2

    Copyrights2005

    PakistanSoftwareExportBoard(G)Ltd.

    MinistryofInformationTechnology

    GovernmentofPakistan

    Printing

    March2005

    Publishedby

    PakistanSoftwareExportBoard

    TheFundingAgency

    TheBestPractices inPakistaniSoftwareSector Project is fundedby the PakistanSoftwareExportBoard (PSEB).

    PSEBistheentitywithinGovernmentchargedwiththetaskofenhancingexportsofsoftwareandITenabledservices

    (ITES) from Pakistan. PSEB is a guarantee limited company totally owned and fundedby the Government of

    Pakistan.AnyquestionsorcommentsaboutthisreportmaybedirectedtoPSEBIslamabadat9251111333666orthroughemailat [email protected].

    Disclaimer

    ThereportispublishedbyPSEBfortheuseofitsmembers&theITindustry.Thisreportisaresultofa3monthlongindependent

    researchstudyconductedbytheprincipalconsultantwithsupportfromPSEB. ThestudyalsoincorporatesfeedbackfromPSEB,

    MinistryofITandTelecom(MOITT)andstakeholdersofthePakistaniITindustry. Itfaithfullyreportswhattheconsultantfound

    theonthegroundrealityofthePakistanisoftwareindustrytobeandaccuratelyreflects(andwhereverpossibleattributestoothers)

    theopinionshewasabletoformonthebasisofhisdiscussionsandonsitevisitstoabout50Pakistanisoftwarecompanies.Tothat

    effect, the reportsolely reflects theviewsof the consultantandmayormaynotreflect thoseofPakistanSoftwareExportBoard

    (PSEB),theMinistryofITandTelecom(MOITT),ortheGovernmentofPakistan(GOP). Thestudyadvisorsorthecontributorsare

    notresponsible,inanywaypossible,fortheerrors/omissionsofthisreport.

    ThisreportisabestintentionedefforttodisseminateinformationaboutthePakistansSoftwareIndustryandshouldnotbeusedas

    asolemeansofadviceformakinginvestmentdecisions.PSEBdoesnotacceptanyliabilityforanydirectandconsequentialuseof

    thisreportoritscontents.ThecontentsofthisreportmaybereproducedonlyafterpriorpermissionfromPSEB.

  • 8/4/2019 Best Practices Study

    3/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 3

    CONTENTS

    1.EXECUTIVESUMMARY ......................................................................................................................................4

    2.BACKGROUND&INTRODUCTION ..............................................................................................................11

    2.1BACKGROUNDANDMOTIVATIONFORTHESTUDY......................................................................................12

    2.2INTRODUCTORYREVIEWOFTHERELEVANTLITERATURE...........................................................................13

    3.THEOBJECTIVES,AUDIENCE,ANDFORMATOFTHESTUDY..............................................................16

    3.1THEANALYTICAGENDA: .............................................................................................................................16

    3.2THEBENEFITSANDINTENDEDAUDIENCE: ..................................................................................................18

    3.3THEFORMATOFTHESTUDY: ........................................................................................................................18

    4.ABRIEFNOTEONPROJECTMETHODOLOGY...........................................................................................19

    5.ASTATISTICALSNAPSHOTOFPAKISTANSSOFTWAREINDUSTRY.................................................22

    5.1ESTABLISHINGAPOINTOFREFERENCEFORPAKISTANSSOFTWAREINDUSTRY ................22

    5.2SOFTWAREDEVELOPMENTINPAKISTAN:STATISTICSONMANAGERIALANDTECHNICALPATTERNS.....24

    5.3SEARCHFORTHEHOLYGRAIL:DOSTATISTICSREVEALAPATTERNOFBESTPRACTICES?...................49

    6.UNDERSTANDINGPROMINENTBUSINESSMODELS&COMPETITIVEDRIVERS...........................53

    6.1ATAXONOMYOFGENERICSOFTWAREBUSINESSMODELS.........................................................................54

    6.2THEEXPORTFOCUSEDLOCALFIRM(THESYSTEMSORNETSOLMODEL) ..........................................59

    6.3THEDOMESTICFOCUSEDLOCALFIRM(THETPSORLMKRMODEL)...............................................68

    6.4THEEXPORTFOCUSEDFOREIGNFIRM(THETECHLOGIXORETILIZEMODEL)..................................80

    6.5THEDEDICATEDDEVELOPMENTCENTER(THEITIMASSOC.ORCLICKMARKSMODEL) .................90

    7. ENVIRONMENTAL,INFRASTRUCTURE &PUBLICPOLICYCHALLENGES....................................101

    7.1TELECOMINFRASTRUCTURECOST&AVAILABILITY .................................................................................1057.2AVAILABILITYOFVENTUREANDRISKCAPITAL........................................................................................106

    7.3UNDERDEVELOPEDDOMESTICMARKET...................................................................................................107

    7.4AVAILABILITYOFPHYSICALINFRASTRUCTURE .........................................................................................108

    7.5INTELLECTUALPROPERTYRIGHTS .............................................................................................................110

    8.CONCLUSIONS&RECOMMENDATIONS..................................................................................................111

    8.1SUMMARYOFRESEARCHRESULTSANDFUTUREDIRECTIONS ..................................................................112

    8.2THEWAYOFTHEFUTURE:SOMETENTATIVECONCLUSIONS...................................................................114

    9.APPENDIXA:LISTOFORGANIZATIONSSURVEYED/INTERVIEWED ..............................................117

    10.LISTOFBIBLIOGRAPHICREFERENCES ...................................................................................................119

    11.ABOUTTHEAUTHOR/CONSULTANT....................................................................................................123

  • 8/4/2019 Best Practices Study

    4/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 4

    PAKISTANSSOFTWAREINDUSTRY

    BESTPRACTICES&STRATEGICCHALLENGES

    AN

    EXPLORATORY

    ANALYSIS

    1.EXECUTIVESUMMARY

    The software industrywidely seen as the great enablerprovides an opportunity to the

    developingcountriestoplayagreatereconomicroleinthefastglobalizingworld.Theexample

    of neighboring Indiawhose ambition and progress towards becoming a mini (software)

    superpowerisnomysteryfromtheworldisoftencitedinthedevelopmentliteratureasan

    evidenceofthefact.Pakistanssoftwareindustrywidelyperceivedtobesharinganumberof

    keyfactorswithIndiahasembarkeduponanambitiouseffortofitsowntoclaimitssharein

    therichesoftheworldssoftwaremarkets.Pakistaniscurrentlyviewedasatier3countryinawidely quoted taxonomy of software exporting nations (Carmel, 2003). It is widelybelieved

    that,withthewealthoftalentandstrengthsavailable,thecountrydeservesabetterplaceinthis

    global pecking order of software exporting nationsatleast a tier2 status like Russia and

    China,orevenatier1statusalongsidearchrivalIndia1.

    Pakistanssoftwareindustryhasbeenasubjectofthecuriosityofinterestedbystandersboth

    localandexpatriateentrepreneursindustryanalysts,andpotentialinvestorsalike.Yet,lackof

    credible dataon thecurrent stateandcompetitive dynamicsof the industry has oftenbeena

    hindrance in engaging these individuals and materializing many prospective ventures. We

    wererecentlyinvolved,ontherequestofanexpatriateinvestor,inanefforttoincubateanIT

    focused venture capital in Pakistan. As we spoke with industry leaders and the financial

    community,werepeatedlyencounteredaseriesoftoughquestions,forexample:

    Why hasnt the Pakistani software industrybeen able to produce a single worldclasssoftwarefirm(e.g.Wipro,InfosysorTCSofIndia)inthelast1015years?

    Why havent webeen able to grow Pakistani software exportsbeyond a certain level($3060millionperannum)forthelast5years?

    DoesPakistanisoftware industrymerelyrepresenta lower levelofdevelopmentoranaltogetherdifferentdevelopmenttrajectoryascomparedtoknownpeernations?

    What constitutes a generalized set ofbest practices in the local software industry (i.e.whatdifferentiatesbetterperformersfromthosethatdontperformthatwell)?

    This study attempts to answer some of these questions. While several factors are widely

    believedtobeahindranceinthecountrysaspirationtobecomeasignificantsoftwareexporter,

    1 A widely quoted GOP target of $1B in software exports by Y2000 would have propelled Pakistan into theexclusive tier-1 club.

  • 8/4/2019 Best Practices Study

    5/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 5

    nottheleastimportantofwhicharemacro andgeopoliticalinnature(e.g.lawandorderand

    securitysituation,imageofthecountryetc.),weadoptaninsideoutapproachthatasks:What

    can the various players, essentially software companies, in the industry learn from each

    other? There is a growing realization that we must truly understand the structure of the

    PakistanisoftwareindustryandthenatureofPakistanscompetitiveadvantageinthesoftware

    arena in order to devise better industrial and organizational strategies and public policy

    interventions.TheBestPracticesinPakistaniSoftwareSectorProjectbeingthefirstofitskind

    andscopeinPakistanisanexploratorystudyofthePakistanisoftwareindustrythatattempts

    todojustthat.

    The study draws upon an onthespot survey of 40 of the most prominent and largest

    software companies in Pakistan, as identified by PSEB and PASHA. We conducted

    organizationalinterviewswithseniorexecutives(CEOs/CTOsorLocalofHeadsofOperations)

    of 47 of these companies to supplement the statistical data with qualitative insights. These

    interviewsfocused

    on

    understanding

    these

    organizations,

    their

    business

    and

    revenue

    models,

    competitivedrivers,strategicchallenges,andpolicybottlenecks.Wealsoconductedinterviews

    ofopinionleaders,policymakers,andseniorexecutivesofotherorganizationalentities(e.g.IT

    MNCs,financialinstitutions,andacademia)thathadasignificantbearingonthelocalsoftware

    industry.Inallweconductedover65interviewsbetweenOct.Dec.timeframe(seeAppendix)

    The substantive findings of the study canbebroadly divided into two components. The first

    part attempts at creating abrief statistical snapshot of the Pakistani software industry, as

    gleaned from the data on organizational, managerial, and technical practices of our

    respondents.Thesecondpartofthestudyusestaxonomyofgenericsoftwarebusinessmodels

    todevelopaqualitativesenseofsoftwaredevelopmentactivityinPakistan.Italsoidentifieskeystrategic challenges (13 in all) typically faced by companies within each of these generic

    business models and managerialbest practices (20 in all) adoptedby various players in the

    industrytomeeteachofthesestrategicchallenges.Thereportconcludeswithadiscussionon

    environmentalandpolicybottlenecksandsometentativeconclusions

    The results of the statistical analysis are quite illuminating. On the whole, the 60 software

    houses included in our statistical sample employ over 4000 technical and professional

    employeesforanaverageof62employeesperorganization.Roughlyone third (32%)of the

    software companies reported annual revenues of more than a million dollars with some

    reportingmorethan$5M,anotherthird(36%)between$200Kand$1M,andtherest(32%)lessthan$200K.6ofthecompanieshadmorethan250employeesandanother8hadbetween100

    and250employees.Onthewholethese60companieshadexperiencedanemploymentgrowth

    ofabout27.5%andarevenuegrowthof37.4%overthelastyearpointingatbetterutilization

    of excesscapacityor valueadditionperemployee,orboth.Around 40%of thecompanies in

    our sample were subsidiaries of foreign companieswith majority of them having a parent

    companyintheUnitedStates.55%ofthecompanieshadoneormorefrontofficesabroad(50%

  • 8/4/2019 Best Practices Study

    6/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 6

    in the US, 11% each in UK and Middle East, and 3% in the Asia Pacific region). 45% of the

    respondents had quality certification (mostly ISO9000 with only 3% having CMM). 73.7% of

    thecompanieshaddedicatedqualityassuranceteams.

    Broadlyspeaking,ourrespondentsderivetheirrevenuesfromexportanddomesticmarketsina

    ratioof60:40.Ontheexportsside,theyderive22.5%and38.5%oftherevenuesfromproductsand services respectively. Although we did not ask directly, our conversations with the top

    leadersoftheindustrysuggestthatamajorityoftheproductexportsarecustomizedrather

    than shrinkwrapped products. On the domestic side, however, the ratios are somewhat

    reversedwithproductsandservicescontributing23%and16.5%respectively.Ourrespondents

    predominantly serve the private sector markets with around 85% of the total sales going to

    privatesector(localandforeigncombined)andtherestgoingtopublicsector,equallydivided

    betweendomesticandforeign.

    We tried to parse the data into various classifications in an attempt to understand the

    organization and dynamics of software industry. For example, we looked at the differences

    betweenexportfocused,domesticfocused,andhybridsoftwareoperations;betweenproduct

    focused, servicesfocused, and hybrid operations;between large and small operations; and

    between operations formed prior to and after the DotCom Bubbleburst in the United States.

    Ourresultsaresuggestiveofseveralinterestingtrends.

    Forexample,on themanagerialpracticesside, there issomesuggestiveevidence thatexport

    focusedsoftwareoperationsaremorelikelytodistributestocks/ownershipamongemployees,

    hold employee bonding activities, and benefit from employeedriven innovation while

    domesticfocusedsoftwareoperationsaremorelikelytoshareprofitswithemployees,provide

    additionalbenefitstofemaleemployees,havegreaterfinancialdiscipline,andprovidetimeto

    employeestoworkontheirowninterests.Despitethelatter,however,theyseemtobenefitless

    from employeedriven innovation and suffer more from a perception of lower delegation

    quality.Hybridsfallinbetweenthetwocategoriesonalmostallthesemeasures.

    Exportfocusedoperationstendtospendmore,onaverage,onqualityassurancewhilehybrids

    tend tohaveagreaterpropensityforseekingaqualitycertification.Allcompanies,across the

    board,prefertouseandexpressgreatersatisfactionwithhighcontactapproachesofmarketing

    (e.g.wordtomouth,oneononecontacts,andpreestablishednetworks).Wedonotfindalot

    of

    differences

    between

    the

    cost

    structures

    of

    export

    focused,

    domestic

    focused,

    or

    hybrid

    operations, except that hybrids seemed to underinvest in productdevelopment to pay for

    expensive marketing and advertising, and training and certification. CEOs of exportfocused

    softwareoperationstendtospendmuchmoretimeintacticalratherthanstrategicmode(doing

    daytodaymanagementratherthanmarketingandbusinessdevelopment).

    Our analysis of other classifications provides few interesting insights. The dedicated

    development centers tend to be smaller, more rigorous (from a technical and process

  • 8/4/2019 Best Practices Study

    7/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 7

    standpoint)thantherestoftheindustry.They,however,seemtoexperienceseriousconstraints

    torevenueandemploymentgrowthafactthatweinterpretasamanifestationoftheirmid

    life crisis. Although we see a trend towards productization in the industry, we found few

    significant differencesbetween productfocused and servicesfocused operations. This lack of

    differentiation (e.g. in the cost structures of services and productfocused operations) is

    problematic, to say the least. There were also few significant differencesbetween large and

    smallsoftwareoperationsandbetweenthosecreatedbeforeandaftertheDotComBubbleburst.

    Onthewholethesefindingsalsopaintapictureoflackoffocusandspecializationwithinthe

    Pakistanisoftware industry. Thoseproductfocused operationsaresimilar to servicesfocused

    operations and preDotCom operations are not qualitatively different from postDotCom

    operations does not speak well for the maturity of the industry as a whole. A related

    substantivefindingisthetrendtowardsthehybridizationofsoftwaredevelopmentactivity.

    Thehybrid firmhasemergedasan importantorganizationalclasson itsownrather than the

    averageof

    the

    two

    extremes.

    While

    the

    hybrid

    firm

    tends

    to

    do

    better

    than

    the

    two

    extremes

    on

    somemeasuresandhencemightbeseenasamanifestationoftheindustryssurvivalinstinct,it

    isnotquiteclearifitistheoptimalmodeloforganizationofsoftwaredevelopmentactivityin

    thelongrun.

    In line with the study objectives, we also asked the question: Do aggregate statistics reveal a

    patternofbestpracticeswithin thesoftware Industry?Weusemultiplecomparisongroups

    (e.g. 40 most prominent companies, top10 companies, 14 fastest growing companies, 14

    companiesthatdescribethemselvesasgloballycompetitiveagainsttherestoftheindustry)and

    find mixed results on thataccount.For example, we find robust evidence to support the fact

    that betterperforming companies tend to adopt a set of employeefriendly managementpractices (e.g. flexibility, stockownership,profitsharingetc.) and haveaccess tohigh quality

    managerial talent (e.g. mix of technical andbusinessbackgrounds, prior venture experience,

    financial disciplineetc.) than therestof the industry.All companies,across theboard, prefer

    highcontact marketing approaches over lowcontact onesbutbetterperforming companies

    reporthighersatisfactionwiththeformerthantherestoftheindustry.Ourresultsonvarious

    measuresoftechnicalandprocessqualityare,however,inconclusive,atbest.Here,wedonot

    find any clear patterns that differentiatebetterperforming companies from the rest of the

    industry.Webelievethatbestpracticeswithintechnicalandprocessrealmsaredependenton

    thetypeofworkperformedandanumberofprojectspecificvariables.Asreportedelsewhere,

    therefore,projectleveldatamightbebettersuitedtoidentifythesedifferences.

    Next,basedonourstatisticalfindingsandqualitativeinsights,wedevisea4parttaxonomyof

    generic business models. The four subclassifications, named after their most prominent

    examples, include: Exportfocused Local Firm (Systems or Netsol Model), Domestic

    Focused Local Firm (TPS or LMKR Model), Export Focused Foreign (Expatriate) Firm

    (TechlogixorEtilizeModel),andDedicatedDevelopmentCenter(ITIMorClickmarks

  • 8/4/2019 Best Practices Study

    8/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 8

    Model).Wepresentasnapshotofeachofthesegenericsoftwarebusinessmodelsandidentify

    key strategic challenges for each13 in all for the entire industry. As we discuss the ways

    relativelymoresuccessful firms in the industryhavecountered thesestrategicchallenges,we

    alsoarriveattwenty(20)managerialbestpracticesthatcouldbereplicatedbyotherplayersin

    theindustry.

    The Exportfocused Local Firm is one founded by a predominantly Pakistanbased

    entrepreneurial team (that may or may not have been aided/encouraged by a group of

    expatriates),butwithanexplicitpurposeofexportingsoftwareproductsorservices.Majorityof

    thefirmsestablishedinpreDotComBubblebursterawithanexpressedpurposeofexporting

    services to North America and

    WesternEuropeancountriesfallin

    this category. Although there are

    somethathavetakentheproducts

    route,their

    numbers

    are

    relatively

    smaller than those focusing on

    export of services. The most

    defining feature of this class of

    companies, namely, the local

    presenceoftheirfoundersandthe

    exportorientation of their

    products/ services, brings a

    number of unique and important

    challenges to this type of a firm.

    We discuss three of these in great

    detailandalludetoseveralothers.

    The ones we discuss in depth include: customer acquisition in a foreign market, setting up a

    foreignmarketingpresence,andunderstandingthedomainandcontextofaforeigncustomer.

    Some salient examples of this type ofbusiness model in action are: ThreesixtyDegreez, Post

    Amazers,AdvancedCommunications,Makabu,Netsol,andAutosoftDynamicsetc.

    TheDomesticfocusedLocalFirm,withanexceptionofafewcompanies,isreallyonebecause

    ofcircumstancesratherthanchoice.Moreoftenthannot,andlogicallyso,thedomesticfocused

    local firm plans to export its products or services abroad and is merely using the domestic

    market as a vehicle to gain a track record with real life customers. Whether a firm is in this

    categorybychoice(Illdodomesticfirst,exportlater)orbycircumstances(Sincetheexport

    market doesnt seem very good right now, Ill survive by selling at home) the strategic

    challengesarequitesimilar.Wediscussthreeoftheseinsomedetail.Theseinclude:operating

    inanunderdevelopedlocalmarket,gettingaccesstocapital,andhavingabusinessplananda

    strategic/domainfocus.Otherchallengesalludedtoinclude:migratingfromthedomestictothe

    export market, developing relationships, delivering quality products/services, and even

    ZRG

    TPS

    Lumensoft

    Yevolve

    2B Technologies

    SI3

    Softech Systems

    Genesis Solutions

    Alchemy Technologies

    AppXS

    Oratech

    Askari Info Systems

    Acrologix

    Comcept

    LMKR

    CARE

    ThreeSixtyDegreez

    Post Amazers

    Advanced Comm.

    Netsol

    Makabu

    Autosoft Dynamics

    Sidaat Hyder Morshed

    Avanza Solutions

    GoNet

    Kalsoft

    Jinn Technologies

    Secure3 Networks

    Systems Ltd

    Progressive Systems

    Millennium Software

    Cressoft

    Etilize

    Prosol

    Adamsoft

    Ultimus

    MixIT

    Techlogix

    Xavor

    Elixir Technologies

    ITIM Associates

    MetaApps

    Clickmarks

    Enabling Tech. (Quartics)

    Trivor Systems

    Strategic Systems Intl

    ESP Global Systems

    DOMESTICFOCUSED

    LOCALFIRM

    EXPORTFOCUSED

    LOCALFIRM

    EXPORTFOCUSED

    FOREIGNFIRM

    DEDICATED

    DEVELOPMENTCENTER

    FIGUREGENERICBUSINESSMODELS&THEIRTRANSITIONSSCENARIOS

    DIVERSIFICATION

    M&A

    W/

    FOREIGN

    FIRM

    BUYOUTBYLOCALMGMT.MATURITY,VALUEADD ELEVATIONOFPAKOPS.SHIFTING

    PRIORITIES

    TRANSITIONSKEY

  • 8/4/2019 Best Practices Study

    9/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 9

    marketing abroad. Some salient examples of this type ofbusiness model in action are: 2B

    Technologies, ZRG, TPS, Lumensoft, Yevolve, SI3, Softech Systems, AppXS, and Genesis

    Solutionsetc.

    The Exportfocused Foreign Firm is one founded abroad (or jointly, in Pakistan), by a

    predominantlyforeign(usually,anexpatriate)entrepreneurialteam,withanexplicitpurposeofusing the Pakistanbased offshore development facility to deliver a product or service

    demandedbytheforeignmarket.Thistypeofbusinessmodelhasbeenadoptedbyservicesand

    productfocusedcompaniesalike.Whilethisclassofcompaniesenjoysseveraladvantagesover

    thoseinearlierdiscussedcategories,namely,qualityofduediligenceonthebasicidea,foreign

    contacts/networksoffounders,andbetteraccesstocapitaletc.,therearesignificantchallenges

    aswell.Wediscussfourofthesechallengesinsomedetailandidentifyanumberofmanagerial

    bestpracticesfollowedbysomeoftheinterviewees.Thesechallengesinclude:dealingwiththe

    imageproblem,counteringthegeographicallyshiftinglaborarbitrageargument,scalingup

    the

    Pakistanbased

    operation,

    and

    getting

    to

    know

    the

    land

    and

    managing

    expectations

    etc.

    Some salient examples of this type ofbusiness model in action are: Elixir, Etilize, Ultimus,

    MixIT,TechLogix,Prosol,andXavoretc.

    TheDedicatedOffshoreDevelopmentCenter,asthenamesuggests,isafairlylimitedoffshore

    operation of a foreign company. It is different from the ExportFocused Foreign (Expatriate)

    Firm inthesense that it isoftenanaddontoanalreadyexistingcompanywhosestrategic

    andmanagerialprocessesandcontrolsarequitewellestablished.Duetoitsuniquenature(i.e.

    limited scope) it faces a number of challenges that are distinct from the earlierdiscussed

    category. We discuss three key challenges facedby organizations in thisbusiness model and

    identify innovative best practices to counter these. These include: managing the parentsubsidiary relationship, setting up an offshore facility in Pakistan, and building a quality

    softwaredevelopmentoperation.Somesalientexamplesofthistypeofbusinessmodelinaction

    are: MetaApps, ITIM Associates, Clickmarks, Trivor Systems, and Strategic Systems

    Internationaletc.

    The taxonomyofgenericsoftwarebusinessmodelsmaybehelpful inseveralways.Firstly, it

    gives us a relatively easy and comprehensive way to classify a particular software operation

    into a broad enough category of organizations and a hence a reference point to compare

    ourselves against. Secondly, it highlights the importance of understanding the strengths,

    weaknesses,prerequisites,andstructural limitationsofeachof thegenericsoftwarebusinessmodels. It is also important here to understand that while transitionsbetween these generic

    softwarebusiness models are possible, they are not necessary or automatic. None of these

    business models is essentially good orbad, they arejust different and one must pick the

    particularmodelthatbestsuitshis/herideaofferingdestinationmix.

    Weconcludethestudywithabriefreviewonenvironmentalandpolicybottlenecksthathave

    hindered the growth and development of the software industry. This is,by no means, an

  • 8/4/2019 Best Practices Study

    10/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 10

    exhaustivestudyorevenacomprehensive listofpolicy issuesbutratheradescriptionofour

    statisticalandqualitativefindings.Thecountrysimage,overandabovethecompanysbrand,

    topsthelistastheproblemidentifiedbyasmanyas68%ofallrespondents.Thisisfollowedby

    qualityofmanpower(56%),thecostofIT/Telecominfrastructure(50%)andlawandorderand

    security situation (48%) as the most important problems from the perspective of alltypes of

    firmscombined.Whiletherearevariationsbetweenhoweachofthesemaydisproportionately

    affectvarioussubcategoriesoforganizations,image,IT/Telecominfrastructure,andHRappear

    to rate consistently as among the top5 problems in all categories. We also faithfully narrate

    severalproposals,putforthbyourinterviewees,toaddresssomeoftheseissues.

    Onthewhole,thereareafewgeneralizedconclusionsthatonecandraw.Thefirstandforemost

    contributionofthisstudyistobringforththeveryvibrantfaceofPakistanssoftwareindustry.

    Pakistantoday,unlikeyesteryears,isfastturningintoahappeningplaceforIT. Althoughthe

    industryhascomealongwaysinceitsfirstcompanyopenedshopin1976,ithasonlybeenin

    thelimelightforinvestorsandpolicymakersalikesincetheearly1990s.TenyearsisaveryshorttimeforthedevelopmentofanentireindustryandtherearesignsthatPakistanssoftware

    industry, having laid the foundations for a tomorrow, maybe in forbetter times ahead. Last

    yearalone, the industryhasgrownataround37% inrevenuesand27% in termsof technical

    and professional employment. Many of the CEOs we spoke to expect abetterthanlastyear

    performance in 2005. Another encouraging sign is the increasing number of Pakistaniowned

    foreign firmsbeing located to Pakistan as well as the reversebrain drainbeing causedby

    returningPakistanientrepreneurswhoseetherelativelylesscompetitiveandvirginmarketat

    home as a tremendous opportunity for setting up a Pakistanbased company. Systems

    Integration, Innovation and Intelligence (SI3) and The Resource Group (TRG) are the posterchildrenofthisundeniabletrend.Noneofthesewouldhavebeenpossibleadecadeago.

    On the domesticfront as well, there is a growing likelihood of considerable opening up and

    modernization of traditionally conservative segments of the economy. If deregulation in the

    financialsector isanycrediblesignofthingstocome,weare likelytoseemassivechanges in

    the shape of the local manufacturing and service industries by virtue of telecom sector

    deregulation and the enhanced competition under the noweffective WTO trade regime. The

    former has already begun to show tremendous promise with around a billion dollars of

    promised

    investment

    in

    last

    year

    alone.

    An

    investor

    whom

    we

    spoke

    to

    sees

    the

    situation

    as

    the

    fading away of the Old Pakistan and the Emergence of the New Pakistan that is effectively

    linked toand a significantplayer of the global economic system.The New Pakistan presents

    considerablepromiseandopportunitytothosewillingtobiteat it.Thereareliveexamplesof

    companiesTRG, SI3, LMKR, Netsol, Techlogix, Etilize, TPS and many morethat have

    capitalizedonthisnewsetofopportunitiesandpositionedthemselvestoreaptherewards.

  • 8/4/2019 Best Practices Study

    11/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 11

    There are, however, considerable, although not insurmountable, challenges too. The industry

    suffers from a serious professionalization and institutionalization deficit. The 200people

    barrier,althoughpsychological, isrealtill it isactuallybrokenandbrokenconvincinglyand

    forever. Inaddition to the200peoplebarrier,wealso facea20peopleanda2peoplebarrier

    thatrequires

    as

    much

    attention

    as

    the

    former.

    Many

    of

    our

    very

    innovative

    firms

    continue

    to

    resistprofessionalizationandthusfailtogrowbeyondaparticularsize.Theindustryishungry

    for capable investors/acquirers to come forth andbring about paradigm shifting structural

    changes to thesecompaniesandenable themtomoveto thenexthigher levelofgrowth.The

    fastmaturingmarketofoutsourcingandoffshoringservicesnecessitatethatourentrepreneurs

    and business leaders think about new ways of doing things. It is unlikely, given the

    consolidation in theoutsourcing industry, thatwewouldseeanewplayerreplacingWipros,

    Infosys, or TCS of this world. Rather than blindly copying the already wellestablished

    countries and players, we must think creatively to devise a model thatbest suits our own

    strengths

    and

    weaknesses.

    Our

    ability

    to

    lead

    in

    the

    business

    model

    innovation

    would

    determine,toalargeextent,ourplaceinthefuturepeckingorderofsoftwareexportingnations.

    Playingthevolumesgame(ITES/BPO),withouttherequisitescalabilityandHR,isunlikelyto

    succeedonanindustrywidescale.Untilwecanresolvethescalabilityissue,wemustlearnto

    playintheequallylucrativeideasgame.

    InadynamicandfastchangingindustrylikeIT/Software,tomorrowcanandwillberadically

    different,andnotmerelyanextensionoftoday.Itwouldrequireinvestorsforesight,business

    managers insight, and entrepreneurs courage to capture the moment and build the next

    generationofnicheplayersand industry leadersandbuild it in theNewPakistan.Profitsare

    certainly to be earned by those who break the rules and try the unthinkable. There is,

    however, a dire need to think deep and hard about the problems, patterns, and strategic

    challenges identified in this report, find explanations for these, and devise strategies to get

    aroundthem.

    2.BACKGROUND&INTRODUCTION

    Pakistanssoftware/IT industryhasshownanunevenpatternofgrowth through itsrelatively

    short history. While Information technology and software industries were not a government

    prioritybefore early nineties, software houses have existed in the country since 1970s. From

    earlytomid 1990s, however, promoting the software/IT industry hasbeen a stated, if not

    alwaysadheredto,governmentpriorityafactmotivatedpartlybyIndiasrisetoprominence

    asamini(software)superpower.Severalpolicyactionsandinfrastructuredevelopmentand

    upgradationprojectshavebeenundertakenbyGovernmentofPakistan(GOP)topromotenot

    onlyadomesticsoftware/ITindustrybutalsoexportsofsoftwarefromPakistan.Manyofthese

  • 8/4/2019 Best Practices Study

    12/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 12

    aredocumentedintheNationalITPolicyanditsaccompanyingActionPlan(MOST,2000).The

    progressontheseactionsandinitiativeshas,however,beensketchy(UNCTAD,2004).Thelocal

    softwarescenedoesnotyetshowthekindofvitalityandgrowththatisacharacteristicofmajor

    tier1oreventier2softwareexportingnationasdescribedinCarmel(2003).

    2.1BackgroundandMotivationfortheStudy

    WhilethecausesofPakistansbelowparperformanceinthesoftwaresectormaybemany,the

    importanceofwithinindustrylearningandanorganicgrowthcannotbeoverlooked.Pakistans

    softwareindustry(anditsancillaryandrelatedindustriese.g.banking,venturecapitaletc.)isin

    direneedofsharingofbestpractices,ofitsownindustryiconsandheroes,andofalotofhope,

    optimismandthefocustosucceed.Itneedsanindepthunderstandingofthecurrentstateofits

    affairs,beyond thegeneralrhetoric,andavisionof the future tomotivate it toupgrade itself

    and capture its due share in the world software/IT market. A formal research study ofbestpracticesandstrategicandcompetitivedriversofthePakistanisoftwaresectorhaslongbeenin

    order. The proposed study would develop a shared understanding of the problems and the

    promise of the Pakistani software sector andbuild a coalition of support around this shared

    reality.Itwouldalsoserveasanauthenticsourceofdataandinformationtoquicklyupgrade

    theunderstandingofpotentialinvestorsintendingtoinvestinthelocalsoftwarescene.Finally,

    andmostimportantly,itwouldhelptheindustryitselfinlearningfromeachotherssuccesses

    andfailures.

    Several factors are widelybelieved tobea hindrance in the countrys aspiration tobecome a

    significant software exporter,not the least important of which are macro and geopolitical in

    nature (e.g. law and order and security situation, image of the country etc.). While resolving

    these issues is critical to developing a strong and robust industry, this study adopts a

    differentinsideoutapproach that asks the question: What can the various players,

    essentially software companies, in the industry learn from eachother? In essence, we are

    attemptingtolearnfromthevariationsinperformanceofcompaniesoperatingunderthesame

    set of geopolitical and policy environment. Secondly, there is a growing realization that we

    must truly understand the structure of the Pakistani software industry and the nature ofPakistanscompetitiveadvantageinthesoftwarearenainordertodevisebetterindustrialand

    organizationalstrategiesandpublicpolicyinterventions.Afirmlevelanalysishasthepotential

    tounearththefactorsbehindwithinindustryperformancedifferentials(e.g.Netsolvs.Cressoft

    vs. Enabling Technologies) and identify best practices that can be adopted industrywide.

    Regardlessofwhatthefinalconclusionmaybe,theonethingthatiscertainaboutthePakistani

  • 8/4/2019 Best Practices Study

    13/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 13

    software industry is that it is not a very well understood and researched one. For example,

    questionslike:

    Why hasnt the Pakistani software industrybeen able to produce a single worldclasssoftwarefirm(e.g.Wipro,InfosysorTCSofIndia)inthelast1015years?

    Why havent webeen able to grow Pakistani software exportsbeyond a certain level($3060millionperannum)forthelast5years?

    DoesPakistanisoftware industrymerelyrepresenta lower levelofdevelopmentoranaltogetherdifferentdevelopmenttrajectoryascomparedtoknownpeernations?

    What constitutes a generalized set ofbest practices in the local software industry (i.e.whatdifferentiatesbetterperformersfromthosethatdontperformthatwell)?

    Answeringthese(andother)questionswouldrequireconsiderableindustryresearch,sharingof

    bestpractices,anddiscussion/debate.Theultimateanswertothesequestionsismostsurelynotgoingtobeasilverbulleteitherbutaformalinquiryhasthepotentialtosetinmotionaprocess

    thatmightgiveussomehints towardsapossibleanswerorenableus toaskmore intelligent

    questionsandthusleadusnearertothetruth.

    2.2IntroductoryReviewoftheRelevantLiterature

    There hasbeen considerable increase in the interest in software industries within developing

    country contexts in the recent years. Proponents of the school of thought that sees IT and

    softwareasagreatenablerhavearguedthatinformationtechnologyingeneral,andsoftware

    industryinparticular,providesanopportunitytothedevelopingcountriestoinextricablylink

    themselves with the developed economies of the west. This globalization of work (or

    production), somebelieve, is a harbinger of subsequent phases of globalization that would

    reduce the disparities across the world and provide an equal opportunity for everybody to

    participateintheglobalproductionandcreativeprocesses.Inmanyinstances,thesepredictions

    havealsobeenvalidatedby initialexperiences insomedevelopingcountries.Mostnotableof

    theseareIndia,IrelandandIsrael,famouslyknownas thethreeIsoftheglobalITrevolution

    andthenewentrantsinthetier1ofsoftwareexportingnationsthatalreadyincludesrelativelymoredeveloped,mostly,OECDcountriesand,andtoalesserdegree,ChinaandRussia(tier2

    countries).Followingtheexamplesofthesetier1and2nations,areahostofotherdeveloping

    countries,namely, Brazil, Mexico, Malaysia, SriLanka,Pakistan, Ukraine,Bulgaria, Hungary,

    Poland and the Philippines (tier3 countries) and Cuba, Iran,Jordan, Egypt, Indonesia and

    Bangladesh(tier4countries)andmanyothers(Carmel,2003).

  • 8/4/2019 Best Practices Study

    14/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 14

    Whiletheboundariesbetweenthecountriesinthis4tieredtaxonomyarequitefuzzy,primarily

    bydesignbutalsoduetolackofcredibledataoneach,Carmel(2003)attemptstodifferentiate

    tier1 countries as having hundreds of companies, more than a billiondollars of export

    revenues,andtheindustrymaturityofmorethan15years;tier2countriesashavingatleasta

    hundredcompanies,exportsrevenuesofmorethan$200million,andgreaterthan10yearsof

    industrymaturity;andtier3countriesashaving tensofcompanies,morethan$25million in

    exportrevenues,andover5yearsofindustrymaturity.Allotheraspirantsthatdonotmake

    thecutfallinthetier4ofthetaxonomy.

    Manyresearchersandanalystshave tried tounderstand thedynamicsof the Indiansoftware

    industry (NASSCOM,2001,2002,2003,2004;Heeksetal,1996,1998,2002;BajpaiandShastri,

    1998;Desai,undated;Aroraetal.,2000).Software industriesofothercountriessuchasChina

    (Tschang and Xue, 2003), Japan (Rapp, 1996), Iran (Nicholson and Sahay, 2003), Romania

    (GrundeyandHeeks,1998),SriLanka(BarrandTessler,2002),Korea(BarrandTessler,2002)

    and Malaysia (Mohan et al., 2004), among others, have alsobeen documented in literature.

    Severalresearchershaveattemptedtotakethisknowledgeandapplyittothecontextofother

    countries(UNCTAD,2002,Tessleretal.,2003).Othershavetriedtodeveloppolicyframeworks

    and draw policy conclusions (Carmel, 2003b, Heeks and Nicholson, 2002) or develop generic

    analyticframeworksforanalyzingthecompetitivenessofsoftwareindustries(Heeks,1999;and

    Bhatnagar,1997).Heeks(1999)describesa2x2theoreticalframework(describedinsection5.2)

    thatclassifiessoftwarecompaniesonthebasisoftheirdestination(domesticorexport)andtype

    ofoffering(productorservice).Bhatnagar(1997),takingadifferentapproach,describesnations

    asgoing through fourstagesofmaturity transitioning frombuildingskillsandreputation, to

    buildingservices,tobuildingproducts.

    Heeks(1999)analyticframeworkisinterestingandusefulandroughlyformsthebasisofthis

    reports analytic framework. The four resultant categories of companies from Heeks 2x2

    frameworksaredifferentintermsoftheirorganizationalcharacteristics,competitivestrategies,

    and enabling conditions and requirements. While it is clear where most companies from

    developing countries would like tobe (i.e. exporting products and services), Heeks (1999)

    arguesthatgettingthereisnotallthateasy.Veryfewcompanieshavebeenabletosuccessfully

    executeonstrategiesdictatedbytheneedsofeachofthesefourquadrantsandHeeks(1999)

    claims that majority of what we see is a constrained kind of an optimizationhe calls them

    survivalstrategiesratherthanafreeplaywithinthesecategories.Drawinguponanearlier

    paper (Heeks, 1998) it also presents secondary and anecdotal evidence to support his

    conclusions.

  • 8/4/2019 Best Practices Study

    15/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 15

    Thatthemuchtoutedsuccessofthesoftwareminisuperpowersmaynotbeasconvincingas

    it is portrayed canbegleaned from the following facts.Firstly, developingcountry packaged

    softwareexportsthe24caratgoldofthesoftwareexportsbusinessareminimalinthe5

    10% range fromeven thebest of thesoftwareexporters like India, with the soleexception of

    Ireland

    and

    perhaps

    to

    a

    lesser

    degree,

    Israel.

    Secondly,

    majority

    of

    the

    work

    doneby

    the

    developing countries consist of lowskilled programming or coding services and while some

    countries,notablyIndia,mighthavedonewellinthistypeofactivity,itseriouslysuffersfrom

    issues of valueaddition and scalability. Thirdly, majority of the workbeing performedby

    developingcountriesislocatedinrelativelyfewconcentratedenclavesofsoftwaredevelopment

    activity worldwide (e.g. IndiasBangalore),beingperformedby foreigntrainedprogrammers

    workinginsubsidiariesofforeigncompanieswhospendamajorportionoftherevenuesonsite

    (in thecountry of their clients) to pay for the travel and living expensesof their consultants,

    leavingmuchtodesiredintermsofvaluegainedbythedevelopingcountryitself.Heeks(1999)

    describes

    major

    challenges

    (orbottlenecks)

    that

    a

    firm

    may

    encounter

    in

    each

    of

    these

    four

    productmarketcategoriesanddescribesthereasonsofthetypeofperformanceweseeineach

    ofthesecategories.

    Stillotherresearchershavetakenamulticountryviewofsoftwareindustries.Rubin(2000)isan

    interesting, thoughdated,overviewofglobalsoftwareeconomics (Pakistan isnot includedas

    one of the countries surveyed). It presents data on several interesting variables (e.g. labor

    productivity,sizeofsoftwarestaff,sizeofportfolio,costperdeliveredanddocumentedlineof

    code,costpersupportedlineofcode,averagesalariesofdevelopersandmaintenancestaff,and

    defects per 1000 lines of code etc.) for a large number of countries. Coward (2003) takes an

    outsourcers view of the software industry looking at the 14 factors that influence the

    decisions of American SMEs to outsource software development activity to developing

    countries. Cusumanoet.al.(2003)isareviewofglobalsoftwaredevelopmentpractices.Based

    onastudysampleof104projects,itcomparesthesoftwaredevelopmentpracticesofAmerican,

    European,Japanese,andIndiancompanies.

    Thisstudyfindsthatconventionalsoftwareengineeringpractices(e.g.functionalspecs,design

    reviews,codereviewsetc.)arepopular in India,Japan,andEuropebutnot theUnitedStates

    wheretheyareusedless,acrosstheboard.ItidentifiesIndiancompaniesasespeciallyadeptin

    mixing theseconventionalapproacheswith therelativelynewerapproaches likedailybuilds,testerdeveloperpairs,andpairedprogrammingtechniques.Overall,thereportfindsJapanese

    andEuropeansoftwareoperationstobemostproductive(intermsoflinesofcodeperaverage

    staff*calendar) followedby US and Indian operations.Japanese projects also produced the

    lowest number of defects, followed closelyby Indian and US projects, and the Europeans

    finishinglastonthismetric.

  • 8/4/2019 Best Practices Study

    16/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 16

    Thisstudyconfirmssimilarfindingsbyotherresearchersthatdescribethetechnicalqualityof

    software developmentprocesses employedby Indian software companies (Duttaand Sekhar,

    2004) and the adoption of standardized quality practices like Six Sigma methodologies

    (Radhakrishnan, 2004) and CMM certifications. These geographical differences in software

    development

    practices,

    however,

    maybe

    attributed

    toboth

    cultural

    and

    type

    of

    work

    related

    factors.Forexample,Cusumanoetal.(2003)observethatIndiaandJapansignificantlylagthe

    AmericanandEuropeansoftwareoperations intermsofthe innovativequalityof theirwork.

    Dutta et al. (1997) finds similar acrosscountry differences within 16 different European

    countries.

    Collectively, thisconstitutesawealthof informationabout thedevelopmentandevolutionof

    softwaredevelopmentactivityindevelopingcountrycontextsfrommultipleperspectives.They

    point towardsanumberof factors,environmental,policyanalytic (e.g.CarmelsOvalModel,

    Heeks National Export Success Model) and organizational (e.g. Cusumano et al., 2003, and

    Cusumano,2004)andidentifymajorbottlenecksthatmightaffecttheexecutionofaparticular

    strategy (e.g. Heeks, 1999). While development planners seek to extract prescriptions, this

    collectivebody of literature falls short of doing so hinting instead at the idiosyncratic factors

    andearlymoveradvantagesthatmightdistinguishsomecountriesprogressfromtherest.

    The overall picture that emerges from various models and frameworks is a complex one. It

    underscores the importance of understanding a large number of policy, environmental, and

    organizational factors, and how they interact with each other, as well as the individualistic

    featuresofeachof thecountriesand their targetmarketsbeforeapolicyoran industrywide

    prescriptioncan

    be

    made.

    Every

    country

    that

    we

    looked

    at

    (e.g.

    India,

    China,

    Japan,

    Ireland,

    Israel etc.) is different from every other country and understanding these unique features is

    importantbeforeany lessonscanbedrawnandapplied fromothercontexts.We takeup this

    challengeinthisreportonPakistanssoftwareindustry.

    3.THEOBJECTIVES,AUDIENCE,ANDFORMATOFTHESTUDY

    TheBestPracticesinPakistaniSoftwareSectorProjectbeingthefirstofitskindandscopein

    Pakistanis an exploratory study of the Pakistani Software Industry. Not only is the whole

    subjectoftheformationanddynamicsofsoftwareindustryaroundtheworld,andespeciallyin

    developingcountries,relativelynewandhenceunderstudied,thePakistanisoftwareindustry

    isa totallyuncharted territoryas faras thestructure,managementpractices, technicalability,

    andtheindustrydynamicsareconcerned.

    3.1TheAnalyticAgenda:

    Thisstudyhasbeenundertakenwithatwoprongedanalyticagenda,namely:

  • 8/4/2019 Best Practices Study

    17/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 17

    Atthemostbasic level, thestudyattempts tocollectqualitative (butalso,wheneverpossible

    within thepurviewof theresearch,quantitative) informationon thecurrentstateofsoftware

    industryinPakistanwithanemphasisonfirmlevelcharacteristicsandcompetitivedynamics.

    This would help in identifying the various organizational success factors, develop a shared

    understanding

    around

    those,

    and

    enable

    stakeholders

    to

    derive

    strategic

    and

    policy

    prescriptions from these. It explores the importance and prevalence of the various structural

    constructs in the Pakistani software industry and documents perceptions ofbusiness leaders,

    entrepreneurs,andinfluentialindividualsintheindustrytowardseachoftheseconstructs.The

    studyattemptstodoaoneleveldeeperanalysisofwhyindividualsholdacertainperceptionto

    move the level of debate within the industry to the next higher level (i.e. from identifying

    problemstoidentifyingsolutions).Forexample,ifwehearalternativeexplanationsoflackofa

    culture of entrepreneurship, we would like to explore why and on what factors are those

    perceptionsbaseduponand,totheextentpossible,corroboratethatwithgroundreality.

    Atthemoreadvancedlevel,thestudyattemptstoestablishbestpracticeswithinthePakistani

    softwaresector.Thisisaproblemriddledwithcontroversies,nottheleastimportantofwhichis

    the identification of highperformers in the absence of credible performance data. Additional

    issueshavetodealwithdefinitional(i.e.whatisabestpractice?)andmaturity(i.e.when

    doesapracticebecomeabestpractice)problems2.Thestudytriestotacklethiscontroversial

    subject in a number of ways. Firstly, we try to identify the relatively more successful and

    prominent software companies in Pakistan and compare their various organizational,

    structural, and process features against several others that have not been as successful.

    Although it is likely that the differencesbetween thebest and the notsogood performers

    may not turn out tobe substantive enough (or worse yet, they may turn out tobe quite

    obvious), the results of the study would, nonetheless, form a documentedbaseline against

    which changing trends in the Pakistani software industry maybe compared in the future or

    againstthatofothercountries(e.g.India).

    To the extent that a (semi) statistical/quantitative analysis is likely tobe of limited utility, a

    qualitative/anecdotalapproachmaystillbeoftremendousvalueinidentifyinganddeveloping

    a shared understanding of best and unique practices (and whats possible) within the

    software sector in Pakistan. Similarly, a valid criticism of our approach maybe that in a

    relativelynascentandimmatureindustrylikeours,asinglecompanymaynotrepresentallthedesirablebestpracticefeatures.Weuseaqualitativeapproachtoidentifyandcherrypick

    specific innovative and successful features of the software development and marketing

    2Accordingtoonelongtimeindustryobserver,itmightbedifficulttoidentifybestpracticesinthe

    relativelynascentPakistanisoftwareindustry,whatonemightgetinreturnforthequestfortheformerwouldbea

    lotofworstpractices.

  • 8/4/2019 Best Practices Study

    18/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 18

    processes (e.g. partnering and alliance building, customer acquisition, and product

    developmentstrategyetc.)todevelopalaundrylistofbestpracticesthattherestoftheindustry

    canemulate.Whileourprimaryfocusismanagerialbestpractices,wedobrieflytouchuponthe

    issueoftechnicalpracticesinthepassing.Thisisdonefortheprimaryreasonthatthereexists

    an

    interplay

    and

    dependence

    between

    the

    latter

    and

    the

    former.

    We

    do

    not,

    however,

    attempt

    an

    exhaustiveanalysisofthetechnicalpracticesoforganizationsbeingstudied.

    3.2TheBenefitsandIntendedAudience:

    The primary purpose of undertaking this study is that ofwithinindustry learning with the

    secondarypurposebeinginvestmentpromotionandfacilitation.Thebenefitsof(andintended

    audiencefor)theaboveanalysiswould,therefore,bethreefold:

    Firstly, the findings of the study would be of considerable value for the existingsoftware entrepreneurs, executives,andmanagers seeking to learnfrom the collective

    experience of their compatriots. This learning could take the form of: What are the

    criticalsuccessfactors,theDosandDonts,sotospeak,ofrunningasoftwarebusiness

    in Pakistan? The industry managers wouldbe able to gauge the performance of their

    companies against the bestinclass companies and derive recommendations for

    correctingcourse,ifnecessary.

    Secondly, the study would also inform the interested (yet skeptic, at times) bystanderspotential entrepreneurs, interested businessmen and managers, and

    investorscontemplatingstartingasoftwareventureand looking foragoodsenseof

    what we can learn from the experiences of tens of successful and notsosuccessfulentrepreneurs.Itwouldalsohelpinspireandilluminatethedecisionsofavastnumber

    of stakeholders, namely, business leaders, industrialists, managers, financiers and

    investors, regulators, policymakers etc, whose decisions to engage or disengage with

    thisnascentsectoroftheeconomycanmeanthedifferenceforthesoftwareindustry.

    Thirdly,thestudybeingthefirstofitskindinPakistancouldbeofpotentialvalueforforeign investors,clients,andpolicymakerswhoseappetiteformeaningfulquality

    informationon thesubjectgoesunsatisfiedforwantof credibleanalysisdoneon the

    subject.Tothateffect,thisstudymayprovideacredibledatabenchmark(orreference

    point)forputtingPakistanssoftware industry in largerglobalperspectiveandgetting

    themessageacrosstopotentialinvestors,clients,andpolicymakers.

    3.3TheFormatoftheStudy:

    Thestudycanbebroadlydividedintotwoparts.Thefirstpartcoversastatisticalsnapshotof

    the industry as gleanedby data on ourrespondents. The second part combines this with the

    morequalitativeinformationtodiscussstrategicchallengesandgoodpracticesintheindustry.

  • 8/4/2019 Best Practices Study

    19/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 19

    The study is formatted as follows: Section3 provides some background that builds the

    motivationforthestudy.Section4dealswiththeobjectives,audience,andformatofthestudy.

    Section5briefly describes the project methodology in a narrative and a graphical fashion.

    Section6startswith theresultsof thesurveyandattempts tobuildastatisticalprofileof the

    Pakistanisoftwareindustryasgleanedfromanonthespotsurveyof60ofitsmajorplayers.

    This section is divided into 4 major parts. The starting part sets the context of this statistical

    analysisbydiscussingresultsfromaverylimitednumberofearlierstudies.Thenwediscussa

    basicstatisticalsnapshotoftheindustryusingexportfocusedanddomesticfocusedfirmsasa

    basis for classification. Next we discuss various other classifications (e.g. productfocused vs.

    servicesfocused, small vs. large, preDotCom vs. postDotCom, and developmentcenters vs.

    rest of the industry) to assess how these varying organizational factors affect the managerial

    and technical processes of software companies in Pakistan. Finally, we assess whether the

    industrystatisticsrevealapatternofbestpractices?Inessence,weusethestatisticaldatato

    answerthe

    question:

    How

    do

    better

    performing

    firms

    differ

    from

    the

    rest

    of

    the

    industry?

    Section7supplements thiswith informationgained fromaround65qualitative interviews. It

    usestaxonomyofgenericsoftwarebusinessmodelsinPakistantoidentifygenericprofilesand

    strategicandcompetitivechallenges facedbysoftwarecompanies inPakistan.We identify13

    suchchallenges,dividedacross4genericcategoriesofsoftwarebusinessmodels,anddiscuss

    ways in which our respondents have innovatively tried to address each of these. There are

    lessonstobelearnthereforthesoftwareentrepreneursandbusinessmen,bothyoungandold

    that could be applied and replicated across the industry. Section8 briefly touches upon

    environmental,infrastructure,andpolicybottlenecksconfrontingthesoftwareindustry.Finally,

    Section9discussessometentativeconclusionsandrecommendations.

    Thisreportcanbereadinitsentiretyorselectivelydependinguponwhatareaderisspecifically

    lookingfor.Initsentirety,wehavetriedtostructurethereportinamannerthatcouldgivethe

    readeracomprehensiveviewofPakistanssoftwareindustry,itscurrentstate,itspeculiarities,

    and themajorchallenges facedby thesoftwarecommunity.Onecanalso,however,pickand

    choose what specific sections to read. For example, the generic profiles of different types of

    softwarebusinessmodelsandthechallengesspecifictoeachcanbereadwithoutreferenceto

    the rest of the report. Either way we hope the report would present considerable original

    informationandgeneratesomethoughtandreflectionamongitsreaders.

    4.ABRIEFNOTEONPROJECTMETHODOLOGY

    Inordertomeetbothqualitativeandquantitativerequirementsofthestudy,weadoptedatwo

    prongedapproachtotheproject,comprisinganonthespotstatisticalsurveyandqualitative

    interviewswithtoporganizationalexecutivesofmajorsoftwarecompaniesinPakistan.Owing

    to the relatively short timeline of the project, a convenience sample of software houses (or

  • 8/4/2019 Best Practices Study

    20/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 20

    softwaredevelopmentoperations)wasselectedandcontactedtobecomepartofthestudy.An

    effort wasmade, however, to includekey largeandprominentplayersof the industry in the

    analysis.Foursourcesofinputwereutilizedforthispurpose.PSEBandPASHAofficialswere

    contactedtoidentify,fromamongsttheirmembercompanies,thelargest,mostprominent,and

    mostsignificant

    software

    operations.

    The

    consulting

    team

    also

    utilized

    its

    own

    knowledge

    of

    the localsoftware industrytoaddto this listofnominations.Finally,severalcompanieswere

    addedtothelistonanongoingbasisasnamesofcompaniesdoinginnovativeandinteresting

    workcameupduringinterviewswithindustryprofessionals.

    In all, 22 companies in Karachi, and 13 each in Islamabad and Lahore (for a total of 47

    companies) were personally visited and surveyed. 13 more companies were added to the

    statistical sample through the Online Survey of Best Practices in the Pakistani Software

    Industry3.Thisincreasedthetotalnumberofsurveyrespondentsto60.40ofthese60companies

    (or 2/3rd of the total) were identified and hence categorized as the more prominent and

    relativelysuccessfulsoftwareoperationsinPakistan.Thisenabledustodeveloptworeference

    groupsandallowedthepossibilityofstatisticalcomparisonsbetweenthesetwogroupswitha

    viewtoidentifyingdifferencesbetweentheminvariousmanagerialandtechnicaldimensions.

    3 The PSEB Best Practices Online Survey is available at: http://www.hostedsurvey.com/takesurvey.asp?c=PSEB

  • 8/4/2019 Best Practices Study

    21/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 21

    Perceived Policy

    Problem &

    Opportunity

    Problem

    Definition

    DataCollectiononContext & Background

    Policy &

    Research

    Questions (RQ)

    Preliminary Literature

    Review (LitR1)

    1.Research

    Questions(RQ)

    2. Survey

    Parameters(sample

    4.

    Instrument

    Testin

    3. Surv

    Instru

    6. Ana

    & Res

    5. Survey

    Administra

    -ion

    3.

    Admin

    Interv

    2. Identify

    Sample /

    Participant

    1.Thematic

    Areas for

    Interviews

    R&DPERFORMANCEMAILSURVEY

    COMPANYINTERVIEWS

    ONGOINGLITERATUREREVIEW(LITR2)

    Figure-I: The Multi-Pronged Research Methodology

  • 8/4/2019 Best Practices Study

    22/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 22

    Severalothercomparisongroupswerealsocreatedtohighlightdifferences inmanagerialand

    technical practices. Throughout the following analysis, where appropriate, we invoke the

    differences between various categorizations (e.g. exportfocused vs. domesticfocused vs.

    developmentcenters,betterperformersvs.rest,productfocusedvs.servicesfocused,andlarge

    vs.small

    software

    houses)

    to

    make

    the

    results

    more

    meaningful

    to

    the

    software

    community.

    FigureI(below)presentsagraphicalsnapshotoftheprojectmethodology.Nextwelookat

    theresultsoftheanalysis.

    5.ASTATISTICALSNAPSHOTOFPAKISTANSSOFTWAREINDUSTRY

    Adiscussionofthesize,structure,anddynamicsofPakistanssoftwareindustrymustbeginby

    setting an appropriate reference for the same. This reference can either come from within

    Pakistan(i.e.comparingthecurrentindustrywithitsstateatsomepointinthepast)oroutside

    Pakistan(i.e.comparingitwiththestateofsoftwareindustryofacomparablecountry).There

    arepotentialproblemswithboththeseapproaches.Fortheformer,barringahandfulofreports,

    welackcomprehensiveandcredibledataofanykind,whatsoever,tosayanythingmeaningful

    abouttheindustryatdifferentinstancesintime.Whileforthelatter,onefacestheproblemof

    finding an appropriate country to make comparisons with. Most often, for reasons of

    prominenceandtradition,theexampleofIndiaisinvokedwhenanalyzingPakistanssoftware

    industrya practice that, although may have some value, can at times be quite

    counterproductiveorleadtowrongpolicyprescriptions4.Wewilldiscusseachofthesepoints

    ofreferenceingreaterdetailbelow.

    5.1EstablishingaPointofReferenceforPakistansSoftwareIndustry

    Looking forpointsofreferencerelevant to thePakistanisoftware industry,wecould identify

    onlyahandfulofstudies/documentsofvaryingcredibilityfromthepast.Theseinclude:A1999

    2000 CSPSEARCC5 ICT Manpower and Skills Survey; a 2002 PASHALUMS Study of

    PakistansSoftware/ITIndustry,a2004UNCTADStudy,anda2004EAC6StudyofPakistans

    ITIndustry. Eachofthesestudiesisfairlylimitedintermsofthescopeandcoverageofpolicy

    4 This has been a case quite a few times in past, for example, the Government of Pakistans $1 Bn. Software ExportTarget by FY2000 was motivated in part by using the Indian software export figure and appropriately discounting itto a smaller value rather than any credible assessment of the industrys present or future capability.5 This study was conducted by the Computer Society of Pakistan (CSP) in collaboration with South East AsiaRegional Computer Confederation (SEARCC) and used methodology and instruments that were used among 14countries of South-East Asia.6 Experts Advisory Cell (EAC) is housed within Ministry of Industries, Government of Pakistan.

  • 8/4/2019 Best Practices Study

    23/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 23

    andorganizational(technicalandmanagerial)issues.Thesestudies,likeanyotherstudyofthis

    nature, also have a fair number of methodological issues and problems. For example, CSP

    SEARCC&PASHALUMSstudiesarequitedated.Whiletheformerdoesfairlywellasfaras

    being representative of the industry and providing a good reference point for crosscountry

    comparisons, it isfairlylimited in itsscope(i.e.onlydealswithmanpower issues).The latter,however,whilebeingmuchbroaderinscopedoesfairlypoorlyonrepresentativeness7.

    The latter two studies (i.e. UNCTAD, 2004 and EAC, 2004) are focused more on the policy

    environmentandlessonorganizationalissues.Theformermakesanattempttoimposepolicy

    prescriptions from other countries without adequately demonstrating an understanding the

    7 The maximum sample size in PASHA LUMS (2002) is 16 with very strong statistical generalizations made, attimes, with as little as 7 observations, without any mentioning of potential non-response biases.

    TEXTBOX#1: SALIENTFINDINGS&METHODOLOGIESOFPRIORSTUDIES

    CSPSEARCCStudyofICTManpower(2000):314of441organizationsresponded(71%

    responserate)ofwhich40.8%wereITsuppliers,14.5%publicsector,and44.7%privatesectorend

    users.2375of5000ITprofessionalsresponded(46%responserate)ofwhich60.3%workedin

    developmentand39.7%inservices.Somesalientfindingsare:

    51.3%ITprofessionalsworkedinsoftwaredevelopmentwhile6.3%inITMgmt. ITprofessionalsagedbetween2529(33%),2024(23%),and3034(19%) Male:Femaleratiois9:1,withroughlyproportionalrepresentationinjobsincl.ITmgmt. Salarylevels:PKR25M)employeddoubletheQAprofessionalsthansmallerona%basis.

    UNCTADStudy(2004):ComprisesreviewofsecondaryliteratureinthePakistaniand

    internationalcontexts.Salientfindingsofthestudy,generallycriticaloftheindustry,are:

    Discernableactionononly18ofthe162(11%)commitmentsofNationalITPolicy Pakistan76thof102countriesinNetworkReadinessIndex ActualspendingunderITPolicy2000lagsallocations,esp.inExports/eCommerce RevenuesinExport:$12.2M(growthof84%)andDomestic:$5M(growthof49%) Currentestimateofsoftwareexportsatabout$12M

  • 8/4/2019 Best Practices Study

    24/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 24

    peculiardynamicsofthe localsoftware industry.Thelatterprovidesa lotofdata,which isat

    best,sketchy,andaggregatesa lotofmundanesecondary information inanunimaginative

    fashion. These weaknesses notwithstanding, these studies provide a starting point for a

    discussionon thecurrentstateofPakistanssoftware industry.TextBox#1 (above)provides

    someofthesalientfeaturesandfindingsofthesereports.

    Moreimportantly,however,thesereportsprovideanimpetusandamotivationtoundertakea

    moreextensiveontheground(handson)analysisofthePakistanisoftwaresector.

    5.2SoftwareDevelopmentinPakistan:StatisticsonManagerialandTechnicalPatterns

    PakistansSoftwareIndustryhascomealongwayfromitsstartin1976whenacompanyby

    thenameofSystemsPvt.Ltd.openeditsofficesinLahore.Overthelastthreedecadesorso,the

    industryhasgrown from zero toan approximatesizeof wellover a hundred milliondollars

    and employs thousands of professionals8. During this time, the industry has seen periods of

    nascence, hope, euphoria, disillusionment, renewal, and rebuilding. The last decade has in

    particularnotonlybeenatimeofgreatpromise,butalsoatestfortheindustrythathasbeen

    throughafullcycleofreversalsfromaninsideout(domesticfirst,exportlater)toanoutside

    in(exportfirst,domesticlater)worldviewandbackagain.Intheprocess,itprobablyhasalso

    been through considerable maturation, not only in terms of its ability to develop good

    innovativesoftwarebutalsobuildsuccessfulbusinesses.Wefindconsiderableevidenceofthe

    fact that the countrys financial communitythe business houses, investors, and business

    managersare learning how to manage the IT and the IT professionals are learning how to

    manage thebusinesspartsof the ITbusiness.The industry,however,hasa longway togo

    beforeitcantrulyrealizeitspotential.Thestatisticalpicturethatwepresentbelow,therefore,is

    asnapshot,ataparticularpointintime,ofwhatessentiallyisamovingtarget.

    Beforewediscussthestatisticalresults,however,adisclaimerisinorder.Thestudyinquestion

    onlylooksattherelativelywellknown50oddsoftwarehouses(ordevelopmentoperations)in

    Pakistanandhencedoesnotclaimtoberepresentativeoftheentireindustry.Totheextentthat

    an

    80:20

    rule

    can

    be

    demonstrated

    to

    apply

    to

    Pakistans

    software

    industry,

    our

    survey

    sample

    8 Although a significant number in its own right, these figures present a picture of an industry that is quiteinsignificant in the bigger scheme of things, namely, its contribution to Pakistans economy both in terms of revenuegeneration as well as employment creation capacity.

  • 8/4/2019 Best Practices Study

    25/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 25

    couldeasilyclaimtocoverthelargestandthemostprominentplayersamongitsrespondents9.

    Wedonot,however,goanyfartherthanthatintryingtoassessorclaimhowrepresentativeour

    findings are for the rest of the industry. For some key statistics, for example, the study can

    providesomeveryaccurateandusefullowerandupperbounds.Thismaybethecasewithdata

    onindustryrevenues,employment,andqualitycertificationsetc.Forotherstatistics,thestudy

    mayonlybeabletoprovideagutfeelestimateofhowthingsareontheground.Thismaybethe

    casewithdataonmanagerial,marketing,andtechnicalpractices,andaccesstofundingetc.For

    others still (e.g. issues specific to smaller companies), the study may not represent the true

    pictureof the industryatall.We leave it to thejudgmentofouraudience todraw theirown

    conclusionsonacasebycasebasis.

    TableI(below)presentsabriefstatisticalsnapshotofPakistanssoftwareindustryasgleaned

    fromoursample60respondents.Althoughthedataisquiteselfexplanatory,someaspectsare

    worthnotinghere.Inacumulativesense,the60softwarehousesinourstatisticalsamplehave

    combinedrevenuesofover$80million(seefootnoteandTableIIfordetails)andemployover

    4000 technical and professional employees. This picture of revenues is, however, merely an

    estimate extrapolated through categorical data. TableII presents more accurate categorical

    estimatesoftherevenuesofourrespondents.Ofthe52companiesthatreportedtheirrevenues,

    slightlymorethanathird(19companiesor36%)hadannualrevenuesbetween$200Kand$1M,

    about a third (17 companies, or 32%) had annual revenues greater than $1M (4 of these had

    annual revenues in excess of $5M), and another third (16 companies, or 30%) had annual

    revenuesoflessthan$200K.Inanaggregatesense,thesesoftwarehouseshaveseenarevenue

    andemploymentgrowthofabout37.4and27.4percentrespectivelyoverthelastyearhinting

    ateitherimprovedcapacityutilizationintheindustryorvalueadditionperemployedtechnical

    andmanagerialemployee,orboth.Theaveragesizeofasoftwarehousecomesouttobeabout

    62employeeswiththeperemployeerevenuepotentialbeingaround$21,800perannum.

    9UNCTAD (2004) makes a similar claim, attributed to PSEB, in that the top-15 or so software companies in

    Pakistan (e.g. the likes of Xavor, Techlogix, Netsol, Systems, Softech etc.) could account for as much as 75% of theoverall industry revenues

  • 8/4/2019 Best Practices Study

    26/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 26

    TableI:KeyAggregateStatisticsOnRespondentCompanies*

    Revenue&EmploymentinSurveyedCompanies #(%)ofSoftwareHouses

    TotalNumberofSoftwareHousesSurveyed 60***

    CumulativeRevenues(calculatedthroughmidpointestimation)** $81.15Million*^

    Total#ofProfessional/TechnicalEmployees 4070

    AveragesizeofCompany(#ofProfessional/TechnicalEmployees) 62

    RevenueperTechnicalandProfessionalEmployee $21,814

    %GrowthinProfessional/TechnicalEmployment(overlastyear) 27.47%

    %GrowthinRevenues(overlastyear) 37.4%

    Ownership Structure and Quality Characteristics of Companies

    %ofCompaniesthataresubsidiariesofForeignCompanies 40%

    %ofCompanieshavingFrontOfficesabroad(US,UK/EU,ME,AP) 55%

    %ofCompanieshavingaQualityCertification(ISO,CMM) 45%(3.3%haveaCMM)

    %ofCompanieshavingaDedicatedQualityAssuranceTeam 73.7%

    Product & Strategic Posture of CompaniesProductProfile****

    ProductfocusedorPackagedSoftwareCompany 56.67%

    Software/ITServicesCompany 48.34%

    Software/ITConsultingCompany 31.67%

    StrategicPosture****

    Nicheproduct/serviceforaNicheMarket 36.67%

    Product/serviceapplicabletoseveralindustries 56.67%

    Product/serviceapplicabletoanindustryvertical 33.34%

    *Thesearebasedonselfreportedannualrevenues

    *^TheStateBankofPakistanestimatesthecountrysexportsfiguresoflastyeartobe$32M.

    **Thisestimateneedstobeusedwithgreatcaution.Thecorrespondinglowerandupperlimits

    are$39.35and$110.95Million. PleaserefertoTableIIforadetailedcategoricalbreakdown

    ***46softwarehousesweresurveyedinpersonwhile14submitteddatathroughonlinesurvey

    ****Thesecategoriesarenotmutuallyexclusivei.e.acompanycanoptforoneormorecategories

    In terms of their product/service strategy and strategic posture in the market, 56% of the

    companies described themselves as productfocused (packagedsoftware) companies, 48% as

    software/ITservices companies, and 31% as software/IT consulting companies. It is worthemphasizingherethatmajorityofthecompaniesthatdescribedthemselvesasproductfocused

    dealtwithcustomizedratherthanshirkwrappedproducts.Intermsofindustryfocus,abouta

    third of the companies described themselves as niche players, another third focused on an

    industry vertical, and about 56% produced a product/service applicable to several industries.

    Clearly,noneofthesecategoriesaremutuallyexclusive.Manycompanies(asmanyas42%and

  • 8/4/2019 Best Practices Study

    27/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 27

    23%) identified with more than one category in the product profile and strategic posture

    respectively.

    TABLEII:SIZEOFRESPONDINGCOMPANIESBYREVENUE*&

    EMPLOYMENT

    AnnualRevenuesinUS$(PKR**) #(%)ofSoftwareHouses

    N=52***

    Greaterthan$5Million(>PKR300Million) 4 (7.69%)

    Between$1and5Million(~PKR60300Million) 13 (25%)

    Between$500Kand1Million(~PKR3060Million) 9 (17.31%)

    Between$200Kand500K(~PKR1230Million) 10(19.23%)

    Between$100Kand200K(~PKR612Million) 6 (11.54%)

    Between$50Kand100K(~PKR36Million) 5 (9.62%)

    Lessthan$50K(~PKR3Million) 5 (9.62%)

    Total 52(100%)

    FulltimeEmployment

    N=60

    Greaterthan250Employees 6(10%)

    Between100and250Employees 8(13.33%)

    Between25100Employees 23(38.33%)

    Between525Employees 22(36.57%)

    Lessthan5Employees 1(1.67%)

    Total 60(100%)

    *Thesearebasedonselfreportedannualrevenues**1US$=60PKR

    ***Eightcompaniesinoursampledidnotreportfullyearrevenueseitherbecauseitwastheir

    firstyearofoperationorbecausetheyweredevelopmentcentersofforeigncompanieswithno

    independentrevenueestimatesoftheirown.

    TableIIIpresentsastatisticalprofileoftheindustrystargetcustomers.Broadlyspeaking,our

    60respondentsderivetheirrevenuesfromexportanddomesticmarketsinaratioof60:40.One

    theexportsside,ourrespondentsderive22.5%and38.5%of therevenues fromproductsand

    services respectively. Because of the preponderance of customizable products in the product

    servicemix,wecontemplatethattheproportionfromexportofproductsmaybeoverestimated

    andthusrepresentanupperboundonly10.

    10Somelocalsoftwarehousesengagedindevelopmentofsoftware(i.e.programmingandcoding)a

    service,fromthestandpointofthelocaloutfitforforeignproductbasedcompaniesmayhaveidentified

    theirrevenuesasarisingfromproducts.

  • 8/4/2019 Best Practices Study

    28/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 28

    TABLEIII:WHOMDOPAKISTANISOFTWARECOMPANIESSELLTO?

    Exportsvs.Domestic&Productsvs.Services %ofTotalRevenues*

    N=54

    ExportProducts 22.56%

    ExportServices 38.52%

    DomesticProducts 23.37%

    DomesticServices 16.53%

    Exportsvs.Domestic&Publicvs.PrivateSectors

    N=54

    PublicSector(Govt.)Domestic 8.51%

    PublicSector(Govt.)Foreign 5.90%

    PrivateSectorDomestic 30.79%

    PrivateSectorForeign 54.77%

    *Thesearebasedonselfreportedpercentagesofannualrevenues

    On the domestic side, our respondents derive around 23% and 16.5% of their revenues from

    products and services respectively. Again, a major chunk of the products revenue would

    comprise customized or customdeveloped products rather than shrinkwrapped products.

    Another factor worth considering on the domestic side is a certain number of hybrid

    companies that, for reasons having to do with the necessities of theirbusiness and revenue

    models, bundle hardware with the software they develop. Examples maybe banking

    automation

    companies,

    callcenter

    solutions

    companies,

    and

    mobile/handheld

    devices

    companiesandotherswhoseofferingsdependonsimultaneoussaleofspecializedhardware.

    This,once again,wouldnecessitate that the figureonrevenues fromdomesticproducts isan

    upperbound rather than an accurate estimate of sales from purely software development

    activity.

    Atthesectorallevel,ourrespondentsderiveanoverwhelmingportionoftheirrevenues (>85%)

    from theprivatesectorwithonly8.5%of thesalescoming fromgovt.orpublicsectoron the

    domestic frontandanother6%on the foreign front.Thisessentiallyconfirms theobservation

    about the relatively insignificant role played by public sector and the government as a

    sophisticatedbuyerofsoftwareproductsandservices inPakistan.Many intervieweesthatwe

    spoke to stressed the need for the government tojumpstart the demand for local software

  • 8/4/2019 Best Practices Study

    29/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 29

    developmentby intelligentlyusing itsdemandcreatingability.Wewilldiscuss this theme in

    muchgreaterdetailinalaterpartofthisreport.

    As we dig deeper into this statistical analysis, it is quite obvious that while these toplevel

    (aggregate)statisticshavetheirownvalue,theytendtomaskthevitalityandheterogeneityof

    the underlying data. That exportfocused operations maybe different from domesticfocused

    operations, larger operations maybe different from smaller operations, and productbased

    operations maybe different from servicesbased operations not only in terms of their

    organizationalandmanagerialarrangementsbutalsothestrategicandcompetitivedriversisa

    foregone conclusion. What we need, therefore, is a much more finegrained analysis that

    focusesontheseimportantsubcategoriesinadditiontotheaggregatelevelstatistics.Thiskind

    ofanalysisalsohasimportantimplicationsforthebusinessmodelandstrategyissuesthatwe

    address,inaqualitativesense,insection6ofthisreport.

    Asweattempttoderive importantsubcategoriesofourdata,ourfirstreferencepointmaybe

    Dr.RichardHeeksworkonsoftwarestrategiesfordevelopingcountries(Heeks,1999).Figure

    II (below) presents a graphical representation of Heeks software strategies framework that

    dividesthepotentialuniverseofstrategiesinto4distinctcomponents,namely,exportservices

    (StrategyA), exportproducts (StrategyB), domesticproducts (StrategyC), and domestic

    services (StrategyD).Forcomparisonpurposes,henames the former (StrategyA&B)as24

    CarotorFoolsGold,andthelatter(StrategyC)asThirdWorldMicrosoftand(StrategyD)

    as Small Fish in a Small Pond. According to Heeks (1999) each of these four strategies is

    distinct in the sense that each has its own set of organizational requirements, competitive

    drivers, environmental prerequisites, and riskfactors. Yet, as Heeks (1999) points out with

    illustrations from Indias case, companies try to adopt each of these strategies to varying

    degreesofsuccess.

  • 8/4/2019 Best Practices Study

    30/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 30

    WhileHeeks(1999)discussesrelevantfactorsthatmightmakeeachoftheseapproachesmore

    orless

    risky

    (and

    thus

    more

    or

    less

    likely

    to

    succeed)

    without

    actually

    presenting

    empirical

    data

    onhowmanyIndiancompaniesadopteachoftheseapproachesandwhatpercentageofthem

    do so successfully, we are able to put data on Heeks framework for the case of Pakistans

    softwareindustry.ThesedataarepresentedinfigureII(above).Inall,37%ofourrespondents

    seem to followstrategyA,20%seem to followstrategyB,9%seem to followstrategyC,and

    another 25% seem to follow strategyD. The overall picture that emerges from overlying our

    dataonHeeksframeworkisthatofoverrelianceontherelativelyriskierofthefourstrategies

    (StrategiesA&B)thatHeekscalls24CarotorFoolsGoldandunderrelianceontherelatively

    less riskier one (Strategy D) that he calls Small Fish in Small Pond. FigureIII presents thisdatainagraphicalformat.Wewilldiscussgenericbusinessstrategiesthataremodificationof

    Heeks4partframeworkinsomewhatdetailinthenextsection.

    C:(Local-Products > 50%)

    Total #: 14

    % of Total: 25%

    B:(Export-Products > 50%)

    Total #: 11% of Total: 20%

    D:(Local-Services > 50%)

    Total #: 5

    % of Total: 9%

    A:(Export-Services > 50%)

    Total #: 20% of Total: 37%

    Software Business

    MarketServed

    Domesti

    Export

    Services Packages

    FigureII: RichardHeeksTaxonomyofSoftwareBusinesses,asappliedtoPakistan

  • 8/4/2019 Best Practices Study

    31/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 31

    100% Exports

    Domestic

    Export

    Fig-III: Product-Market Profile of Pakistani SoftwServices Product

    0 25 50 75

    0

    25

    50

    75

    100

  • 8/4/2019 Best Practices Study

    32/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 32

    Another way to look at the data is to classify software companies solely according to their

    marketorientationi.e.thosethatarepredominantlyexportfocusedvs.domesticfocusedwitha

    lot of hybrids working in between in both export and domestic markets in almost equal

    proportions. From the standpoint of being able to identify differences in organizational,

    managerial, marketing, and technical processes, this seems tobe a more promising approach

    thanHeeks(1999)asitremediesfortheabruptboundarychangesacrossthefoursegmentsin

    Heeksclassification11.Weuseanarbitrarylimitof75%(exportsvs.domesticandproductsvs.

    services) to define a new typology. Thus, we define an exportfocused company as one that

    derivesmore than75%of itsrevenues fromexportsandadomesticfocusedcompanyasone

    that derives more than 75% of itsrevenues from the domestic market (products and services

    combined). Inbetween these twocategoriesareabunchofcompanies thatarecategorizedas

    hybrids(almostequallyactiveinexportanddomesticmarkets).Asimilarscheme,basedona

    75%cutoff,canbedevisedforproduct andservicesfocusedcompanies.

    Given the importance of marketorientation (rather than productservice orientation) as a

    definingfactorintheconceptualizationofsoftwareventuresinthePakistanienvironment,we

    first look at that in greater detail.

    FigureVI (above) presents a

    breakdown of our respondents

    between domestic, export

    focused, and hybrid operations.

    Applying the above classification

    on our sample, we get 18 (37%)

    companiesasdomesticfocused,20

    (41%) of the companies as export

    focused, and 11 (22%) of the

    companies as hybrids. We now

    look at the technical, managerial, and marketing practices of companies in our sample from

    bothanaggregateandacategoricalperspective.

    11 One can speculate that the company doing 51% product-exports is not likely to be very different from doing 49%product-exports, yet the former would be placed in a different category than the latter under Heeks (1999) scheme.

    Fig-IV: Market Orientation of

    Software Companies in Sample

    Hybrids

    22%

    Domestic

    37%

    Export

    41%

    Domestic Export Hybrids

  • 8/4/2019 Best Practices Study

    33/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 33

    Before we discuss the results of this analysis, however, it is important recognize that almost

    none of the differences between export and domesticfocused operations are statistically

    significantatthe5%

    significance level

    and, therefore, may

    atbestbetermedas

    suggestive. In the

    following

    discussion, when

    we talk of

    significance, we

    would mean a

    finding of

    significance from

    practicalratherthan

    statistical

    standpoint. Also,

    what is practically

    significantmayvaryfromsituationtosituationandconstructtoconstruct(e.g.a10%difference

    maybe of littlevalue inonecontextbutof greatvalue inanother). With thatcaveat in mind,

    herearesomeofthestatisticalfindings:

    Theformation&fundingstrategiesforexport &domesticoperationsarechanging.Goingback

    toourdata,exportfocusedsoftwarehousesaremuchless likelytobefundedwithsavingsof

    localfoundersthaneitherthedomesticfocusedorhybridoperations(TableIV).Webelievethis

    tobe a representation of an afterthefact conclusion i.e. it is not that there is a dearth of the

    desiretoexploretheexportrouteamonglocalfoundersbutratherthanthelatterhavenotbeen

    able to successfully do so, eitherbecause of inadequate capital or lack of networks abroad.

    Investmentbyaventurecapitalfundoralocalpartner(e.g.abusinesshouse)isalmostequally

    likely to result in a domestic, an exportfocused or a hybrid software operation. Domestic

    focused software houses are much more likely tobe focused on financial and automation

    systems, and sell a mix of hardwaresoftware offerings than exportfocused software

    TEXTBOX#2:WHATDOESITMEANFORARESULTTOBE(NOT)

    STATISTICALLYSIGNIFICANT?

    0

    10

    20

    30

    40

    qaproll

    1 2 3 4

    generated from the same underlying population. For example, for a 5%

    significanceleveltohold,onemustbesurethat19outof20times,adraw

    fromonesubpopulationwouldcomeouttobedistinctlydifferentfroma

    drawfromtheothersubpopulation.Wedonotgetstatisticallysignificant

    results when either the subpopulations are not dissimilar in which case

    there is littlevariationbetween them or there is too muchvariation,as is

    thecaseinthefigureabovewherethestandarddeviationbands(variation)

    around the means are so large that none of four subpopulations is

    distinctly different from others. Smaller sample sizes can sometimes, not

    always,beahindranceingettingstatisticalsignificance.

    Two subpopulations

    e.g. different types of

    software houses, are

    considered to be

    different,atastatistically

    significantlevel, if one

    can rule out, with a

    measure of confidence

    thatindeedtheyarenot

  • 8/4/2019 Best Practices Study

    34/123

    PakistansSoftwareIndustry:BestPractices&StrategicChallenges2005 34

    operations.Theyhave,onaverage,smallerrevenuesizesbutalsolesserdependenceonasingle

    client.

    Exportfocusedsoftwarehousestendtosufferfromlackofgrowthinprofitability(operations

    continue to grow in revenuesbut not in profitability) much more than domesticfocused or

    hybridsoftwarehouses.Thismightbeduetoadversetermsoftradearisingfromrecessionin

    majorsoftwareexportmarketsintherecentyears,aninabilitytoclimbupthevaluechain,oran

    overrepresentationofdevelopmentcentertypeworkdonebycompaniesinthiscategory.

    There is no clearcut winner among domestic/exportfocused operations in managerial

    practices.>Fromthestandpointofmanagerialpractices(seeTableV,below),thereseemstobe

    virtually no statistically identifiable differencebetween domestic, hybrid, and exportfocused

    software operations in terms of the technicalbackgrounds of the entrepreneurial team (i.e.

    founders).Thereissomesuggestiveevidencethatexportfocusedsoftwareoperationsaremore