kmetz chap one

7/27/2019 Kmetz Chap One

1/42

Mapping Workflowsand Managing Knowledge


2/42


3/42

Mapping Workflowsand Managing Knowledge

Capturing Formal and

Tacit Knowledge to Improve

Performance

John L. Kmetz, MBA, DBA

President, Transition Assistance Associates

Associate Professor of Management

Department of Business Administration, and

Faculty Director, Advanced Project Management Certificate Program

University of Delaware


4/42

Mapping Workflows and Managing Knowledge: Capturing Formal and Tacit

Knowledge to Improve Performance

Copyright Business Expert Press, 2012.

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted in any form or by any

meanselectronic, mechanical, photocopy, recording, or any other

except for brief quotations, not to exceed 400 words, without the prior

permission of the publisher.

First published in 2012 by

Business Expert Press, LLC222 East 46th Street, New York, NY 10017

www.businessexpertpress.com

ISBN-13: 978-1-60649-454-7 (paperback)

ISBN-13: 978-1-60649-455-4 (e-book)

DOI 10.4128/9781606494554

Business Expert Press Operations and Supply Chain Management

collection

Collection ISSN: 2156-8189 (print)

Collection ISSN: 2156-8200 (electronic)

Cover design by Jonathan Pennell

Interior design by Exeter Premedia Services Private Ltd.,Chennai, India

First edition: 2012

10 9 8 7 6 5 4 3 2 1

Printed in the United States of America.


5/42

To my grandsonsOwen, Nicholas, and Carson


6/42

Abstract

Tis is a book that does what the title says, and is diferent rom most

business process mapping inormation in three key ways. First, it lets

users capture all the knowledge that goes into a workow in any kind o

organization, including the most di cult kind o all, the tacit knowledge

people bring to the job and carry in their heads. Second, it is simple,

powerul, exible, and easy to learn. Tird, it does not require installing,

learning, and applying a complicated program (sometimes requiring reor-

ganization to support the sotware rather than the sotware supporting

the organization). It was developed by the author in a teen-year longprogram o studying, analyzing, and improving avionics maintenance

processes or the U.S. Navy and the Royal Canadian Air Force, and then

applied to organizations o all kinds ever since, or more than two dec-

ades. It has been taught and applied by the author and others in many

short courses. It works.

Keywords

Business process, business process mapping, workow mapping, knowledge

management, tacit knowledge


7/42

Contents

List of Figures ......................................................................................viii

List of ables ...........................................................................................x

How to Use his Book............................................................................xi

Introduction ....................................................................................... xiii

Objectives of his Book...................................................................... xvii

Chapter 1 Knowing What We Know ..................................................1

Chapter 2 Systems, Processes, Organizations, and Worklows ...........25

Chapter 3 Worklow Mapping Fundamentals ...................................45

Chapter 4 WFMA Data Collection and Analysis ..............................97

Chapter 5 WFMA and Knowledge Management ...........................141

Chapter 6 WFMA and Dynamic Modeling ....................................169

Appendix 1 A Brie Summary o the NAVAIR Study.......................189

Appendix 2 A Partial List o Process Mapping Sotware ...................203

Notes..................................................................................................205

References .......................................................................................... 209

About the Author ................................................................................215

Index .................................................................................................217


8/42

Figures

1.1 An exhaustive model o states o inormation ..............................3

1.2 Properties o inormation ............................................................4

1.3 Functional consistency lag and cost ...........................................11

1.4 A system and its environment ....................................................15

1.5 Formal knowledge, tacit knowledge, and

organizational unctioning .........................................................23

2.1 Examples o actions resulting rom

interactions between locus and mode.........................................28

2.2 Examples o inormation resulting

rom interactions between locus and mode ................................29

2.3 Te basic system model .............................................................30

2.4 Te basic system model with regulatory inormation ows ........33

2.5 Path and synchronization efects on workow outcomes ............40

3.1 Te WFMA symbol set .............................................................54

3.2 Single-cycle process ow ............................................................67

3.3 Branching process ow ..............................................................68

3.4 Multiple-cycle (looping) process ow .....................................69

3.5 wo levels o detail in WFMA (drill-down) ............................70

3.6a High-level view o the process o making bread .........................74

3.6b Four principal processes in making bread ..................................75

3.6c Making bread showing election o multiple risings i desired .....76

3.6d Making bread detailed workow map ........................................77

3.7 Incorrect (a) and correct (b) mapping o parallel processes ......79

3.8 wo illustrations o organizational response to

external events with simultaneous internal processes ...................81


9/42

LIST OF FIGURES ix

3.9 Processing a und trade .............................................................87

3.10 Opening a corporate retirement account ..................................88

3.11a Client process map or Laptime V dinners .............................89

3.11b Kmetz method process map or Laptime V dinners ................89

3.12a Client process map or new book selection ...............................90

3.12b Kmetz method process map or new book selection..................90

3.13a Client Data Flow Diagram map or travel accounting ..............91

3.13b Kmetz method process map or travel accounting .....................914.1 Workow or customer cash-transer order

by method o receipt, ca. 1989 ...............................................115

4.2 Minutes required or cash transer by method o order ...........116

4.3 Te Shewhart experimentation cycle ......................................119

4.4 Histogram o 163 card shu es in 10-second intervals ............126

4.5 Whats in a number? Te composition o elapsed

maintenance time (EM) in avionics maintenance ................131

5.1 Simplied view o intended VAS shop test workow ...........147

5.2 Actual VAS shop workow ..................................................150

5.3 Complex eedback and eedorward relationships ...................154

5.4 Positive and negative outcomes rom workow knowledge .....158

6.1 Manuacturing quality control process workow ....................173

6.2 Exhaustive categorization o process outcomes .......................174

6.3 Basic iTinkmodel o manuacturing with quality control .....175

6.4 Te VAS shop workow as a system .....................................185

A1.1 Te avionics repair cycle in the U.S. Navy aircrat

intermediate maintenance department (AIMD) .....................190

A1.2 A simplied diagram o the avionics repair cycle .....................193

A1.3 Te logistics tail o aircrat maintenance ..............................195


10/42

Tables

1.1 Forms o inormation imperection .............................................7

1.2 Positive and negative outcomes as a unction

o imperect inormation ..........................................................19


11/42

How to Use This Book

his book has been written primarily as a how-to guide to Worklow

Mapping and Analysis, and is ocused on mapping organizational pro-

cesses. At the same time, it explains whywe need to do certain things, and

provides some thinking tools to help achieve that.

As we will see in Chapter 2, processes are combinations o actions

and inormation. o really understand a process and map it, you have to

understand both o these parts. he actions, and the material we workon (oten inormation itsel), are the easy parts to capture. Inormation

is trickierthe ormal part is the inormation we can see in rules, proce-

dures, policies, and the like, and that is usually pretty visible. But people

do things in dierent ways, or many reasons, and so the really hard part

is capturing this tacit knowledge, which gets built into how everyone

does his or her part o a process. I a company or organization wants to

understand its processes, and perhaps set up a knowledge management

system at some point, it will have to include this tacit knowledge.

For the person who needs to get a handle on processes right now, the

place to start is Chapter 3this is the core chapter on process mapping.

Chapter 4 then helps igure out how to collect data on a worklow rom

the maps we create, and analyze and interpret it to do things like process

improvement and change. A great deal depends on intelligent use o our

eyes, but a spreadsheet helps. Simplicity is the key to both making and

analyzing the maps.

At some point, the user ought to back up and read Chapter 2 to have

a uller understanding o organizations as systems, which they all are,and o processes within them. Finally, since inormation is undamental

to everything we are trying to do, the user should take some time to

read Chapter 1. Doing what Chapters 3 and 4 show rom the perspec-

tive o the irst two chapters will make or better worklow maps and

the applications users put them to later. In comparison with Chapter 3,

Chapters 1 and 2 may seem somewhat academic, but those chapters pro-

vide thinking tools that are as important as the mapping tools, and they

work better together; i you have time, start with Chapters 1 and 2both are short.


12/42

Chapters 5 and 6 extend our capabilities in two useul directions.

I we want to get into more ar-reaching knowledge-management eorts,Chapter 5 gives some insight into that and helps shape expectations or

what we can and cannot do with it. Chapter 6 gives us a preview o the

next level o mapping, which is to use it as a precursor or dynamic simu-

lation o our organizational processes. here is already a lot o this kind o

simulation being done, and I believe it will become a source o competi-

tive advantage or many irms in the uture.

Can we actually do worklow mapping without sotware, as I suggest

many times in this book? o draw the maps, we really dont need anything

more than the drawing programs that are built into word processors and

spreadsheets, but a dedicated program like SmartDrawor Visio (among

others) can be very helpul. For data analysis, you need a standard spread-

sheet. What you dont need is a costly, complex, proprietary process

mapping program (and oten all the support sta that comes with it).

xii HOW TO USE THIS BOOK


13/42

Introduction

7:00 AM is a miserable time o the morning to have to start work; it is an

even more miserable time to begin a presentation on a diicult job to an

even more diicult audience, only to have it end by nearly being booed

o the platorm about 15 minutes later. I know this because it happened

to me a long time ago, when I irst tried to present a complex diagram

o a low o work to a meeting o the Common Automatic est Equip-

ment Integrated Logistics Support Management eam (CAE/ILSM,to insiders) in San Diego one beautiul summer morning in 1980. Out o

this experience, in some ways, grew the procedure I am going to share in

this small volume. his technique, called worklow mapping and analy-

sis (WFMA), is a way o visually capturing a low o work by using a

small set o symbols in a very consistent way; it was born rom a need to

visually portray a challenging low o work in the maintenance o aviation

electronics (avionics) or the U.S. Naval air orce (NAVAIR) during

the Cold War. his is a method that was developed rom a need to com-

municate important inormation about that worklow to military and

civilian navy personnel who needed to know it, and not get booed o the

platorm beore that communication could be accomplished.

Within a ew years, I had developed a revised approach to WFMA

which met with a much better reception, and hence was much more

eective in helping me and several colleagues meet our objectives in our

research with the CAE/ILSM. A large part o this had to do with map-

ping a hugely diverse range o worklow activities and procedures, along

with their supporting lows o inormation. his required a techniquewhich was lexible and robust enough to capture everything that hap-

pened in that low o work. I reined this technique even urther when

I began working on an article or the academic journal Administrative

Science Quarterlythat was published in 1984. By that time, I had come

to realize that while WFMA used a small set o standard lowcharting

symbols that I had learned in my college programming courses, WFMA

was not lowcharting by any stretch o the imagination; that was one o

the things I had to unlearn to really make this technique work. O much


14/42

xiv INTRODUCTION

greater importance was the realization that WFMA was a potent tool to

capture all kinds o knowledge embedded in a worklow.Among the chie lessons I had also learned was that the symbol set

necessary to communicate graphically about lows o work and inorma-

tion could also be the greatest obstacle to success. he symbols are neces-

sary, but they must be as simple and unobtrusive as possible i they are

not to smother the inormation they provide. he particular symbols and

the disciplined approach to WFMA that I present here may seem overly

simple at irst, and overly rigid at other times; I argue that neither o these

is true, but rather that my approach has been developed through many

years in the inamous school o hard knocks, and what has emerged is

what works. Some o my experience tends to support an old, somewhat

cynical deinition: Experience is that which makes you wonder how

it ever got a reputation or being the best teacher. But it was, and what

I learned in those years is one o the principal inluences in shaping my

approach to WFMA.

My experiences in the NAVAIR environment were given another chal-

lenge in the early 1990s, when I was asked to examine the avionics mainte-

nance processes used by the royal canadian air orce (RCAF). In 1980, theCanadians purchased a new wing o CF-18 Hornets rom McDonnell-

Douglas (now Boeing) to replace an aging mix o various aircrat in the

RCAF. he CF-18 is an export model o the U.S. Navys F/A-18, and is

supported by a similar type o automatic avionics tester, which was the

subject o my studies in the U.S. Navy; within a ew years, the Canadians

had begun to experience the same kinds o avionics maintenance work-

low problems as had the U.S. Navy, and similarly suspected the automatic

tester to be the culprit. hey came to the tester manuacturer or help

with solving the problem; I was asked to take this on, and I ound that

analyzing their worklow through WFMA was again the necessary irst

step, just as it had been or NAVAIR. he end result was a 1991 man-

agement manual that helped the RCAF deal with the challenges o this

worklow.

By the early 1990s, I had begun to develop this approach into a course

which I still oer, and applied it in a number o companies and non-proit

organizations. I ound that my approach unctioned just as well in service

worklows as it did in maintenance or manuacturing environments. his


15/42

INTRODUCTION xv

is not very surprising in retrospect, since the original NAVAIR worklow

was a repair process; while it handled physical materials, it was really aknowledge-intensive service activity that required the diagnosis o aults

and the replacement o parts, with ollow-up testing to be sure a repair

was successul. All o the major issues in that worklow, we came to ind,

were issues o inormation and the processing o it, and not issues with

the hardware or sotware; the avionics maintenance process is really an

excellent example o knowledge workers at work.

A inal realization rom these years o experience and investigation

was that one o the principal beneits o WFMA is that it makes many

aspects o the knowledge embedded in a worklow explicit, even when

these involve inormation processing and decision making that are the

results o many cycles o trial and error, perhaps involving many people

and long periods o time. In todays discussion o knowledge manage-

ment this has been recognized as a principal orm o tacit knowledge.

WFMA is capable o capturing both ormal and tacit knowledge

and making it accessible to a company or organization. In the case o

NAVAIR, it became an absolute necessity or resolving a very serious

problem o Cold War readiness or the leets carrier aircrat.Knowledge management (KM) is one o the most challenging tasks

acing many companies today, and is oten one o the most rustrating.

WFMA can not only make both ormal and tacit knowledge visible and

accessible, it can provide a repository or storing both types o inormation.

hus, it can be a tool or creating and storing job descriptions; it can serve

as a training tool; it can be a key tool in personnel accession planning;

and, o course, it is a undamental requirement or Six Sigma programs,

quality management, and process improvement.


16/42


17/42

Objectives of This Book

he ollowing chapters are designed to achieve these objectives:

1. Provide a general ramework or understanding knowledge and

inormation, the oundations o both processes and worklow maps.

2. Deine Worklow Mapping and Analysis (WFMA) as a descriptive

and analytical tool.3. Illustrate the need or WFMA to be understood and applied as a

disciplined approach to the analysis o worklows and processes

4. Demonstrate the utility o WFMA or analyzing and collecting data

on organizational processes.

5. Illustrate the applicability o WFMA to universal organization issues

such as eiciency, quality, and control; it also supports many unc-

tions in HRM, training, certiication, and process improvement.

6. Illustrate how WFMA enables users to capture both the ormal

knowledge required to do the work o an organization, and embed-

ded process-relevant tacit knowledge as well.

7. Illustrate how WFMA may support development o more advanced

dynamic models o organization, tools that permit the testing o

ideas and changes on a virtual model o an organization prior to

their implementation.

WFMA is ageneralisttool or managers, supervisors, and proessionals

needing to understand how work is processed through an organization.WFMA is nota specialist tool such as programming, data low diagram-

ming, and so on, and it does not requiresotware training or extensive sot-

ware mastery. Nevertheless, like nearly any current inormation-intensive

task, sotware can be a powerul tool that can support WFMA i it is used

to proper advantage.

Using WFMA eectively takes a little practice, like most new things

in the world, and I have provided some simple exercises in the context o


18/42

xviii OBJECTIVES OF THIS BOOK

amiliar terms and situations in Chapter 3; these exercises also illustrate

some o the dierences between WFMA and general lowcharting, whichare helpul to keep in mind. For those who might be curious about the

origins o this approach, I have added a short appendix on the original

NAVAIR work. Enjoy, and I hope you ind this helpul.


19/42

CHAPTER 1

Knowing What We Know

Knowledge and Information: The Framework

his is a book about organizations and understanding how they work.

A undamental idea that we are going to use is that o a process becauseeverything that organizations do, in one way or another, can be described

as a process. In organizations, we design processes to accomplish speciic

goals, and I ind it useul to think o these designed processes as work-

lows, the term in the title o this book.

Organizations and worklows are both critically dependent on inor-

mation; in act, I am going to argue that organizations themselves are

inormation processors, in a very undamental way. he inormation

they process consists o two main typesinormation about what we are

producing, whether tangible goods or intangible products like services,

and inormation about how we do that. he latter can be thought o as

inormation necessary or coordination, which is an absolute requirement

in organizations because their reason or existing is to do work that is

beyond the capability o a single individual. Coordination requires both

ormal and tacit knowledge, two other key terms we will hear much o

in this book.

he worklow used by an artisan to crat an item o jewelry is

entirely up to that artisan; as soon as the artisan hires help or that lowo work, however, it becomes necessary to think about who does what, in

what order, what happens when things do not go as planned, and much

more. hat is what we mean by coordination, and there is no escaping

it. Some o this inormation may be ormal rules and policies, but a great

deal o it is individual and worked out on the basis o day to day interac-

tion, the way that masters and their apprentices did it or centuries. For

every person added to the organization, the coordination requirements go


20/42

2 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE

up geometricallycoordinating our people takes much more than twice

the inormation processing needed or two.Organizations do not just happen, and thereore neither do work-

lowsthey are designed. Neither one is static nor unchanging over time,

so what was designed at one time will need to be modiied in the uture.

We are constantly changing organization structures or one reason or

another. One consequence is that the worklow designs that made good

sense at one time no longer do, but they persist and oten become seri-

ously out o whack with the goals o the organization.

So, i there isnt a Second Law o Organizational hermodynamics,

there should be. In physics, the Second Law o hermodynamics

says that everything eventually winds down until energy is evenly distrib-

uted throughout the universe and everything comes to a stop. Entropy

rules! My experience with organizations suggests they ollow this law, and

this chapter is going to present some underlying reasons why this is the case.

We build on this in Chapter 2 to set the stage or the tools and techniques

we will see in Chapters 3 and 4, tools that not only help us to manage some

o the chaos but actually change and improve processes and perormance.

Understanding worklows also requires some undamental under-standing o inormation and knowledge, and in this way KM is related to

worklow mapping. Given this relationship, one payo is that mapping by

the Kmetz method becomes a valuable way o capturing both ormal and

tacit knowledge in the worklow. We will discuss KM in more detail in

Chapter 5our immediate concern is to know more about inormation

and knowledge, two words that we use all the time but seldom appreci-

ate or their richness and complexity.

Knowledge Is Information Is Knowledge

I want to begin with an idea that in some ways is the entire point o

this introductory chapter. In the perspective o this book, inormation

is knowledge, in the sense that it is a product o human intellect; it is

structured, rather than random; and it is communicable to others.1 his is

a utilitarian perspective on the deinition o knowledge, in the sense that

i you do not know you have inormation, then you do not have it. wo

simple models help explain this perspective.


21/42

KNOWING WHAT WE KNOW 3

he irst model posits that all inormation (and thus all knowledge)

can be represented by a simple 22 ramework, shown in Figure 1.1.

his simple model categorizes all inormation into one o our cells.

Known knowns (KK) are those items o inormation we consider to be

acts, or to which we attach so little uncertainty as to make them eec-tively actual; known unknowns (KU) are essentially questions we know

to be unanswered. Unknown knowns (UK) are inormation which we

may have but cannot unambiguously interpreta classic illustration is

the problem aced by intelligence analysts, who are conronted with myr-

iad acts that cannot be easily evaluated or truth or accuracy, or what they

collectively mean. he inal cell comprises unknown unknowns (UU),

eectively an undeined area o inormation, the existence o which might

be surmised but cannot be orced to yield to analysisor example, whatis the likelihood that a speciic person will break his or her let leg in

exactly 27 days; the probability that the Yellowstone volcano will erupt

with the same orce as its last eruption (and on a historical basis, it is due)

and potentially end advanced civilization; the odds that we are actually on

a surace in 11-dimension space-time, and that none o the universe we

see can even begin to be understood in the our dimensions o space and

time? All o these are serious questions, but with the exception o theoreti-

cal physics we have no way to rame a serious question in terms that we

can comprehend, let alone a meaningul answer.

Figure 1.1. An exhaustive model of states of information.

UnknownUnknowns

KnownUnknowns

Unknowns

State of the universe

UnknownUnknownKnowns

KnownKnowns

Known

State of ourinformation

Knowns


22/42


Figure 1.1 provides a way o characterizing the overall state o the inor-

mation we have in terms o both knowns and unknowns. he contents othese cells are not the same or dierent observers, however, because inor-

mation is a product o human intellect and dependent on the observer.

One aspect o this content is that or each observer, any item o inorma-

tion may be described as a vector, which is our second model. In the ter-

minology o linear algebra a vector may be thought o as an expression o a

single path through a multidimensional matrix. In terms o human experi-

ence, at least seven properties o any item o inormation might deine a

vector, as shown in Figure 1.2.

he vector in Figure 1.2 is the dotted line connecting each scale or

continuum or seven properties o inormation. Each o the seven properties

is an opposite pair (truealse, consistentinconsistent, and so on.), where

the extreme end o each scale might be deined by the associated wordor

example, only statements at the extreme let o the irst continuum are really

true. Where the dotted line intersects each o the scales deines the value o

the vector or a speciic item o inormation as seen by one observer.

How could inormation have a vector like that shown in Figure 1.2in

particular, how can inormation be partly true and partly alse, as shownon the irst continuum? Consider the ollowing statement: I dont know

whether to believe them entirely or not, but the numbers coming out o the

rare-earth explorations weve been doing at Site X, even though they dont

agree with a number o other prospectors whove looked around the same

area, seem to make a strong case or spending some serious development

Figure 1.2. Properties of information.

Source: Adapted by permission of the publishers, in The Information Processing Theory of

Organization by John L. Kmetz (Farnham: Gower, 1998), p. 16. Copyright 1998.

True

FactConsistentSpecificParticular

Explicit

Information is a vector representing the location of an informationelement on all of these properties, each of which is a continuum.

ObjectiveFalse

FeelingInconsistentGeneralAggregated

Implicit

Subjective


23/42


money. Going rom top to bottom o Figure 1.2, all seven vector properties

are embodied in this statement. his is the way inormation usually comesto usit is a bundle o qualities that are not necessarily reconcilable with

each other, let alone the basis or a irm conclusion or immediate action. We

literally need time and thought to igure out what we consider to be a KK.

Moreover, it is highly unlikely that any two individuals will perceive

an item o inormation in identical terms or each o these vector proper-

ties; that is, the meaningo inormation (knowledge) to one person will

inevitably not be the same as or another. Depending on where one per-

son considers an item o inormation to all on each vector, a bit o inor-

mation may be considered highly credible and be placed in cell KK in

Figure 1.1; another observer who evaluates the vector properties or that

item dierently places that item in cell KU. For example, source credibil-

ity will strongly aect where one places inormation on these continua,

as any ollower o marketing or political science can easily attest.

he idea o known knowns may ultimately be an oversimpliica-

tion. Very ew things are truly known in the sense o being ixed and

inalcourts review verdicts, analysts recalculate the books or businesses,

research outcomes are reviewed, and so on. Because inormation is a unc-tion o both inherent content and human perception and processing, eve-

rything is subject to reinterpretation. Much o the tacit knowledge in

organizations is derived rom these kinds o highly individual processes.

What both o the models in Figures 1.1 and 1.2 emphasize is the

importance o thinking about what we know, and also about how much

conidence we have in that knowledge. In his highly recommended book,

he Black Swan, aleb points out a number o very important character-

istics o human inormation processing which may lead to error in our

conclusions about things.2 We have a tendency to tunnel, as he terms

it, to look at one or more sources o inormation and disregard others.

An immediate implication is that we need to be as receptive to inorma-

tion as we can, perhaps especially to that we do not really want to hear.

he absence o inormation itsel may have valueabsence o evidence

on a subject is not the same as evidence o absence. We are also strongly

persuaded by stories, or narratives, which oten have the property o

making rough knowledge appear to be more smooth and complete than

it really is i we look at it closely. hese two igures give us some simple


24/42


ways to think about what we think we know. In terms o Figure 1.1, cat-

egorizing important inormation on a perormance problem into three othe our cells o that model can be a high-payo application o it, as may

evaluating arguments on the basis o relevant vectors in Figure 1.2.

So does this mean we never really know anything? Perhaps in the

philosophical sense it does, but in the world o working organizations we

deal with the complexities o inormation dierently. Much o what we

know is a social reality, meaning that through usage, experimentation,

and learning, we come to agree on what something means to the extent

that we can use it as i it were a KK. Working knowledge evolves. Most

ormal policies and procedures develop in response to a perceived need,

to ill a vacuum when it becomes evident; they are changed and replaced

in the same way. acit knowledge does the same thing, only on the part

o individuals and small groups. acit worklow knowledge develops in

the environment o ormal organizational knowledge, which has many

implications (one o them being the old bromide that we get things done

around here not because o the rules, but in spite o them). So we may not

have inal answers to anything, but we agree on the inormation we need

to make progress, and that inormation always includes tacit knowledge.

All Information Is Imperfect

What constitutes KU or UK in Figure 1.1 depends considerably on the

individual making the judgment about the contents o these cells. As a

commitment to aith, one observer may reject the entire construct o

Figure 1.1, since it rejects the potential or all unknowns to rest in the

hands o a higher power. Over time, each o the cells with known elements

is a uzzy set, in that the content and classiication system may change.

A humble example o this is the deinition o dishwasher sae kitchen

equipment and cutlery. As a wooden-handled knie (located in the KK

cell as not dishwasher sae when acquired) becomes older, duller, and

less prized, it is less likely to be hand-washed and more likely to be put

in the dishwasher; dishwasher sae is partly a matter o who makes the

determination as well as the physical properties o the item. hereore,

inormation imperection can be summarized as either a problem o

incompleteness, where at a minimum the UU cell in Figure 1.1 can never


25/42


be eliminated, or as a unction o the uzzy set problems induced byvector

propertiesshown in Figure 1.2.

Either o these two orms o inormation imperection may be

the product o active or passive sources, as summarized in able 1.1.

Tese may result in simply incomplete inormation or difering vector

properties.

Active processes are shown in cells 1 and 3. In cell 1, active distor-

tion o inormation or misleading inormation may be provided by acompetitor as a deliberate method or concealment o strategy or inten-

tions; in cell 3, various kinds o analytical error may result in imperect

inormationthese could include incorrect weighting o inormation

content, misinterpretation o vector properties, and simple mathemati-

cal error. Passive orms o imperection are shown in the other two cells,

and are relatively straightorwardthe lack or loss o inormation in

cell 2, rendering what we think we know to be incomplete, and the

unconscious iltering o vector properties or addition o unintended

vector properties to inormation in cell 4. hese our archetypal pro-

cesses are interdependent or any observer; or example, jamming inor-

mation about a source (person) may create emotional ilters that aect

the vector properties o all inormation rom that source. Examples o

such interactions can easily be imagined or all our sources o imperec-

tion, and these interact over time.

We hear all the time that knowledge is oten the most critical asset

any organization possesses. he lengths taken to protect the ormula o

Coca-Cola, to protect innumerable trademarks, and the global concern

Table 1.1. Forms of Information Imperfection

FormSource of imperfection

Active Passive

Incompleteness 1. Misinformation,

disinformation, jamming

2. Uncertainty, lack of infor-

mation, signal loss, or noise

Vector properties 3. Analytical error 4. Values, feelings and

emotions, source-specific

responses, culture

Source: Reprinted by permission of the publishers, in The Information Processing Theory of



26/42


over protection o intellectual property are abundant testimony to that

act. hus, the cells in able 1.1 where inormation is incomplete may bethe product o active processes on the part o external agents who do not

want knowledge to be ull or complete, in addition to imperections rom

our internal thought processes.

It is also important to recognize that actively derived imperections

do not necessarily imply bad intent. Businesses keep at least three sets o

booksone to report to shareholders, one to use or internal decision

making, and one or tax collectors. While we might view this cynically

and suggest that each is intended to keep inormation away rom people,

it is equally true that compliance with a hugely complex tax code may not

always tell the most accurate story o how the business is doing or the

shareholders, and that neither o these is what a manager needs or day-

to-day operations. Changing the way we keep accounts changes the prop-

erties o the knowledge we have to work with, and we need to actively

create dierent versions o a single truth. In the wrong hands, o course,

this same need opens the door or the kinds o abuses we have seen with

the Enrons and WorldComs o the business community.

Inormation imperection is a major issue in the mapping o work-lows, as we will see in Chapter 3. Much o the tacit knowledge in a

worklow becomes so deeply embedded in individualized behavior that it

becomes diicult to extract. Everyone has experienced the startling realiza-

tion o having driven a long distance without really being aware o it until

some point near the end. We overlearn a amiliar route to the extent that

conscious attention to driving it is not necessary, and we navigate by using

waypoints and landmarks; i we are asked how we travel, we suddenly

realize we no longer know route numbers or street names, but these land-

marks. he same thing happens in our work, and a type o uncertainty

is the inevitable result. But we should be aware that in cases where people

eel threatened by a new worklow-mapping project in their company, they

may respond by engaging in jamming and providing disinormation.

Organizations Are Information Processors

When we talk about organizations, the type o social creation we will

be ocusing on in this book, that word conjures up many images. he


27/42


one that I personally preer is that an organization is an inormation pro-

cessor. Organizations, small or large, are groups o people using varioustechnologies to accomplish something that cannot be done through indi-

vidual eort alone. Because we have multiple players, dierent materials,

dierent objectives, dierent constituencies, and all the myriad things

that come with an organization, it is necessary to process inormation to

coordinate everything that has to be done.

his need is easy to understand. A small team o people can coor-

dinate their actions or a small project relatively easily (especially since

they are likely to be sel-selected members or the job at hand). hey

simply ask questions and make suggestions to each other as circumstances

require, and with everyone in contact with each other, processing inor-

mation to coordinate the team is easily managed. But when a job gets

bigger, takes more people and more specialized skills, extends over a long

time, and so on, the capacity or inormation processing activity suicient

to coordinate a small team will simply not be adequate.

he solution to this problem is also easy to understandwe break the

big organization down into smaller groups (typically by the type o skills

people have or the type o output they produce), and have a specialistin charge o each group, so that the amount o inormation that has to

be processed within each group will be dramatically reduced relative to

the whole organization. Each small team will only have to coordinate its

actions within the group, and between-group coordination can be done

by the team leaders. hey may need a higher-level team leader, and i so,

we have just created a three-level hierarchy.

his spontaneous hierarchy is hardly new, and the discovery o the

inormation processingeiciency and eectiveness o the hierarchy is as old

as organization itsel. he Romans are oten credited with invention o the

hierarchy (a centurion was the leader o 10 groups o 10), but hierarchi-

cal military organization was used by the early Assyrians, Genghis Khan,

and the Mongols, among others. his eiciency is also why the hierarchy

is durable, despite the eorts o many thinkers and advocates o alternative

orms o organization to discredit itit persists because it works.

When we design an organization, we have a signiicant impact

on the way the organization will be able to process inormation, and

how much and what kind o processing it will have to do. here are


28/42


signiicant tradeos. I we organize our basic units by skill or type o

work, as opposed to grouping people and skills around production oa particular type o output, we create specialist units that tend to pay

most attention to their specialization, and oten lose touch with the

customer; organizing by product may keep us closer to the customer,

but at the cost o losing our skill (and innovative) edge. I we make

the hierarchy tall and keep all the decision power at the top, it makes

it easier or the whole organization to adapt its overall goals over time,

but at the cost o buy-in and much valuable knowledge that stays

at lower levels; i we reverse that and keep decision making at lower

levels, we risk having the overall goal lost in the cracks between goals

o the individual business units. hese are never-ending problems,

and they are a constant challenge to large organizations because the

tradeos between them are important.

As an example, Gore Associates, the maker o Gore-ex and many

other nonconsumer products, decided to commit to an organizational

orm that relected William Gores experience and preerences rom his

early career in a large company. His decision was to orm production units

o 200 or ewer people, and when a site grew beyond that size, he openeda new physical unita new plant at a new site. he reason was that in his

early experience, large organizations always lost touch with individuals

and were simply not much un; he wanted plants small enough to let eve-

ryone get to know everyone else. He also abolished hierarchy and status

dierentiation, so that everyone who works with Gore is an Associate.

As a result, this global company now has small units in roughly similarly-

sized buildings scattered around the globe. hey have their own unique

problems in trying to coordinate this kind o operation, but have learned

how to do it successully through several long-term business cycles and

the end o patent protection or a major product line.

he scattering o task-related inormation through an organization

thus induces a new type o inormation imperection. Anyone who has

been in an organization knows that keeping both the let and right hands

inormed o what the other is doing is an endless job; moreover, what is

important at any given time depends considerably on the point o view

o both the individuals and the unit they represent. It requires time and

money to process inormation to achieve unctional consistency, where


29/42


goals can be met with enough success to keep the wol rom the door over

the long term.

Figure 1.3 illustrates the problem in general terms. Knowing what

we know is not ree. From the vertical axis, two organizations (A and B)

might start rom much the same level o internal inormation consistency,and both might agree that this level o consistency is inadequate or their

needsthey need to get on the same page. Doing that requires time

and money, and as they move through time to the right, they improve

their consistency, but at increasing cost. (It is also worth noting that the

problem they are working on is less and less current.)

he two organizations may start in a similar position with respect to

their internal degree o unctional consistency, which might be thought o

as increasing the relative size o the KK cell in Figure 1.1. o increase the

size o the KK cell requires eort and expense, as does urther resolution

o the UK and KU cells. How much processing toward these outcomes is

justiied, and how do we know? o what extent is acquiring the nth item

o inormation worthwhile? What is the cost o the time to do this, and

what is the time value o money relative to all o these tasks? hese are

questions that are undamental to any organization design. he existence

o the UU cell means that there is also an irreducible system-level cost,

where spending ininite amounts o money will not gain much by way o

new inormation.

Figure 1.3. Functional consistency lag and cost.



Low

Time

Consistencyand cost

High Organization BSystematic

(irreducible)inconsistency

Organization A


30/42


Several decades ago, Aaron Wildavsky coined the term uncertainty

absorption to describe what happens when inormation in raw or nearlyraworm enters the organization, and decision makershave to deal with

the unknowns and imperections in it.3 Raw data and inormation at any

level o an organization is partially a mess, and what to do in the ace o a

problem is requently not clear. Wildavsky argues that managers acquire

much o their inormation as summaries o it rom the level below (an

interesting process in its own right), and use this to make decisions that are

hopeully consistent with the organizations goals. From the subordinates

point o view, once the management has made a decision and passed it

down, uncertainty about what to do has been absorbed by the manager or

the subordinateI may or may not agree with managements decision, or

example, but my job is to comply with it. his pattern is repeated through

all levels, with all the potential or organizational politicking and inight-

ing one could imagine. Such battles need not originate rom an outside

problemhaving been through several wars over technical design in the

aerospace industry, I can personally attest that the technical battles in the

labs and engineering divisions are as bloody as they come.

Against this backdrop, it should come as no surprise that organiza-tions are destined to constantly struggle with the problem o internally

getting their act together. estament to the diiculty o this job is pro-

vided by the popularity o the Dilbertcomic strip, which parodies the

role o managers and the problems o running a company (a ormer

MBA student, an engineer by training like Dilbert, once told me in

all sincerity that all he needed to know about management in the real

world could be learned by regular reading o the strip; the strip author,

Scott Adams, has oten noted that most ideas or his strips are sent to

him by readers in the working world). he challenge to managers is that

no matter how hard they try, there will always be some things that slip

through the cracks.

Even in the world o international spying and intelligence, the unda-

mental need or eective inormation processing cannot be escaped. he

how did we miss that? or how could we not have known? investiga-

tions that oten ollow intelligence ailures are as predictable as rain.

Even attempts to resolve the problem by restricting inormation access

ailthey only create dierent types o perormance and coordination


31/42


problems. Harold Wilensky made an observation in his 1967 book that

is as true today as ever:4

he more secrecy, the smaller the intelligent audience, the less

systematic the distribution and indexing o research, the greater

the anonymity o authorship, and the more intolerant the attitude

toward deviant views.

Organizations must constantly struggle to get everyone on the same

page, and it is a never-ending battle. Organizations are constantly

restructuring, and by some accounts the average time between reorgani-

zations is at a record low. All o this is driven by the need to process the

right inormation in the right place at the right time, and stay competitive

in a rapidly changing world. From basic hierarchies with a decision maker

at the head o each group, we experiment with pre-made decisions in the

orm o rules, policies, and procedures; we add sta specialists to take

some o the processing burden o line managers; we add inormation

technology; we split up by region or customer group or on some other

basis that makes sense in our industry. Every one o these changes has animpact on our processes, and it is not uncommon to ind that a large part

o the body o tacit knowledge is directed toward patching the cracks in

the worklow let by the last reorganization. Most o these patches, actu-

ally, are taken care o through voluntary action on the part o employees,

who use their tacit knowledge o customers and situations to ix things

when they get out o whack.

Organizations Are Systems

he next o the undamentals to discuss is that organizations are sys-

tems. On one hand this will seem intuitively obvious when explained,

but on the other, it is a powerul and useul way to think about organi-

zational processes and the critical roles o ormal knowledge and tacit

knowledge in them.

he idea o a system that I am using here is based on the everyday

observation o complex, organized, oten sel-regulating entities in the

world around us. Systems are the subject o a body o knowledge known


32/42


as General Systems heory, and they are ormally deined in several ways,

but they all built on the idea that a system is a whole made o componentparts, and which is relatively stable and is both recognized and unctions

as a whole. here are our properties associated with systems as they

are deined in General Systems heory: (1) the whole is greater than the

sum o its parts; (2) the whole determines nature o the parts; (3) the

parts cannot be understood in isolation rom the whole; and (4) the parts

o the system are dynamically interrelatedthey are interdependent and

interact with each other over time.

Much o this sounds theoretical, but the essence o these deinitions

is captured in my own somewhat tongue-in-cheek deinition: A system

is a thing made up o other things, all connected to each other and all

other things. Examples are everywhere. A person is a system; so is a town

or city, on a larger scale, or a gut bacterium, on a much smaller scale.

What is evident rom consideration o these three systems is that any

system is on one hand a subsystem o a larger entity, while at the same

time a supersystem or smaller entities within it. Bacteria in the human

gut are independent organisms on their own right, but as subsystems o

a human body they are essential, and without them the survival o thehuman would be impossible. A political entity like a town or city has

speciic governing bodies which give the entity o town the ability to

regulate behavior, repair itsel, protect itsel rom hazards, and so on, even

as its human subsystems come and go.

In one respect, I tend to preer my inormal deinition because it

orces one relationship to the orerontthe relationship between the

system and its environment. Is the human body the environment or the

bacterium; the town or the citizen? Each system we examine has this rela-

tionship to larger and smaller entities, internally and externally. Where

the system ends and the environment begins is a matter o both scien-

tiic and philosophical debate. his is something we will not attempt to

resolve here, but it has signiicant implications or the way we deine any

system we want to examine through worklow mapping.

Figure 1.4 shows the basic relationship a system maintains with its

environment, and this will be discussed in more detail in Chapter 2, par-

ticularly what goes on inside the system. All systems have a permeable

boundary with their environments, and take inputs, transorm these


33/42


using internal processes into outputs, and return these outputs to the

environment. As I suggested above, the boundary is not ixed or imper-

meable, and how we deine that boundary may have signiicant implica-

tions or worklow characteristics.

Examples o the importance o these boundary relationships can be

ound in modern supply-chain or just in time (JI) management. Inorder to make these methods work, companies must share inormation

with outsiders on a level that a ew years ago would have been con-

sidered an unacceptable breach o corporate conidence. What was the

organizational environment a ew years ago is now part o the operating

system, and cannot work any other way.

Understanding organizations as systems has a number o important

implications or understanding what goes on inside them. First, the

inputs to one system are the outputs rom one or more other systems. We

sometimes think o knowledge as an economic stock o inormation, to

be categorized and accounted or as a bakery would with dierent lours

being prepared to make bread and pastries, or in other cases, knowledge

is treated more as an input variable, without regard to its source(s). My

approach to capturing knowledge does not treat it as only a stock or an

input lowrather, knowledge takes on both roles in dierent times and

circumstances in a worklow.

What dierentiates the system and its environment is oten worth

careul consideration. Most companies (and perhaps most organizations

Figure 1.4. A system and its environment.

Inputs Outputs

Systemboundary

Thesystem

Theenvironment

Transformationprocesses


34/42


in general) want to be as selectively open to the outside world as they can,

while at the same time protecting the intellectual property (knowledgebase) that makes them successul at what they do. his creates interesting

problems and interesting opportunities. A ew years ago many companies

that used telephone back-oice customer support elt that it was a no-

brainer to take those unctions oshore; what was once considered a nec-

essary internal part o the business had become redeined as a routine

unction that could be done by contract employees on another continent.

Since then, many o those irms have had to rethink that decision inso-

ar as critical unctions and customers are concerned. What constitutes a

core body o knowledge, how porous the boundary should be, and how a

company manages the relationships between them is quite important to

worklow design and perormance.

he environment is not static. Companies and organizations must

adjust to shocks and environmental disturbances all the time, and many

o them show remarkable resilience. In addition, there are things that go

wrong internally or all manner o reasons (we oten reer to these as excep-

tions, since they were not what we planned), and we have to adjust to these

as well. Both types o adjustments require improvisation, jerry-rigging, andthe like; they are heavily dependent on the expertise o people at the scene,

at the time. hese adjustments oten become institutionalized because they

worked, and they are both an important orm and important source o tacit

knowledge. From the perspective o the system, however, these tacit-knowl-

edge adjustments are oten nearly invisible, simply because they worked.

Another property to recognize is that complex systems, like compa-

nies and organizations, exhibit a high degree o sel-regulation and adapt-

ability. hese properties are critically dependent on the knowledge base

within the organization, which itsel has to change and adapt as both

internal and environmental orces require. Every individual and every

group or unit within an organization possesses bodies o ormal and tacit

knowledge, the latter oten a large body. hese not only enable the organ-

ization to meet its immediate objectives, but to regulate its processes to

do that and to react to problems in its worklow and change as necessary.

Indeed, Senge and his colleagues argue that mastery o these knowledge

bases and the ability to learn over time is a major competitive advantage

and a requirement or long-term survival.5


35/42


Both the sel-regulation and adaptability o complex systems depends

on what may be thought o as the economic stocks o inormationmentioned earlier, and also lows o inormation. Every part o an

organization depends on a knowledge base o ormal knowledge, which is

principally ocused on the technical aspects o work; this knowledge base

consists o many components, each o which is closely associated with the

dierentiated units that make up the organization. Each o these units

applies its knowledge to the material in the low o work, transorming

raw inputs into inal outputs. But much o this is heavily dependent on

the tacit knowledge base, which is partly brought to the organization by its

members, and partly created within it as the members interact with each

other. It is primarily in this tacit knowledge base where we ind lows

o inormation, in the broadest sense meaning any inormation mobilized

or used in a way that the ormal knowledge base could not anticipate.

Much tacit knowledge is also associated with units o the organization,

but much is not, and it is ree to move and be applied when and where

it is needed to make the organization lexible and adaptable.

here is a good bit more to say about organizations as systems, but or

the present time we should appreciate that internal processes are parto the connected things that make up a system. he inputs we bring

into the organization rom its environment not only include inorma-

tion about suppliers, markets, and so on, but the people who process

it; they bring with them many other inormation inputs, along with a

body o skills and interests. Some o these are unknown when we hire

them and have unanticipated impact on the organizations inormation

processingthey are both an input to carry out ormal processes, and

a stock o their own knowledge which will inluence how they do these

processes. o ully understand an organization requires recognition o the

openness o the system to its external environment as well as the ull

extent to which ormal and tacit knowledge are necessary to meeting its

goals. We will expand on this idea in Chapter 2.

Information, Processes, and Performance

o pull the previous our points about organizational processes and

inormation together, we need to consider the relationship o these to


36/42


perormance. Perormance or goal attainment is not a oregone conclu-

sion in a world o imperect inormation, and this is one o the reasonsthat we oten discuss perormance in terms o the degreeo goal attain-

ment. he linkage between what an organization plans and projects on

the one hand, and what actually happens on the other, is neither a sure

thing nor a straight path. hus, any discussion o perormance must take

actors that cause perormance variations into account; these variations

and deviations in the path to the uture necessitate inormation process-

ing, just as elements o other organizational processes do.

At the same time, the variability o process outcomes and the act

that we are always dealing with imperect inormation makes it diicult

to rigorously link perormance to inormation. For example, the ability

to demonstrate the payo o investments in inormation technology has

been a major challenge or decades. Strassmann argued that much o

the early investment in inormation technology ailed because it simply

automated obsolete methods o doing work.6 Since then, inormation

technology has been argued by some to be a key to the rapid increases

in productivity o the U.S. economy during the late 1980s and early

1990s.7,8 But the time lags and lack o one-or-one correspondencebetween variables in a complex system always make such relationships

diicult to identiy or measure.

he desired or planned level o perormance or a company or

organization might be thought o as the outcome that would be attained

under conditions o perect inormationbut we know that is impos-

sible because we have only imperect inormation. Imperect inormation

causes deviations rom the outcomes that we would obtain with perect

inormation, in the orm o both gains and beneits on one hand, or

as costs and losses, on the other. Considering both positive and nega-

tive outcomes caused by imperect inormation, the perormance o an

organization may be described in terms o the ollowing relationships:

Perormance =

outcomes as planned or

projected (assuming perect

inormation)

+

net payof o outcomes

resulting rom imperect

inormation

What is that last term on the rightthe net payo o outcomes

resulting rom imperect inormation? As shown in able 1.2, this


37/42


payo is the net value o all beneits or gains, and costs or losses, result-

ing rom both proactive and reactive organizational responses to imper-

ect inormation. Companies cannot simply sit and wait or everything

to be known, so we take both proactive and reactive steps to deal with

risks and unknowns. Costs or losses may be incurred whether the organi-

zation attempts to deal with imperect inormation through proactivesteps, such as planning, market research, and orecasting; or they may

be incurred through reactive steps, such as missing market share or

having to correct or compensate or the costs o delay. In either case,

there are planned costs or coordinating organizational activities in the

ace o this imperect inormation, and there are unoreseen costs and

losses. Similarly, beneits may be gained rom both proactive and reac-

tive approaches to dealing with imperect inormation, either through

gains rom anticipation and exploitation o new opportunities and com-

petitive advantages, or through the avoidance o costs or unnecessary

inormation and inormation-processing activities. An organization o

any size usually does most o these things, and obtains many individual

payos. he sum o all outcomes in cells 14 makes up the net payo

o imperect inormation.

Consider the payo o what I reer to as coordination costs. hese

costs may be to acquire inormation or decision making, or may be

the costs o tightly coordinating activities within and between organiza-

tions. Again, an excellent example o the latter is JI vendorcustomer

Table 1.2. Positive and Negative Outcomes as a Function of

Imperfect Information

Response

mode

Payoff value

Cost or loss () Benefit or gain (+)

Proactive steps 1. Planning and forecasting of

future outcomes; coordina-

tion costs

2. Avoidance (errors), an-

ticipation and exploitation

(payoffs)

Reactive steps 4. Opportunity costs of fore-

gone outcomes and payoffs;

coordination costs

3. Passive opportunism (a.k.a.

IIABDFIIf It Aint Broke,

Dont Fix It)




38/42


relationships, where considerable initial cost is borne by both parties to

tightly coordinate their production and logistics lows across companyboundaries. he net payo o that investment in JI, however, is so

great that or many manuacturers any other approach to doing business

is inconceivable.

But imperect inormation oten pays o in terms o beneits. For

those companies able to ind a competitive advantage in their technology

or market niche, returns ar above those obtained by competitors may

be earned. For those who adopt a wait-and-see approach to dealing with

unknowns, problems oten go away and the unnecessary costs o coor-

dination and attempted mastery o new technologies and new markets

are avoidedi it aint broke, dont ix it. O course, many irms using

either approach guess wrong, and ailneither proacting nor reacting are

totally ree o risk.

An interesting implication o all o this is that companies can adjust

to the challenges o imperect inormation through lowered peror-

mance, that is, that i an organization lacks the inormation processing

capacity to cope with all its knowns and unknowns, then an adjustive

reaction is to reduce the level o output relative to what it might havebeen with adequate capacity. he hard question in this is what might

have been, either in terms o opportunity costs or oregone beneits.

Most organizations would not choose to lower perormance levels, but

many do so by not knowing how ormal and tacit knowledge interact

in their worklows.

he obverse, o course, also holdsi inormation processing capac-

ity is increasedin a system, then at a later time there should be a measura-

ble increase in perormance, which has clearly been the argument o both

the inormation technology and business process consulting industries

over the years.

he bottom line to this is that organizations are systems that unc-

tion through inormation processing, and what we know about the or-

mal and tacit aspects o this in our worklows has both direct and indirect

eects on perormance. I this seems obvious at this point, that is excel-

lent; i not, we need to be clear about this undamental point, which

we will expand on in Chapter 2. For now it is necessary to recognize


39/42


that inormation is both the stu o much organizational work, and the

glue that holds the organization together so that it can work.

Summary and Implications

In some ways, it might be appropriate to return to Figure 1.1 and use that

as the summary o this entire chapter, since the real issue is, as the chapter

title says, knowing what we know. By now it should be clear that this is

a more complicated question than it might irst seem, and that realization

is a good thing.

It is a good thing or two major reasons. First, a undamental assump-

tion o this entire book is that as organizations change and evolve over

time, their internal processes need to do the same. Much experience has

shown that this evolutionary change aects not only the overall struc-

ture o the organization, but has many subtle and requently unknown

eects on the worklows within it. Indeed, in later chapters we will hear

about a number o these eects rom many dierent kinds o compa-

nies and organizations. he inormation we do have about processes is

seldom complete since part o that evolution is because people bringoutside knowledge into the organization with them, and use it in crea-

tive, but oten unexpected and unknown ways, to get their work done.

Unless we understand the role o this tacit knowledge in our processes,

we never really know what those processes are. So in short, it is quite

reasonable to ind that in many organizations, we really do not know

how we do things, even though we may think we do beore we take a

careul look.

Second, much o the inormation we use or making decisions and

controlling the day-to-day activities o a productive enterprise, the kind

we consider to be in the KK cell o Figure 1.1, is seldom really examined

or questioned as to whether that designation is accurate. Who has not let

a meeting wondering what the whole thing was about? Who has not had

the experience o being told to manage a inancial decision on the basis

o a policy that, with little analysis, can be shown to be less cost-eective

than an easy alternative? One does not have to look very hard to ind

examples o companies that spent millions o dollars on an Enterprise


40/42


Resource Planning (ERP) system, entirely on aith that it will work, only

to ind that in some respects it never really did. In reality, we do a lot othings in organizations on the basis o because. As long as our cash low

enables us to absorb the costs o because, we can get away with it, but

that may not work over the long term, and we will hear some stories in

this book about that, too.

Knowledge is inormation, and inormation is always partly incom-

plete and in some ways imperect, so it can only be rendered useul

through processing. One o the major unctions o organizations is to

process inormation and knowledge, so that coordinated progress toward

goals is enabled despite the limits to the inormation we ace. Organiza-

tions are also systems, and are thereore open to all manner o inside

and outside shocks and internal changes, all o which require them to

be adaptable. Many organizations do this rather well over the long term,

while many others have short, i interesting, lives. How well an organiza-

tion perorms depends on all the outcomes o its actions, whether proac-

tive oensive behaviors or reactive deensive behaviors. Both o these

may result in costs or beneits, and it is the net payo o these that deter-

mines how we do in the long term. I generally dislike sports analogies ortheir oversimpliication o complex issues, but the idea o batting aver-

ages in baseball applies here. A batter can strike out whether he swings or

not, and or the batter who produces a respectable average o hits in his

at-bats, along with the occasional home run, there is a realistic chance o

making the World Series.

his chapter has ocused on some basic propositions about knowl-

edge, inormation, and organizations. In some ways I have stressed the

limits to our knowledge and our ability to cope with them. his does not

mean that useul management o knowledge is beyond our reach, how-

everquite the opposite. I have ocused on limitations and boundaries

because it is important to know what we know as well as what we do not

know. We may have to give up on the idea o a ull and comprehensive

database or boundless wellspring o innovation based on an open organi-

zation structure, but there are tools and methods that can be very helpul

in increasing the extent to which we know what we know. We will always

have to deal with the reality o unknown unknowns, and the conundrum

that we cannot know what these are; there will always be questions about


41/42


the value o inormation and the value o obtaining more o it, without

ully knowing what the payo o additional inormation might be. Nev-ertheless, there is also the potential to capture more o what we have dis-

covered and learned, and to use what is requently an unknown known

to much greater advantage. One o the key unctions o worklow map-

ping is to help the organization know what it knows.

acit knowledge is always a key to how organizations cope with their

limits to knowledge. Consider three types o organizationsa glass prod-

ucts company, a sotware developer, and a hospital. At the beginning o

this chapter I pointed out that organizations have to process inorma-

tion to achieve both technical and coordinative unctions, and Figure 1.5

shows how ormal and tacit knowledge both contribute to these objec-

tives. First, ormal knowledge is the basis o technical perormance. he

properties o materials that make various glasses, the programming rules

and syntax or computer code, and sources o inection, are all among the

many elements o the ormal knowledge base that technical perormance

depends on; at the same time, coordination depends on related ormal

knowledge o how glass behaves in its molten state, so that a success-

ul production line can be designed; how (and to whom) to assign code

Figure 1.5. Formal knowledge, tacit knowledge, and organizational

functioning.

Formal

Knowledge

Tacit

Technical

Performancefunction

Coordination

Properties ofmaterials;programmingsyntax;sources of

infection

Opening blownglass;transportablechunks ofcomputer code;

modes ofinfection

Design of glassproduction line;assignment ofcode modules;sterile-zone rulesand procedures

Everything elseall other modesof adaptation andadjustment notpredicted in othercells


42/42


modules or new programs; and on the steps and procedures medical sta

ollow to keep sterile zones sterile, since hospital-derived inections are amajor medical problem.

acit knowledge, shown in the right column o Figure 1.5, is equally

important to organizational capabilities. Much o the technical success o

organizations is entirely dependent on what people learn in what might

be thought o as apprenticeships. One learns to open blown glass

through trial and error; knowing how to apply transportable chunks o

computer code is oten a matter o deeply knowing how a piece o code

works, by the programmer; and how and where inections get started is

oten as important as the bug that causes it, and sometimes more so.

What is most important to realize about tacit knowledge, however,

is the bottom-right coordination cellthis is literally everything else

we know how to do. It is where individual and group learning and

knowledge give the organization response capabilities it never could

have anticipated needing, let alone designed. In a universe where we

can never have complete and perect inormation, an absolute neces-

sity is the ability to compensate and adjust when the UUs and other

unknowns in Figure 1.1 reveal themselves. In many situations this celldeines how organizations survive.

he next chapter provides an expanded ramework or understanding

how organizations unction, and that understanding is the basis or the

simple, robust, and widely applicable method o graphically describing

worklow processes, in a orm that can quickly be mastered and applied

to a wide variety o organizations, which is the subject o Chapter 3. he

combination o conceptual tools in this chapter and Chapter 2, and applied

tools in Chapters 3 and 4, will enable managers and analysts to compre-

hensively describe all that is done with material and inormation in a pro-

cess. he ability to accurately capture both ormal and tacit knowledge in

our worklows has a big payo. While it will never solve the undamental

limitations to ull and complete inormation, WFMA will certainly go a

long way toward letting us know what we know, and experience clearly

shows that improvements in the quality o inormation rom that increase

our ability to improve perormance.

kmetz chap one

Documents