7/27/2019 Kmetz Chap One
1/42
Mapping Workflowsand Managing Knowledge
7/27/2019 Kmetz Chap One
2/42
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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.
7/27/2019 Kmetz Chap One
5/42
To my grandsonsOwen, Nicholas, and Carson
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
10/42
Tables
1.1 Forms o inormation imperection .............................................7
1.2 Positive and negative outcomes as a unction
o imperect inormation ..........................................................19
7/27/2019 Kmetz Chap One
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.
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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.
7/27/2019 Kmetz Chap One
16/42
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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.
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
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.
7/27/2019 Kmetz Chap One
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
7/27/2019 Kmetz Chap One
22/42
4 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
23/42
KNOWING WHAT WE KNOW 5
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
7/27/2019 Kmetz Chap One
24/42
6 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
25/42
KNOWING WHAT WE KNOW 7
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
Organization by John L. Kmetz (Farnham: Gower, 1998), p. 17. Copyright 1998.
7/27/2019 Kmetz Chap One
26/42
8 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
27/42
KNOWING WHAT WE KNOW 9
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
7/27/2019 Kmetz Chap One
28/42
10 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
29/42
KNOWING WHAT WE KNOW 11
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.
Source: Adapted by permission of the publishers, in The Information Processing Theory of
Organization by John L. Kmetz (Farnham: Gower, 1998), p. 352. Copyright 1998.
Low
Time
Consistencyand cost
High Organization BSystematic
(irreducible)inconsistency
Organization A
7/27/2019 Kmetz Chap One
30/42
12 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
31/42
KNOWING WHAT WE KNOW 13
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
7/27/2019 Kmetz Chap One
32/42
14 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
33/42
KNOWING WHAT WE KNOW 15
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
7/27/2019 Kmetz Chap One
34/42
16 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
35/42
KNOWING WHAT WE KNOW 17
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
7/27/2019 Kmetz Chap One
36/42
18 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
37/42
KNOWING WHAT WE KNOW 19
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)
Source: Adapted by permission of the publishers, in The Information Processing Theory of
Organization by John L. Kmetz (Farnham: Gower, 1998), p. 43. Copyright 1998.
7/27/2019 Kmetz Chap One
38/42
20 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
39/42
KNOWING WHAT WE KNOW 21
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
7/27/2019 Kmetz Chap One
40/42
22 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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
7/27/2019 Kmetz Chap One
41/42
KNOWING WHAT WE KNOW 23
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
7/27/2019 Kmetz Chap One
42/42
24 MAPPING WORKFLOWS AND MANAGING KNOWLEDGE
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.
Recommended