Proceedings - LEAD 2009 Conference
KNOWLEDGE AS A CONTINGENCY VARIABLE FOR ORGANIZING
KNOWLEDGE MANAGEMENT SOLUTIONS
Amy Javernick-Will1 and Raymond Levitt2
ABSTRACT
The knowledge-based-view (KBV) of the firm has received wide acceptance and support. This
view recognizes that knowledge is an asset with as much importance as capital to an
organization. With this recognition, many scholars and organizations have stressed the
importance of sharing their knowledge globally. As a result, many organizations have attempted
a variety of knowledge management programs. However, important knowledge to an
organization (or subset of an organization) can vary drastically based on knowledge
characteristics, including the half-life of the knowledge. This can impact the ways that
companies organize and share their knowledge. As a result, the rate of obsolescence of
knowledge can be viewed as a contingency variable that influences effective organization design
and strategy. Using KBV, organization theory, and contingency theory, this paper begins to
analyze how knowledge characteristics of important knowledge for different types of firms and
different types of communities within firms influences the ways they manage knowledge. It
suggests that organizations that account for the characteristics of important knowledge when
designing and implementing a knowledge management program can have a strategic advantage
over competitors.

KEYWORDS: Knowledge Management; Contingency Theory; Organizations
INTRODUCTION
Organizations have begun to recognize knowledge as a resource that can be a competitive
advantage if managed effectively. In order to integrate and capitalize on the combined
knowledge of employees across the organization, many organizations have attempted to

1 Assistant Professor, Department of Civil, Environmental and Architectural Engineering, University of Colorado at
Boulder, Campus Box 428 UCB, Boulder, CO 80309 USA; amy.javernick@colorado.edu
2 Professor, Department of Civil and Environmental Engineering, Stanford University, Stanford, CA

---

Proceedings - LEAD 2009 Conference
implement knowledge management initiatives and programs. Unfortunately, although estimates
vary, over 50% of knowledge management systems implemented in practice fail to achieve their
original goals (Akhavan et al. 2005). Thus, many firms spend time and money implementing a
strategy that is ineffective for their organization.
There are many reasons why knowledge management systems fail. Reasons abound,
including a lack of resources and time, the culture of the organization (McDermott and O'Dell
2001), lack of employee support for the program, and an overemphasis on technology
(Damodaran and Olphert 2000). In many cases, it appears that an organization fails to take a
holistic view of people, processes and technology (Tregaskes et al. 2004), or, in other words,
connecting people, the organizations structure and the right way to transfer this knowledge.
We acknowledge that there are many reasons why knowledge management programs
succeed or fail; however, for this paper, we specifically argue that knowledge must be treated as
a contingency variable for effective organization and community knowledge management
design. The characteristics of knowledge that is important to an organization or subset of an
organization should influence the choice of design and structure for managing this knowledge.
We build upon past work regarding knowledge types (Nonaka 1994; Winter 1987) to include the
half-life of knowledge, emphasize the contextual nature of knowledge, and relate this to
knowledge management design and structure. We start by reviewing literature and our proposed
research methodology. We conclude with a discussion of preliminary results and our proposed
plan to enlarge the study with other firm types.
POINTS OF DEPARTURE
Our focus on knowledge as a contingency variable for organization and knowledge management
design leads us to build upon literature related to knowledge management and organizational
theory, including contingency theory.
Knowledge Management
Knowledge has always been fundamental to organizations; however, the knowledge-based-view
of the firm brought new meaning to the value of knowledge to an organization by identifying
knowledge as a resource with as much importance as capital (Conner and Prahalad 1996; Grant
1996; Spender 1996). Viewing knowledge in this regard leads many firms, scholars and
consultants towards managing the knowledge like any other valuable resource. As a result,

---

Proceedings - LEAD 2009 Conference
knowledge management systems have proliferated in recent years. Even the construction
engineering industry, which tends to be slow to adopt change, has been embracing knowledge
management programs in an attempt to combine and share their knowledge more effectively. A
recent survey of firms in the UK found that approximately 40% of engineering design and
construction organizations have a knowledge management strategy and another 41% plan to have
one within a year (Carrillo et al. 2004).
Although companies are anxious to manage and exploit their organizational knowledge
more efficiently, many of the knowledge management systems implemented in practice fail to
meet their initial goals (Akhavan et al. 2005). Many studies have identified a conflux of barriers
for sharing knowledge effectively (Carrillo and Chinowsky 2006; Fong and Chu 2006), etc..
Some of these reasons include a lack of time and resources, a tendency to rely on technology,
organizational and cultural issues, and individual personnel issues such as hoarding knowledge
for power.
Recent studies have also identified differences in the implementation and choice of
knowledge management strategy depending on the type and size of the firm. For instance,
Carrillo and Chinowsky (2006) found that the large design firms involved in their study had
specific initiatives and funding for sharing knowledge whereas the construction companies
tended to have less of a focus on knowledge management, combining it with business
development initiatives. We found similar results during a recent study regarding the sharing of
institutional knowledge amongst developers, contactors and engineers (Javernick-Will and
Levitt). In this study, all six of the engineering firms had varying degrees of knowledge
management initiatives. Three of these firms had a formal intranet system that combined people
and technology. In comparison, only one of the five construction firms was starting to develop a
formal knowledge management intranet system and none of the four developers had a formal
system in place.
There is no definitive answer regarding why some knowledge management systems fail
for organizations; however, it is likely that a “one size fits all” solution will not work. The
observation that different types of firms rely upon different types of knowledge sharing practices
appears to be promising. Because these firms have different types of important knowledge to
share, there is likely a strong correlation between knowledge type and choice of knowledge
management strategy. Thus, instead of adopting a “one size fits all” strategy, additional research

---

Proceedings - LEAD 2009 Conference
is needed to examine the link between the variables of knowledge types that are important to an
organization or community within an organization and its knowledge management strategy. We
provide a brief overview of Contingency Theory before discussing knowledge dimensions and
types.
Contingency Theory
Simply put, contingency theory argues that there is not one best way for all companies to
organize. Rather than arguing that all organizations should be structured similarly, contingency
theory argues that an organization must chose a structure that is contingent on a wide variety of
factors, including environmental complexity (Lawrence and Lorsch 1967), organization strategy
(Chandler 1962), and technology and techniques of production (Thompson 1967; Woodward
1965). Under this view, contingency theory indicates that a relationship exists between various
contingent variables and an organizations structure. Thus, superior performance should result if
an organization can appropriately link the correct structure to the contingent variables.
The recent interest in knowledge as a competitive resource has many firms racing to
implement knowledge management systems. However, with so many firm’s knowledge
management systems failing to meet their goals, it seems likely that knowledge can be viewed as
a contingent variable for organizing these systems. Recent work by Birkinshaw and colleagues
studied R&D units and found a strong association between the dimension of knowledge and
organization structure, indicating that organizations should take into account the characteristics
of a firms knowledge base (Birkinshaw et al. 2002). Their work focused on two dimensions of
knowledge: observability, or how easy it is to understand the knowledge or activity, and system
embeddedness, the extent to which the knowledge is embedded in a particular context. We seek
to add to this work by expanding attention to knowledge embeddedness and focusing on how
additional knowledge dimensions, including obsolescence, influence the ways companies
organize to share their knowledge.
Knowledge Dimensions and Transfer Mechanisms
Winter (1987) proposed four key dimensions of knowledge that affect the ease of transfer of the
knowledge asset: complexity, tacitness, observability, and system dependence. Thus, if a type of
knowledge is complex, tacit, hard to observe, and embedded in a system or location, it will be

---

Proceedings - LEAD 2009 Conference
hard to transfer. On the other hand, if the knowledge is simple, explicit, easily observable and
not dependent on a location or system, it will be easier to transfer.

Table 1: Winter's Knowledge Dimensions
Hard to Transfer
Easy to Transfer
Complex
Simple
Tacit
Explicit
Hard to Observe
Easily Observable
Embedded
Independent
Nonaka’s theory of knowledge conversion (1994) expanded from Winter’s classification
to address the transfer of knowledge, specifically focusing on tacit and explicit knowledge. He
theorized that knowledge is created and converted through a spiral-like process involving four
steps:
• Socialization: tacit to tacit knowledge transfer through shared experiences such as
mentoring and on-the-job training
• Combination: explicit to explicit knowledge transfer through mechanisms such as
meetings, information processing and technology
• Externalization: the conversion and transfer of tacit knowledge to explicit knowledge
through questioning and reconstruction of perspectives and decisions
• Internalization: the conversion and transfer of explicit knowledge to tacit knowledge
through learning and the awareness of knowledge
Some scholars have looked beyond knowledge characteristics to identify other
influencing factors in an organization’s knowledge strategy, including focusing on not only the
knowledge the organization possesses, but how it practices (Cook and Brown 1999)(Brown and
Duguid). However, for the purpose of this paper we focus primarily on knowledge
characteristics and relate this to knowledge transfer strategy.
Half-life of Knowledge
In addition to Winter’s (1987) dimensions, we propose that organizations must also consider the
“half-life” of important knowledge to the organization. The definition for the half-life of

---

Proceedings - LEAD 2009 Conference
knowledge is the time span in which half of the knowledge becomes obsolete. The amount of
knowledge in the world is expanding rapidly—some estimates indicate that the amount of
knowledge in the world has doubled in the past 10 years and is doubling every 18 months
(Gonzalez 2004). However, the rate at which knowledge becomes out of date varies enormously
from industry to industry (Haindl 2002). For instance, certain scientific principles have been
valid since they were conceived. However, other types of knowledge become obsolete quickly.
In the IT sector, for instance, there is general acceptance of a half-life of approximately one-year
(Haindl 2002). In comparison, the principles of structural engineering remains relatively
unchanged with the exception of new technologies to make the work faster.

Drawing from Winter’s knowledge dimensions and expanding this to include the concept
of the half-life of knowledge, our research seeks to explore the question: How do characteristics
of knowledge that is important to an organization influence their knowledge transfer strategies?
RESEARCH METHODOLOGY
The novelty of this topic and the need to gain informative insights regarding knowledge sharing
practices necessitated a case based methodology for this study. The research topic originated
from insights from our past case studies with 113 participants in fifteen international firms. We
collected data from these case studies regarding the mechanisms companies used to share various
types of knowledge, including technical, institutional and company processes.
For this research, we selected two of the engineering companies who participated in the
original research for this study. These companies were selected because they have well-known
and established “interactive online systems” that have been in use for many years. These
combine social and formal methods of transferring knowledge through an online platform that
contain processes and procedures, but also connect people through forums, people searches and
other means. In addition, these companies have many sub-communities, or communities of
practice, which use this system. A "community of practice", described by Lave and Wenger
(Lave and Wenger 1991) refers to formal and informal groups of practitioners where members
learn from and acquire the sociocultural practices of the community. The communities in our
study are based upon functional (i.e. structural, mechanical) or business (i.e. strategy) lines.
Being able to compare communities within the same firm, which has the same knowledge
sharing processes, attention to knowledge management, resource commitment, etc. helps us to

---

Proceedings - LEAD 2009 Conference
eliminate some of the external factors that could skew the data. Finally, and most importantly
for our data collection, these two firms collect statistics regarding the use of the system and
provided access to this data. Many knowledge transfer processes cannot be tracked to provide
statistics. Because these interactive online platforms provide these statistics, we can compare the
use of the system with the community type to compare knowledge characteristics with transfer
mechanisms and community structure.
Information was collected from the company’s knowledge management divisions
regarding the number of participants in each of the communities and statistics regarding the
various functions of the interactive online platforms each community uses to share their
knowledge. In addition, the first author is currently asking selected community leaders questions
regarding the ways knowledge sharing methods change based on the type of knowledge and
community via email and telephone interviews.
To expand the study, we plan to recruit additional companies from other sectors to
compare the methods they use to share various types of knowledge. After the conference, we
plan to contact two additional companies to collect additional data. One company manages
federally funded research and development centers in four critical areas, including aviation
systems development and homeland security, with eleven different community-based subsets,
such as cybersecurity and health transformation. The other is a science and technology firm that
focuses on five critical areas of expertise, including energy and national security, with twelve
community-based areas, such as industrial products and medical devices. These firms were
involved in a “Working Knowledge Group” with one of the Engineering Firms in this study and
have a formal knowledge management strategy.
PRELIMINARY RESULTS AND DISCUSSION
We focus on two primary knowledge dimensions—contextual/tacitness and our proposed new
dimension of the half-life of knowledge—and relate this to knowledge management strategy—
principally whether the knowledge is shared through technical or social means.
Contextual/Tacitness
Nonaka’s work focused on combining and converting tacit and explicit knowledge. As a result,
he found that the transfer of tacit knowledge occurs primarily through shared experiences and
social mechanisms while explicit knowledge can be transferred through information technology
and other processes. As part of our study on how multinational firms transfer institutional

---

Proceedings - LEAD 2009 Conference
(regulative, normative and cultural-cognitive) knowledge (Javernick-Will and Levitt 2009), we
observed differences based on the knowledge type and how employees within the firm obtained
knowledge from others for their projects. Studying institutional knowledge meant that most of
the knowledge would be specific to the given location from which the knowledge was derived,
but that it would vary based on the degree of tacitness and contextual nature of the knowledge.
From our data and analysis, as the knowledge changed from regulative to normative to cultural-
cognitive, the knowledge generally became both more tacit and more embedded within a given
context, requiring different means and richer media channels to transfer the details of the
knowledge. This falls in line with media richness theory (Daft and Lengel 1986), which argues
that richer, more personal forms of communication are more effective for communication and
transfer than less rich media. For instance, formal processes to transfer knowledge, such as
project databases, written reports and company processes and procedures were used more
frequently for regulative knowledge, which tends to be more explicit and, in many cases, more
easily understood and translatable. As this knowledge became more tacit and difficult to
understand for normative and cultural-cognitive knowledge, the relative frequency of using
formal processes declined. On the other hand, social mechanisms, such as personal discussions,
meetings and reviews, which can often translate the contextual and tacit details of knowledge to
the recipient, were used to transfer all types of institutional knowledge.
Organizations that had interactive online knowledge management solutions facilitated
social connections through forums and people searches. For instance, an employee could search
for a topic and find answers to forum posts posted by a particular employee who they could then
contact regarding their question. Alternatively, they could search for people with experience in a
particular area. One of the people on the knowledge management team described this process:
“Employees search for people approximately one out of four times… in
approximately 80% of the cases, employees first search for content and then
connect to the author. People will contact others directly when they have very
specific project related questions”.
In order to facilitate searches, this same company asks its employees to tag their forums
or knowledge objects with a title, description and context. Tagging the contextual details allows
explicit, context-specific knowledge to be transferred through formal, IT processes. In addition,
it eases the process of searching for and applying the knowledge. With so many different
businesses, they want to make sure that employees understand the application and context of the

---

Proceedings - LEAD 2009 Conference
knowledge posting, whether it is a particular area, project type, etc. to avoid using data for
projects erroneously.
Proposition 1:
If knowledge that is important to an organization is tacit and context-specific, the organization
should focus its knowledge management strategy on social processes to facilitate person-to-
person connections to transfer this knowledge.
Proposition 2:
If knowledge that is important to an organization is explicit and context-specific, the
organization can focus its knowledge management strategy on an IT solution provided that they
require employees to “tag” the knowledge with appropriate details and allow social processes for
employees to obtain additional contextual details.
In order to add other dimensions for consideration of knowledge management strategy for
organizations based upon knowledge type, we requested statistics from the two large, global
engineering companies on the global use of their interactive programs based upon community
type. We then requested the number of employees within each community in order to get
statistics on the use of each of the features per community member. Table 2 presents data
regarding the communities that use forum reads and forum submits the most and least for
Company 1. Table 3 presents data regarding the communities that use forum threads most and
least for Company 2. In addition, Company 2 has a feature that formalizes knowledge through
an in-depth review process. Only two communities currently use this feature for more than one
piece of knowledge: the Structural and Mechanical Community.
Half-Life of Knowledge
In addition to confirming the contextual dimension discussed above, after reviewing the
statistics, it also appears that forums are read most frequently for communities that have a longer
half-life of knowledge. For example, some of the communities with the least read forums
include Contract Management, Strategy & Business Intelligence and Risk. These communities
contain knowledge that can change quickly, require contextual specific details or sensitive
information. On the other hand, the communities with the most read forums—Process
Technology, Coker, and Civil/Structural/Architectural—contain knowledge that has a longer
half-life and may contain knowledge that is less contextual across projects and locations. In
addition, for Company 2, the two communities that are using the formalized knowledge feature

---

Proceedings - LEAD 2009 Conference
(this feature requires a formalized review process including outside company citations) are from
disciplines (Structural and Mechanical) whose knowledge is not changing as rapidly as, for
instance, their sustainability community. Communities that have knowledge that is quickly
changing and becoming outdated are less likely to want to invest the time and resources required
to go through the process of formalizing their knowledge.
Proposition 3:
If knowledge that is important to an organization has a long half-life, that is, the knowledge
remains valid for a lengthy period of time, the organization should invest in technology that can
collect and formalize the knowledge through a technical platform to allow the knowledge to be
easily transferred amongst all of its employees.
Table 2: Company 1 Interactive Online Platform Data for the last 90 days (printed 5/14/2009)
Communities with the Greatest Number/Community
Communities with the Least Number/Community
Member
Member
Forum Reads
Process Technology 9.15
Contract Management 0.00
Coker 6.73
Strategy & Business Intel igence 0.06
Civil/Struct/Arch 6.45
Knowledge Manager Training 0.09
Piping 5.86
General Corporate 0.10
Control Systems 4.60
Risk 0.11
Polysilicon 4.40
Upstream Oil & Gas 0.12
Mechanical 4.33
Infrastructure 0.12
Electrical 3.03
Life Sciences 0.14
Forum Submits
Polysilicon 0.20
Community Relations 0.00
Process Technology 0.18
Contract Management 0.00
Piping 0.16
Infrastructure 0.00
CSA 0.12
Life Sciences 0.00
Ops & Maintenance 0.12
Offshore structures 0.00
Mechanical 0.11
Risk Strategy & Business Intelligence 0.00
Travel 0.00
Upstream Oil & Gas 0.00

Table 3: Company 2 Interactive Online Platform Data over lifetime (2009 employee count)
Communities with the Greatest Number/Community
Communities with the Least Number/Community
Member
Member
Forum Threads
Mechanical Services 6.46
People & Org Change 0.00
Project Controls 4.63
Computational Optimization 0.07
Acoustics 4.05
Knowledge 0.09
Electrical 3.70
Risk Forum 0.18
Structural Skil s Forum 3.51
Logistics 0.37
Fire Engineering 3.44
Applied Geology 0.43
Number of Formalized Knowledge Postings
Structural 370

Mechanical 18

---

Document Outline

• Abstract
• INTRODUCTION
• Points of Departure
  • Knowledge Management
  • Contingency Theory
  • Knowledge Dimensions and Transfer Mechanisms
    • Half-life of Knowledge
• Research Methodology
• Preliminary Results and Discussion
  • Contextual/Tacitness
  • Half-Life of Knowledge
• DISCUSSION/CONCLUSION
• References

---