Practices Underlying One Piece of the Learning Organization
David I. Levine & April Gilbert
management" is something of a fad in today’s business press. At
the same time, it addresses key issues that can lead to success
within organizations. Knowledge management addressed how organizations
can manage the knowledge embedded in its systems, and contained
in the heads of its employees. It also systematically considers
how advanced information technology can be used to leverage existing
knowledge and create new knowledge. This essay discusses one aspect
of knowledge management, knowledge transfer: how to move good ideas
from one part of an organization to others that can use the information.
The Growing Importance of Knowledge Transfer
Knowledge transfer has always been a challenge for organizations.
Its importance has grown in recent decades for three related reasons.
First, knowledge appears to be an increasing proportion of many
organizations’ total assets. Second, organizations have moved away
from hierarchical methods of control toward more decentralized organizational
structures and increased employee involvement (Levine, 1995.). This
has resulted in more creativity by frontline employees and subunits,
but fewer obvious organizational paths through which the transfer
can occur. Finally, advances in information technology have created
new means of knowledge transfer. Innovations such as Lotus Notes,
the Internet, and intranets all hold the potential for increased
diffusion of innovations. However, technology alone cannot solve
the problem of knowledge transfer; organizational structures and
practices must facilitate and motivate transfers.
Knowledge transfer is only valuable when it is integrated into a
set of policies for knowledge generation and capture. In what follows,
we analyze the process of knowledge transfer and outline steps that
managers can take to increase information flow within their organizations.
The Components of Knowledge Transfer
In principle, knowledge transfer can be broken down into distinct
stages. We've chosen five steps to describe the process: idea creation,
sharing, evaluation, dissemination, and adoption. These stages often
overlap, are combined, or are skipped; they also have important
1) Idea creation:
Can Promote Knowledge Transfer
A massive literature exists on how to promote creativity. Robert
Sutton has studied creativity in groups and offers the following
list of questions to ask when assessing a group's potential for
- Is the
knowledge in the group varied enough?
- Is the
group’s attitude about its knowledge include respect for what
it knows and searching for what it does not know?
- Does the
group know how to fight so that new ideas are encouraged?
- Does the
group engage in constant experimentation?
- Does the
group’s status order support innovation, or do a few bosses
2) Idea sharing:
In practice, sharing (step 2) is often combined with validation
and dissemination (steps 3 & 4). For example, a work group might
share its ideas in a meeting, where their merits are discussed
and relevant potential adopters hear the new methods. Here, sharing
refers to the need to expose others to the idea in order for it
to be evaluated. Dissemination takes place once
the idea has passed some minimum level of evaluation.
For information sharing to occur, two conditions must be satisfied.
First, ideas must be in a form that others in the organization
can interpret. Dissemination is easier when the knowledge can
be made explicit or formal. For many skills and ideas, this involves
transforming the idea into a codified, often written, format.
Tacit, or informal, knowledge can be shared as well but the means
of sharing are different, requiring face-to-face contact and opportunities
for experiential learning. Apprenticeships often follow this time-intensive
and sensory-rich means of transmitting knowledge. Nonaka has emphasized
the rich interactions between tacit and explicit knowledge (1994).
While conventional wisdom on why knowledge is difficult to transfer
within firms has focused on motivational barriers, Szulanski (1996)
found that features of the knowledge itself and the receiver's
inability to interpret it were two of the most important factors
in inhibiting knowledge transfer.
The second condition required for sharing to occur is that employees
with ideas must be willing to share them. Sharing takes place
at multiple levels, with overlapping but distinct concerns: from
a worker to a workgroup, between workgroups, between departments,
between business units, and between organizations. Unsurprisingly,
Szulanski (1996) found that when the relationship between the
source and recipient was distant or problematic, knowledge transfer
was more difficult.
3) Idea evaluation:
Far more ideas exist than good ideas. Thus, organizations must
evaluate their new ideas -- see whether they have worked
in the past, are likely to work at new places, and actually work
at new places. Employees must have the capability, incentives,
and structures to perform the validation studies. At Xerox, for
example, skilled technicians evaluate new ideas; the best are
added into a best practices database for others to learn from.
4) Idea dissemination:
In principles, more information is better than less. At the same
time, too much information creates overload. The Internet is a
classic example, where nobody can read even a fraction of what
is there. The key to disseminating knowledge is that people
receive it who can use it. Several solutions exist to targeting
information, ranging from the primarily technological to the purely
5) Idea adoption:
In the best of all worlds, if people knew the right thing to do,
they would do it. However, we are not in such a world. Scholars
of organizational inertia have developed complex theories of why,
even after knowledge has been transmitted to the right
people, it may not have been transferred to the organization.
These theories fall into the categories of inadequate capability
(known as "absorptive capacity" in the literature), poor incentives
(the famous "not invented here" syndrome), and inadequate structures
(for example, rigid operating procedures that are difficult to
section outlines how managers can encourage knowledge transfer within
an organization through the use of training, incentives, organizational
structures, and technology. Under each section, we outline steps
that will promote each of the stages of knowledge transfer outlined
To effectively generate new ideas, employees need to be trained
in problem solving, including an ability to think "outside the box."
A typical program includes how to identify problems, prioritize,
analyze root causes, identify possible counter-measures, implement
the solution, and check whether the solution actually works. Companies
must also provide people information on the business and its environment
so their ideas are appropriate. In addition, employees need modern
organizational skills such as how to work effectively as a team.
To share articulated or explicit knowledge, workers need
to be literate in the languages in which ideas are expressed in
their work. In addition to spoken and written language such as English,
this may involve high-order "literacy" in more technical languages
such as blue prints or statistics.
Managers and workers must be trained to evaluate new ideas.
Just as importantly, they must be trained in systematically understanding
what evidence should be convincing -- for example, the difference
between correlation and causality, and the problems of small samples.
As everyone who has ever studied statistics knows (and especially
everyone who has ever taught it), these basic concepts are often
difficult to apply in practice. Once these basics have been mastered,
formal procedures such as statistical process control and the design
of experiments can be useful in creating new knowledge. Importantly,
for most employees and managers, statistical and problem-solving
training will usually be more effective if it is coupled with resolving
an actual problem, instead of classroom training in statistics.
Training workers to both disseminate and adopt new
ideas may revolve around making them aware of where else in the
organization their ideas may be useful and where else ideas may
arrive from. Workers must also know how to use technology to post
and search for new ideas. A receiver's ability to understand an
idea, "absorptive capacity", can be a barrier. This can only be
resolved through increasing the worker's own knowledge base, requiring
an increased emphasis on substantive ongoing education and training.
One difficulty with existing training efforts is their lack of integration.
To be most effective, training on creativity should include designing
solutions that include opportunities for validation and dissemination
To create an environment that encourages the generation of
new ideas, managers should consider the following policies: incentive
pay for ideas generated by groups or individuals; no layoffs for
productivity improvements that follow from new ideas; job duties
that include tinkering; permitting or encouraging experiments that
are well-conceived but fail; and giving credit to employees who
generate new ideas.
Employees are most likely to spend energy sharing what they
know if they are in a single workplace with group incentives. Thus,
extra incentives can be helpful when employees are in different
units without common objectives. Both monetary rewards and recognition
can prompt people to be more open with information and can create
corporate cultures in which sharing of information is valued.
For example, at Buckman Laboratories (See www.buckman.com), everyone
sees who answers problems on the open bulletin boards. Those who
contribute to solving company problems in public are praised, those
who do not become conspicuous. Bob Buckman emphasizes the benefits
of there being "no place to hide". Similarly, when Jack Welch, CEO
of General Electric, sees a new idea, he always asks: "Who else
knows about this?" People know that their reward for cleverness
depends on being able to explain how their idea has been shared.
Managers can also be rewarded for subordinates’ participation. For
example, at NUMMI first-level supervisors’ job evaluation depends
in part on their subordinates’ participation in the suggestion program.
In order to encourage not only sharing but also evaluation
and dissemination of ideas, knowledge-creating divisions
must be rewarded for creating knowledge that other divisions use.
Corporate headquarters cannot monitor the value of the knowledge
transfer between units, or even whether any knowledge is shared.
Knowledge-creating divisions face costs of creating an idea, posting
it to the corporate computer network, posting it carefully (for
example, avoiding division-specific jargon, being complete, creating
helpful keywords, providing appropriate pointers to people who can
supplement the report), and helping the knowledge-using unit implement
One idea is to pay for each posted idea. This promotes quantity
but not quality of ideas and provides no incentive for idea creators
to help adopters in implementation. A more complex alternative is
to pay for the measured quality of each idea. This provides better
incentives for quality ideas but is expensive due to the costs of
evaluation. In addition, there is still no incentive for idea generators
to help adopters. A third alternative is to pay bonuses based on
knowledge-using units’ claimed results. Variants on this process
include having knowledge-using units nominate knowledge-creating
units for internal awards, or giving each knowledge-using unit a
fixed number of prizes it can award to knowledge-creating units
that help it out.
Adoption depends in part on validation because ideas that
are clearly effective are more likely to be adopted. But even effective
ideas are sometimes not adopted and there are several psychological
reasons for this. Potential adopters may find it hard to believe
that one’s own ideas are not better than those from elsewhere. In
addition, many people find it difficult to see the applicability
of ideas from elsewhere because understanding how ideas can work
in new contexts can be difficult to perceive. Finally, it can be
embarrassing to say others did it better since rewards typically
go to "can do" people. These tendencies may be reduced if people
are exposed to lots of stories of stolen, adopted, and adapted ideas,
and of those using these techniques are acknowledged and rewarded.
The most important structural component that encourages creativity
or idea generation is often providing time to experiment
and tinker. This may run counter to other productivity measures
that emphasize efficiency. Also, formal employee involvement structures
such as brainstorming, suggestion programs, quality circles, and
self-directing teams support both creating and sharing knowledge.
People need the power and the responsibility to make improvements.
Another key element is to make the knowledge explicit. Many Japanese
firms stress the importance of formalizing knowledge, turning it
from tacit to explicit. At the same time, these same firms often
stress the importance of being "on site" so one can use most of
the five senses to understand a problem (MacDuffie, 1997).
To promote evaluation, companies must institutionalize means
of learning from past experience: "Companies must review their successes
and failures, assess them systematically, and record the lessons
in a form that employees find open and accessible. One expert has
called this process the "Santayana Review", citing the famous philosopher
George Santayana, who coined the phrase 'Those who cannot remember
the past are condemned to repeat it'." (Garvin, 1993)
A variety of organizational structures can promote the dissemination
and adoption of ideas. Despite the current emphasis on technology
and new methods of idea transmission, much knowledge remains tacit
and is most efficiently transmitted in person, renewing the importance
of decidedly low-tech practices including job rotation across units,
cross-functional meetings, cross-unit or cross-group meetings (e.g.,
sales convention), mentoring, training, and free time during coffee
What technologies support knowledge transfer? How can we integrate
new technologies such as Intranets, groupware, the Internet, with
other managerial practices?
One way in which technology may promote idea generation is through
its ability to provide information, including real-time tracking
of results, and communication with customers.
Groupware promotes sharing by tracking the status of ideas
and communicating them across a group, or further with wide-area
networks. Technology helps the quick evaluation of new ideas
by capturing actions and transactions and computing their effectiveness.
Personal computers can assist through the use of statistics.
Technology can help with the dissemination of ideas by making
it easier to target appropriate recipients such as 1) a group defined
formally by a common product, job title, or project, 2) a group
formed by management, or 3) an ad hoc group formed by workers such
as a mailing list, with either public or private membership. As
the technology develops, the groupware itself should help determine
who is likely to need a piece of information. Some examples are
bulletin boards, Web pages, and newsgroups where people self-select
to read. Complementing newsgroups and email, Internet "push" technology
(currently implemented by Pointcast, for example, on the World Wide
Web) deliver news of the sorts requested by users.
Those adopting new ideas can use email to communicate with the disseminators
of new ideas and ask for help in implementation.
Conclusion: The Importance of Integration
Organizations worry a lot about promoting creativity and innovation
but, in many cases, useful ideas already exist in some form. The
key is to capture the existing knowledge from within, and outside,
the organization and adopt those ideas that are relevant. This essay
steps through five stages of knowledge transfer: creation, sharing,
evaluation, dissemination, and adoption. For each stage, we examine
the how training, incentives, structures and technology can be used
to enhance the process. In order for an organization to be a true
"learning organization", it must acknowledge the importance of all
phases of knowledge creation and transfer and endeavor to create
a culture of sharing and continuous improvement.
Focusing on some stages but not others is less effective than moving
along with all stages in an integrated fashion. Creating knowledge
but not sharing it, or finding that other groups cannot learn it,
makes knowledge creation less relevant. If knowledge is transferred
successfully, but not first validated, then lots of costly fads
will sweep companies. Finally, true integration involves self-reflection
-- doing cost-benefit and cost-effectiveness analysis, and continuous
improvement of the learning and knowledge processes.
Garvin, David. Building a Learning Organization. Harvard Business
Review, July-August 1993
Levine, David I., Reinventing the Workplace: How Business and Employees
Can Both Win, Brookings Institution, Washington DC, 1995.
MacDuffie, John-Paul, "The Road to ‘Root Cause’: Shop-Floor Problem-Solving
at Three Auto Assembly Plants," Management Science, 43, /4,
Nonaka , Ikujiro, A Dynamic Theory of Organizational Knowledge Creation,
Organizational Science, Vol. 5, No. 1, February 1994
Stewart, Thomas A., Intellectual Capital, Currency/Doubleday,
Staw, Barry, editor, Creativity in Management, special issue of California
Management Review, v. 40, n. 1, Fall 1997.
Szulanski, G. "Exploring Internal Stickiness: Impediments To The Transfer
of Best Practice Within the Firm," Strategic Management Journal,
v. 17, Winter 1996, 27-44.