Originally published on May 23, 2016
The late Frank Leistner generously gave me a copy of his book, Mastering Organizational Knowledge Flow: How to Make Knowledge Sharing Work. Christopher Parsons quotes from it in What is “knowledge flow management?”
After a long time of playing with alternative terms, the one that actually fits best with my understanding is knowledge flow management, because the thing that you can manage is the flow of knowledge.
When I started playing with the notion of knowledge flow, the analogy of knowledge flowing through the organization like a river flowing through its bed seemed to fit for a number of reasons. Flows find their own way, but they can also be guided and stopped by barriers. You can have some individuals steering the direction of the flow on a daily level and others providing the main bed of the river by setting strategic goals for the longer run.
From Implementing a Successful KM Program, the five steps to follow for starting a KM program are:
- Create a Top 3 Objectives List of challenges and opportunities which your KM program will address. These objectives align business direction with program goals.
- Provide 9 Answers to questions about people, process, and technology. This information defines who will participate, which processes will be required, and how tools will support the people and processes.
- Define the KM Strategy. These are specific actions which will be taken to implement the program.
- Gain the sponsorship of your senior executive through The 10 Commitments. These commitments from the leader of your organization will enable the KM strategy to be implemented.
- Create and execute the Implementation Plan. This plan spells out the details of implementing the initiative. Contained in the Implementation Plan are program governance; desired modes of knowledge flow; people, process, and technology component selection; and implementation plans for some of the components, such as training, communications, and change management. Each one of these needs to be followed as part of implementing the overall plan.
Part of creating and executing a knowledge management program plan is implementing people, process, and technology components that will achieve your top 3 objectives. In order to do so, first think about which types of knowledge flow are needed.
There are five key ways in which the flow of knowledge can be tapped:
- Collection: processes and repositories for capturing explicit knowledge. This involves attempting to codify and encapsulate knowledge in writing or some other form of stored data.
- Connection: collaboration, communities, and social networks for sharing tacit knowledge. Connecting people allows them to exchange knowledge by communicating with one another.
- Boundary spanning: bridges across organizational boundaries for enabling knowledge to flow between previously-isolated groups. Building bridges to connect otherwise unconnected networks makes available previously unknown sources of knowledge.
- Discovery: processes for learning from existing sources of information, including systems, databases, and libraries. Scouring established knowledge bases in order to gain insights, distill trends, and uncover useful nuggets can provide a competitive advantage.
- Creation: processes for stimulating innovation and facilitating invention. By using the other modes of knowledge flow, creative ideas can be developed into useful new products, services, and ways of getting work done.
Putting knowledge to work in order to solve a problem, save time, make a sale, inspire innovation, improve quality, lower costs, increase profits, meet customer needs, and otherwise improve the world requires knowledge to flow between people.
In The Wealth of Knowledge: Intellectual Capital and the Twenty-first Century Organizationby Thomas Stewart, Chapter 8 is “A New Agenda: Managing Knowledge Projects.” On page 175, Stewart makes the following important point.
Connection, not collection: That’s the essence of knowledge management. The purpose of projects, therefore, is to get knowledge moving, not to freeze it; to distribute it, not to shelve it.
Chapter 6 is “The Case Against Knowledge Management.” On page 116, Stewart describes “the Kraken,” a Lotus Notes email list for general questions and answers.
The founders imagined that people would spark discussion by uploading white papers and the like — that is, they expected that users would pile logs of content in the fireplace, generating fire in the form of questions, critiques, and the like. Instead, the spark comes first: 80 percent of Kraken traffic starts with questions: Does anybody know? Does anybody have? Has anybody ever done something like?
The Kraken differs from KnowledgeCurve. The latter is supply-side; it’s full of documents, artifacts, and other explicit knowledge… The Kraken’s a conversation; KnowledgeCurve and its cousins are compendiums. KnowledgeCurve is about teaching; the Kraken is about learning.
In Volunteer not conscript, Dave Snowden writes:
Many years ago I formulated three rules or heuristics of Knowledge Management:
1. Knowledge will only ever be volunteered; it cannot be conscripted
2. We only know what we know when we need to know it
3. We always know more than we can tell, and we will always tell more than we can write down
The first of these reference the fact that you cannot make someone surrender their knowledge in the way that you can make them conform with a process. It was originally coined in reference to individuals, but I have come to realize that it also applies to organizations… So a new formulation… of the first rule would be:
If you ask someone, or a body for specific knowledge in the context of a real need it will never be refused. If you ask them to give you your knowledge on the basis that you may need it in the future, then you will never receive it.
For more about this from Dave Snowden, see:
- Volunteer not conscript — full
- Complex acts of knowing: paradox and descriptive self-awareness — condensed
- Complex Acts of Knowing: Paradox and Descriptive Self-Awareness — full
Many KM programs emphasize capture too much — collecting lots of documents, but not being able to effectively reuse them. So which types of knowledge flow should you emphasize, and to what extent? Here is a look at each type and when to incorporate it.
1. Collection: processes and repositories for capturing explicit knowledge
Explicit knowledge is formal knowledge that can be conveyed from one person to another in systematic ways. Examples include books, documents, white papers, databases, policy manuals, email messages, spreadsheets, methodologies, multimedia, and other types of files.
Based on the points made by Tom Stewart and Dave Snowden, it is reasonable to question the value of devoting significant energy to document collection in advance of a need. But there is still value in capturing some information in easily-retrievable repositories.
For example, before beginning a new project, it is useful to ask the question “has anyone ever done anything like this before?” If information on all prior projects has been collected in a searchable repository, then this question can be answered. Not all of the documents created by previous projects may have been captured, but if the names of the project team members are available, then it is possible to contact them to find out more and to request any relevant documents. This is an example of how collection and connection can work together to deliver important knowledge at the time of need.
Another example of how collection and connection complement one another is asking a community for help. In responding to a request from one community member, another member can point to a previously-stored document which meets the needs of the first member.
One way of minimizing the need for collection is to use connection to identify a need and then respond with a document only upon such a request. Another way is to rely on discovery to ferret out information from existing databases such that additional collection is not required. For example, if information on previous projects is automatically captured as part of the organization’s business management system, then it can be retrieved without the need for additional data entry.
Collection provides the supply side of knowledge. If you decide that it is needed, try to keep it to the absolute minimum needed to support the Top 3 Objectives List. Rely on other modes of knowledge flow as much as possible. And be sensitive to Dave Snowden’s point: “If you ask them to give you your knowledge on the basis that you may need it in the future, then you will never receive it.”
2. Connection: collaboration, communities and social networks for sharing tacit knowledge
Tacit knowledge is personal knowledge that resides in an individual. It is content that has not been recorded or exchanged. It relies on experiences, ideas, insights, values, and judgments and usually requires joint, shared activities in order to transmit it. Individuals possess tacit knowledge and must learn to verbalize that knowledge. The art of talking about a problem or opportunity causes it to take shape and to be defined. Once defined, it can be solved or developed.
Dave Snowden wrote “we will always tell more than we can write down.” And according to Tom Stewart, “connection is the essence of knowledge management.” So this mode of knowledge flow should be a key part of your KM plan.
Connection supports the demand side of knowledge. It enables demand-driven or just-in-time knowledge management.
Dave Snowden asserts: “If you ask someone, or a body for specific knowledge in the context of a real need it will never be refused.” And Tom Stewart states: “80 percent of Kraken traffic starts with questions: Does anybody know? Does anybody have? Has anybody ever done something like?”
This argues for including communities and threaded discussions or enterprise social networks (ESNs) in your selected list of KM components. Communities are the people who connect, and threaded discussions/ESNs are the mechanism for the connection. These should almost always be part of any KM program.
3. Boundary spanning: bridges across organizational boundaries for enabling knowledge to flow between previously-isolated groups
In Building Smart Communities through Network Weaving, Valdis Krebs and June Holley define boundary spanners as “nodes that connect two or more clusters — they act as bridges between groups.” They go on to observe: “When left unmanaged, networks follow two simple, yet powerful driving forces: 1. Birds of a feather flock together. 2. Those close by, form a tie. This results in many small and dense clusters with little or no diversity. Everyone in the cluster knows what everyone else knows and no one knows what is going on in other clusters. The lack of outside information, and dense cohesion within the network, removes all possibility for new ideas and innovations.”
To overcome this tendency, it is important to make explicit efforts to establish links between different groups. Examples include different regions of the world (e.g., North America, Latin America, Europe/Middle East/Africa, and Asia Pacific), functions (e.g., engineering, manufacturing, marketing, sales, logistics, and service), business units (e.g., paper products, cleaning products, and health products), roles (e.g., interns, retirees, and contractors), and organizations (e.g., employees, customers, and partners).
An example of how boundary spanning can help overcome organizational barriers is product development and introduction. Marketing tells Engineering to develop a new product to meet a customer need. Engineering designs the product, which is produced by Manufacturing. Marketing promotes the product, which is sold to customers by Sales and delivered by Logistics. Service installs the product, and it repairs it if the customer has a problem with it. A community focused on a specific product which includes members from all of these functions can help them collaborate across boundaries.
One of the following collaboration conditions typically exists in an organization. These are listed in increasing level of connectedness:
- There are no communities. Small work teams collaborate, but there is limited collaboration beyond the teams.
- There are some communities within functions. For example, a community of engineers who help each other out with designs.
- There are some communities which span some functions. For example, a community for engineers and service people for a specific product.
- There are some communities which span all functions. For example, a community with everyone involved in some way on a specific product.
- There are communities for all offerings which span all functions, and include customers and partners. This is true boundary spanning.
In the implementation plan, identify all groups which need to connect, and include boundary spanning as a required knowledge flow. The higher the level of connectedness you can achieve, the more knowledge will flow between groups. You can use social network analysis to help determine the current state of social networks and to identify boundary spanning opportunities.
4. Discovery: processes for learning from existing sources of information, including systems, databases, and libraries
In most organizations there are information systems, transaction processing applications, and databases which are used to run the business. There is data captured in these systems which can be used to distill trends, answer queries, and support decision making. And this can be done without the need to capture data redundantly. For example, if customer purchase information is entered into the order processing system, it can be fed to a data warehouse for use by all departments.
Many organizations have libraries of information obtained through outside sources. These may include competitive intelligence, analyst reports, industry news, and benchmark data. Providing access to this information supports analysis, strategy formulation, and planning. If such libraries do not exist centrally, you should consider providing them to prevent individual departments from purchasing information on their own. If they do exist, then your plan should incorporate them into the resources provided through the user interface.
Many of the 100 Knowledge Management Specialties can be used to support discovery, including:
- Analytics, text analytics, visualization
- After Action Review, sensemaking, ritual dissent
- Appreciative inquiry, positive deviance, Most Significant Change
- Big data, databases, repositories, business intelligence, data warehouses, data lakes
- Competitive intelligence, customer intelligence, market intelligence, research
- Cognitive computing, artificial intelligence, natural language processing, machine learning, neural networks
5. Creation: processes for stimulating innovation and facilitating invention
Creating new knowledge is an important goal for most organizations, but it is difficult to enable. By using the other modes of knowledge flow — collection, connection, boundary spanning, and discovery — and adding explicit processes to use these flows to create knowledge, innovation can be stimulated.
Let’s look at an example. In a consulting firm, information about customer projects is captured in a repository (collection). Communities for each type of consulting service are active (connection), and include consultants, partners, contractors, and sales people from all regions of the world (boundary spanning). Details on the win rate, delivery time, and profitability of each service offering are available in a data warehouse (discovery). Competitive and industry trends are available in a corporate library (discovery).
The leadership team has been asked to increase the gross profit margin of the consulting business. They take the following steps:
- Search the project repository to see which customers are doing the most repeat business. Survey those customers about their upcoming needs.
- Ask the communities for each service offering to offer their suggestions for improving profits. Select the best ones for implementation.
- Analyze the information in the data warehouse to see which service offerings are the most and least profitable. Improve on the profitable ones and develop new offerings with similar attributes. Discontinue the unprofitable ones and deny approval to future proposals for offerings with similar attributes.
- Review competitive and industry trends to see which competitors’ offerings are the most profitable and what the analysts predict will be profitable. Use these findings to help shape new development efforts.
- Combining all of these inputs, the leaders decide to drop their three worst-performing service offerings, invest in further developing their top three, and decide to create two new offerings based on customer input, community feedback, and analyst predictions.
By institutionalizing the process used in this case, a knowledge creation process can be reused for future innovation. It is not simple or intuitive to create new knowledge, but it is worth perfecting because the potential benefits are significant.
Examples of Knowledge Flow
Demand-driven knowledge sharing, which can also be called just-in-time knowledge management, emphasizes connection instead of collection. It assumes that knowledge will be provided at the time of need.
Here is how it works. Someone has a question, problem, or need to know who, what, when, where, why, and/or how about a topic. They search existing repositories and threaded discussion archives to see if there is an existing answer. If so, they use it. If no answer is found, they post their question, problem, or need to one or more relevant threaded discussions. Other members of the threaded discussion respond with their answers. The answers may include links to content in other repositories. The answers are automatically archived so that future searches will produce useful results.
Tacit knowledge can be shared through connection, and it can be turned into explicit knowledge through collection. Communities and social networks are the usual mechanisms for sharing tacit knowledge.
Here is an example. Someone wants to share an insight, a nugget of knowledge, or a solution to a problem which others may face. They post to a relevant threaded discussion. They may choose to write up their knowledge more formally, thus turning it into explicit knowledge.
Explicit knowledge is captured through collection and shared through connection. Repositories are typically used to capture this form of knowledge.
For example, someone wants to share reusable content such as a document, presentation, recording, process, procedure, template, tool, software source code module, or some other form of data. They upload the file containing the content to the appropriate repository. They post to the related threaded discussion to let the members know about the file, including a link to it.
By considering all five modes of knowledge flow, and incorporating them appropriately into your plans, you can decide how to enable and support all needed flows. This will be incorporated in the corresponding people, process, and technology elements which you design as part of the program. And encouraging demand-driven knowledge sharing can help knowledge flow in very practical ways.
What other types of knowledge flow have you seen?