Sharing Knowledge and Organizational Learning

Actually, this is yet another extract, because I found out that it is also relevant to the topic of this blog, and because I want to be able to quote it in the future ...

On the management literature best-seller is P. Senge’s book The 5th Discipline, which I have quoted earlier. This great book focuses on the collective intelligence of an organization (and, by consequence, collective learning). This leads naturally to the concepts of collective skills and collective knowledge. Looking at analysis and thinking as collective activities is not as easy or simple as it sounds. This type of collective behavior emerges in companies as a response to the increasing demands and the increasing competition. A major part of his book is devoted to the analysis of systems and their complexity, which P. Senge considers to be the most significant issue of organizational (team) learning[1]. Sharing knowledge is a topic which is neither simple nor intuitive. There exists a wealth of dubious quotes or proverbs that claim that sharing knowledge is a way to get richer, that knowledge is the only wealth that is not lost when given away, and so on. In practice, sharing one’s knowledge is difficult (which is why one must learn to become a teacher), it is time-consuming (thus it is nothing but free) and it may yield to feel a loss of status to the one who is “sharing” away her knowledge. A necessary condition is to establish a culture that promote knowledge passing, both in the general sense (as a principle) and in a practical sense (so that the time which is spent is recognized and rewarded). The fame of “knowledge management” has declined over the years since it has been associated with heavy and costly projects, which came with new technologies that appeared overly complex. The “knowledge management” (KM) expression is used to designate both a process (which is our focus here) and a project. Ten years ago, projects to extract knowledge out of a group of specialists (through interviews to build a so-called “knowledge base”) were quite popular. They are still useful and relevant in a few specialized cases, but this approach is too heavy and too static to be applied in a continuous manner[2]. The next figure is an attempt – some readers may find it unduly abstract or complex – to represent the complex transformation which an enterprise must go through, as explained in this last section. There are two parts: the left part is a “traditional view” of the enterprise, the right is a newer vision centered on collective intelligence. The left part represents a traditional perspective on the employee, who is appreciated for his actions. We use a classical representation of knowledge as a process which transforms information into action (small rectangular box associated to each employee on the figure). In a traditional perspective, knowledge is a personal matter (even if some information is shared). The domain of the company is the coordination of individual actions into collective ones, which is represented, in Figure 6.3, with a dashed-line triangle. The “bubble” on top of the figure represents the “governance mode”, which is the “control-command” approach that we discussed earlier. The right part of the figure represents the result obtained when the transformation towards collective intelligence is achieved. The first difference with the previous situation is that knowledge is no longer an individual matter. A new part of collective knowledge (the grey rectangle in Figure 6.3) has appeared, which is precisely the field of knowledge management. Since the company is immersed into a constantly evolving world, this collective knowledge needs to be continuously revisited and updated. This continuous activity (circle arrow) is the knowledge management process’s goal. Let us notice that, when the surroundings change constantly, knowledge management and organizational learning become synonymous. The goal for companies is not simply to “know the right things” but to be able to “learn all the necessary things”. The second difference with the left part lies with the distributed management style (illustrated with the bubble). Top-down hierarchical command is replaced by a dual flow: a situation occurs that triggers a reaction, while a common goal influences the way situations are analyzed. The company goals are re-evaluated according to the situation changes. Moreover, they may be interpreted locally by “field managers” who enjoy more responsibilities. A distributed control emphasizes the interest of managing collective knowledge, which supports the coherence control that may no longer be enforced by a centralized and autocrat vision.

Figure 6.3: Collective Management of Knowledge and Learning

As far as the process of knowledge management within the enterprise is concerned, we may summarize the key principles as follows:

• Knowledge capitalization is mostly an internal communication process. This is why tools that support “practice communities” (cf. the previous discussion about blogs) are playing a key role.
• A global management of knowledge throughout the company is made necessary by the distribution of management and decision taking[3].
• Structuring collective knowledge is an emerging process, which is not directed towards goals, but, in the opposite direction, yields opportunities[4].

Managing knowledge may not be separated from skill management, nor, more generally, from human resource management. Managing skills require producing assessments, to draw skill maps and to articulate a growth strategy. In a way that complements the emergent process of knowledge management, skill management is better carried in a planned and top-town manner. Necessary skills may be bought, recruited or trained … All this takes time, on a multi-year scale. Skill management is an operational responsibility that devolves upon business divisions, but which requires the methodological assistance from the human resource division. This is not simply because skills are a person-related matter, but also because the time scale that is required to implement a strategy is too long for a purely operational perspective.

[1] Peter Senge makes a key difference between detail complexity (many sub-parts) and dynamic complexity (many steps for the associated processes). Chapters 6 and 7 provides with a clear illustration of what « system thinking » (a favorite expressions from this book) is. On the other hand, Chapter 12 deals with « team learning », which is precisely about organizational learning. To P. Senge, organizational learning is the fifth step of a maturity scale, which follows more classical maturity stage that are borrowed from total quality management.
[2] On this topic, see the previously quoted book from E. Awad et H. Ghaziri, Knowledge Management, which is very thorough, even though on the descriptive side. The reader will find there definitions for terms such as “information” and “knowledge”, as well as a complete taxonomy of knowledge, together with the associated extraction and capitalization processes.
[3] These ideas are brilliantly developed in L. Morris’s book, Managing the evolving corporation, which was introduced in the previous chapter. I especially recommend the fifth chapter which has been a key source for my own thinking. It explains the relationships between information, knowledge and decisions with much more detail than what is given here.
[4] About this fascinating but complex topic, one should read the book from F. Julien: “Treatises on Efficacy: Between Western and Chinese Thinking”. He starts with a brilliant comparison between strategic military thinking in Occident versus China. He then deals with the concept of efficiency and contrast our goal-directed strategies with « situation intelligence » which is the trademark of Chinese thinkers. A large number of fashionable theories, proposed in management books, about the « networked enterprise » may be analyzed and explained with this treasure of knowledge and erudition.

Organization and Information Flows

A last extract from my book, from the fourth chapter.

If organization is seen as an information transfer tool, it becomes obvious that both the « management system » (i.e., the hierarchical structure) and the «corporate meeting system » (i.e., the set of all planned meetings and committees) must be considered as one system. We shall take a closer look at the “corporate meeting system” in Section 4.5. The influence between the two is mutual: hierarchical management generates scheduled information sharing meetings (either as one-to-one or team meetings), committees are an alternate form of embodiment for decision and management power, as we just noticed. Using the meeting system to complement the hierarchical structure is much more flexible than a matrix organization, but one must be careful not to overuse it (cf. Section 4.5.1 as well as the next chapter) [1].

This chapter will present some key ideas about the relationship between organization and information flows, which may be gathered into three categories:

• A key feature of an enterprise organization is the information propagation latency (the time it takes for a piece of information to propagate). A major goal of organization re-engineering should be to produce a more reactive enterprise.
• A dual key notion is the “connectivity degree”, which we can define as the average length of the communication paths. A key principle from communication theory is that fidelity declines with the number of exchanges (i.e., intermediaries). A related goal for organizational architecture is to ensure the existence of short paths which may be used for high priority messages[2].
• Time is the most critical resource for information management. The act of communicating takes time. The duration may vary according to the channel that is used, the medium or the protocol, but communication is a process, not simply some information transfer. Many communication attempts fail when the receiving end simply does not have the time to process the information. An efficient organization must optimize global time management (share when possible, sort according to priorities, etc.).
  

An interesting consequence of this relationship between communication and enterprise organization is the impact of electronic communication channels (email, IM, …) on the enterprise management. The next chapter will focus on electronic tools such as email, intranets, phones, etc and their integration with the information system. A balance between face-to-face and electronic communication must be found, since electronic channels have their own advantages and limitations[3] (cf. Section 4.5).
The impact of internal communication on the enterprise operations and efficiency is, nevertheless, strongly dependent on the size of the company. Many of the issues that we raise in this chapter are of little significance for a 10-employees company. A very crude model would say that the work load (amount of time spent in activities) varies linearly with the number of employees, whereas coordination requires a share of the total time which is evolving in a quadratic manner. The same reasoning goes with process complexity: the coordination load grows faster than the simple number of tasks that need to be managed. A similar remark may be made about the opportunity, or the constraints, to share functions or resources within an organization. This sharing is only profitable if a critical mass is reached. The creation of a department associated with a special skill becomes relevant only when the coordination management load has reached a given threshold. Hence a major part of the dilemmas reported here only apply to large companies.
Speaking about company size is actually a gross simplification, since what matters is the size of the teams which are necessary to achieve the company’s goals and to operate the business processes. There are large enterprises which are heavily distributed, for instance according to geographical zones, or by projects (such as construction sites), for which operating teams remain rather small. On the other hand, service companies which reach a mass market often need to assign a large amount of its resources to each business process (that is, each process involves a fair amount of coordination between the different departments).
In this later context, coordination is often more time-consuming than processing the activity itself (the so-called “individual work”). That is to say, a “knowledge worker” spends more of her time passing information along in meetings than producing its own (information, or value addition). Such a behavior is often tagged with the French expression “bucket carrier” (a bucket carrier is someone whose role is limited to passing information around). Such an (implicit) criticism is partially unjust, since it is perfectly normal that a fair amount of time is spent transferring information in a large-scale project or a process (i.e., with a large number of participants). One way to reduce this coordination load is to reduce this number of participants through avoiding specialization (at least partially), which we shall discuss in the next section. Another approach is to optimize the management of information flows, especially through the corporate meeting system, which we shall discuss in section 4.5.1.
The issue of coordinating a large-scale team is made even more complex when dynamic variations occur. Load variations naturally create local over-capacity pockets. In a small structure, this over-capacity is resolved by capillarity: the under-utilized employee can find additional work in the same, local, environment. There is an implicit levelling of capacity by the needs. In a larger structure, on the other hand, there is a risk of “self-employment”: if over-capacity reaches a critical mass, it may generate its own activity. In some situation, this may turn into a “management” activity. Some employees focus on “improving” the work of others: they create models, scorecards, monitoring tools, etc. This translates into a “Brownian movement” around the enterprise main course. Projects become more complex, requirements grow richer, new avenues are explored.

[1] The time which is necessary to learn about one another and how to work together should not be underestimated. This crucial point is emphasized in the conclusion from B. Nardi’s article “Beyond Bandwidth: Dimensions of Connection in Interpersonal Communication », in Computer Supported Cooperative Work (2005), vol. 14, Springer. Here is a relevant quote: « The need for speed and cost saving encourages distributed work, necessitating mediated communication, and yet the clocks tick faster, the deadline grows shorter. The use of short term “virtual teams” and matrixed organizational schemes means workers have less time to get to know one another. We do not yet know the long term effects that attenuated social relations in the workplace may have, but there are certainly hidden costs involved.”
[2] This topic is explained brilliantly in M. Gladwell’s book “ The Tipping Point”. M. Gladwell’s book deals with « the propagation of contagious ideas » and the underlying social networks which are required for this propagation.
[3] This issue of information and communication technology may be placed in the context of transaction costs, as defined by Ronald Coase. The economy Nobel Prize, in his famous article “ The Nature of the Firm” reconstructs the enterprise concept from the « cost of transaction » principle. Transaction costs are smaller within a company, which creates a competitive advantage of internal collaboration as opposed to outsourcing. Information and communication technologies (ICT) have reduced these costs considerably, with consequences such as «networked company », outsourcing, « extended enterprise », etc. Optimizing these internal transaction costs through the better use of ICT is a major competitiveness issue.