30 April 2007

The search for the unified definition of Knowledge...

I could not resist!

Knowledge has nearly as many definitions as the number of authors who wrote about it or about a related subject such as Knowledge Management (by the way, you can find quite a few definitions of KM as well and maybe this explains that).

Well, guess what, I had to give “my” own definition!

On the ActKM listserve ( ) a still on-going debate on the definition of knowledge generated very valuable insights. I have extracted a few extracts and composed a definition attempting to synthesize the thinking of all these bright individuals:

Dave Snowden ( ) quoting Prusak and Davenport in their book “Working Knowledge”:

"Knowledge is a fluid mix of framed experience, values, contextual information, and expert insight that provides a framework for evaluating and incorporating new experiences and information. It originates and is applied in the minds of knowers. In organizations, it often becomes embedded not only in documents or repositories but also in organizational routines, processes, practices, and norms"

Joe Firestone’s definition ( ):

Knowledge is a tested, evaluated and surviving structure of information (e.g., DNA instructions, synaptic structures, beliefs, or claims) that is developed by a living system to help itself solve problems and which may help it to adapt.”

Han Van Loon’s version of Joe’s definition ( ):

"Knowledge is a learned and analysed structure of awareness based upon
information (e.g., DNA instructions, synaptic structures, beliefs, or
claims) that is developed by a living system

A suggestion of a synthesis:

Knowledge is a learned and evaluated fluid mix of framed experience, values, contextual information and expert insight; that is developed by a living system. In organizations, knowledge often becomes embedded not only in documents or repositories but also in organizational routines, processes, practices, and norms.
Living systems use knowledge either instinctively or in a state of awareness (typically the latter first then the former through acquisition of reflexes for repetitive actions) to compete for resources and survival, through solving problems, adapting to challenges and setting objectives.

Still on the ActKM listserver, Kaye Vivian ( ) writes:

The abstraction we call "Knowledge" has three aspects (I propose):
(1) what is acquired by learning
(2) the sum of everything known
(3) a state of awareness
and perhaps a fourth for those who argue that knowledge can exist in neural, hormonal or sensory systems:
(4) instincts

Kaye goes on then asking us all if we can think of an instance of knowledge that does not fit into any of these four aspects. Assuming for a moment that none can be found, I think that my suggested definition above does try to cater for all 4 of these aspects.

But hold on! Don’t get over excited! I surely have not suddenly stumbled on the unified definition of knowledge. I will post it on ActKM as well and I anticipate quite a bit of constructive criticism. I will keep you informed and will amend the definition accordingly.

So watch this space...

Peter-Anthony Glick

21 April 2007

Questions to Verna Allee on how to start a Value Networks analysis

This week I attended a (tel. conf.) presentation by Verna Allee on Value Networks. I then sent her 3 questions (in blue) and here are her answers below:

Can a VN approach be used to initiate a cultural change in an organizational context where hierarchy is prominent, departmental boundaries are strong and guarded and knowledge tends to be protected rather than shared? Or is it an approach only effective when the value of collaboration and knowledge-sharing is already recognized?

Absolutely it can be helpful in culture change. Value network analysis (VNA) zeros in on the most mission critical and essential intangible exchanges that support the work. This is not just "nice to do" stuff and it is not some vague encouragment to "share your knowledge." What people spell out, through conversations, are the specific deliverables and behaviors that they need and expect from each other in order to work effectively and build good relationshps. "Knowledge" is not a deliverable - it is an asset - but you must convert that asset to some negotiable form of value in order to put it into play to create value. You extend your knowledge in very specific forms: a report, professional advice, market intelligence, referrals, etc. VNA forces people to negotiate around intangibles such as various forms of knowledge in a very clear,specific direct way. Forget "knowledge sharing" - that doesn't meananything. "Timely input of market intelligence," is a specific knowledge output that can be delivered, can have performancestandards, and that someone can be held accountable for. The very process of this negotiation "loosens up" the knowledge flows because there is an intuitive sense of fairness and reciprocity that kicks inonce people talk about knowledge sharing in this way. It is a very interesting dynamic. You aren't beating on this big concept of "culture" but are focusing on the very specific behaviors that are essential for people to successfully work together. Making those "visible" in this way is very powerful, not only for affirming how important they are but also for making it much more likely thatthey will happen. For more tips on naming intangibles see ValueNetwork Mapping Tips at

2."What is the most effective method for collecting the informationneeded to build a VN diagram? Is it assumed that a relatively highproportion of subjectivity will be captured?"

There are two different approaches for different purposes. If you are using VNA as a collaborative sense making tool then facilitating the mapping as a group process is absolutely the way to go. You WANT the subjective nature of value to become clear to the group - this is a major "ah ha" and learning that can help a group of people dramatically reframe what they are doing and start thinking about value in a whole new way. It also is a beautiful way to surface unspoken expectations and the "mental models" of how people are really thinking it all works and about different issues. I have used it this way for years with great success. In this case most of what people need to discover happens in the mapping process, not in the deeper analytics.

In other cases you absolutely do not want the value network analysis to be skewed by subjective input - and you will want to run the more advanced analytics in order to see patterns and opportunities. This is especially true when you are doing something like a market space analysis or a large scale analysis like we are doing for the European Commission. There we are evaluating innovation networks across all 255 regions and 25 nation states. We now have the technical capability to use real corporate data sets to generate the value network visualization. For example we took a month of call data for all the calls handled by Cisco's Customer Interaction Network and used the data to generate the value network visualization. These kinds of data sets can be coded for tangible and intangible exchanges and the "real" network patterns are revealed. In the case of the European Commission we have identified 4 network archetypes within these large scale innovation networks depending on the real purpose of the network. In some cases the stated intent ofthe funded project was to do one thing, but the data shows that they actually created something different. Very interesting. The application I am referring to is the open source GenIsis application developed by my brilliant colleague Oliver Schwabe. The open source version uses an Excel-based workbook to capture and organize the data. If you are interested in the enterprise level database version you will need to contact me off-line, but the open resource version is readily available and you can learn about it, and download it. Some of you may also be interested in the value network data model that is being supported by the Value Networks Consortium, which is sponsored by companies like Cisco and is leading standards and open source tools for value network analysis. . Value network language and models are popping up everywhere and someof it is very good work, but there is also a lot of "junk" or old wine in new bottles masquerading as value network analysis. The consortium is a good way to stay grounded in the quality work and avoid some of the pitfalls.

3. "What is the typical scale and scope of the very first "proof> of concept" VN implementation for an organization? Scale meaning> how large is the section of the organization to consider. Scope> meaning how many processes to consider.

Where do we start? I suggest you start by focusing on a particular activity where you can easily identify somewhere between 4-8 roles that are required for that activity. The Mapping tips above have some tips on scope, scale and boundaries and there are other tip sheets in the How To Guides at . They are designed to address FAQs such as this one. That said, I just want to share that at the beginning of a workshop in the Netherlands a clearly skeptical partipant asked, "I want to know where I would actually use something like this." At the end of the day he said, "I get it - you use it with whatever is on the table." You can use VNA at virtually every level from shop floor (example Mayo Clinic looking at how to reduce scheduling time for a procedure) to business unit level (exampel AT&T getting ready for a product launch) to business webs (the eBay value network) or at the macro economic level as we already mentioned. Youdo need to decide what level you want to look at - ground floor, rooftop, helicopter, jetliner or satellite and not mix your level. Doread the "how to guides," sit down with a couple of friendlies and try it out. It doesn't do any good to just read about it - just do it. One of our practitioners suggests that people start by simply mapping their own most important role and their everyday key interactions. There are also a number of case studies on the site that will give you a good sampling of how people are using it. The community goal for that site is to really give people an on-ramp to learn all the trade secrets to the method and provide enough tools and examples that they can get started. Hope this is helpful. I am glad you enjoyed the presentation and encourage you to jump in and get your feet wet. Like anything there is a learning curve but it can be a very fast track to some good results. An executive with Openwave came to one three hour workshop on VNA, went back home and completely reorganized a company of three thousand people over the next two months using the method. He said it was the smoothest reorg they ever had with zero productivity loss the entire time. There is a slide deck on that in the case studies tab on the resource site, too. I appreciate that you asked your questions in a forum as they really are FAQs and this gets the word out to more people on where to find the support. Oh, yes, I always forget to mention we do have a commercial "deep dive" ValueNet Works Practitioner Qualification for a fee. You can learn about that at . You can tell I am a maniac on a mission because most of the time I forget to even mention the commercial offerings, but I have a cat (insert dog, child, partner) to feed just like everyone else. Our qualified practitioner community stretches literally from Iceland to Tasmania and lucky for you Peter there is a nice little "hotbed" of practitioners on the UK who are also very supportive of new folks.
Good luck.

Verna Allee

I will get my feet wet now and try out VN. I particularly like the idea that VN can be used to bypass the culture issue by focusing on the value-adding interactions. Knowledge-sharing is a byproduct of the process, not the objective. Knowledge is the asset, added-value the output.

Peter-Anthony Glick

09 April 2007

“Knowledge management strategies that create value”

I found a very good article with the same name as this post on the Accenture site. It was written in 1999 by Leigh P. Donoghue, Jeanne G. Harris and Bruce E. Weitzman.
( article) It presents a rather visionary KM approach considering it is now about 8 years old.

The article starts with this statement I totally agree with: “There is no one-size-fits-all way to effectively tap a firm's intellectual capital. To create value, companies must focus on how knowledge is used to build critical capabilities”. I would add that the more pervasive a Company’s organizational culture is, the more this is true. So many technological solutions have been presented as THE knowledge-sharing solution, and nearly as many have failed.

“[…] Knowledge management is complex and multifaceted; it encompasses everything the organization does to make knowledge available to the business, such as embedding key information in systems and processes, applying incentives to motivate employees and forging alliances to infuse the business with new knowledge. Effective knowledge management requires a combination of many organizational elements—technology, human resource practices, organizational structure and culture—in order to ensure that the right knowledge is brought to bear at the right time”. Well, this is what I (and many other KMers) have been writing for some time now. You cannot count on technology alone, or on a structural change alone, or on a new reward and recognition mechanism alone, to instigate a deep, long-lasting and effective leveraging of an Organization’s Knowledge. You need a holistic approach with both top-down leadership and bottom-up initiatives, being aware along the way that different core processes will require different KM solutions. The authors then present a framework created and used by the Accenture Institute for Strategic Change. Its aim is to associate “specific knowledge-management strategies with specific challenges that companies face”.

Well, my first impression of this framework presented this way was: whow! That looks simple (if not simplistic). I was reassured a bit when reading two paragraphs down that “[..]It is important to note that there are no hard-and-fast connections between a certain core process and a work model, because the same process can be performed in different ways”. In other words, you cannot actually plot core processes on the table above to build a model to fit all companies. This would also contradict the initial statement that there is no one-size-fit-all solution. So this is where the Accenture consultant comes in. The way the work is performed in the organization must be defined in order to select the right KM approach.

In the above diagram, the authors show how an Organization’s work processes can be aligned with a specific KM model.

I think this framework is approaching the issue in the correct manner, i.e. holistically and with a good deal of flexibility in order to adapt to any organizational context. However, there is a level of flexibility that I believe is missing. It could be that it was omitted by the authors in this rather short presentation. Nevertheless, I can only judge on what is given here. The flexibility that seems to be lacking is the consideration that within a specific work process, say Retail operations, you can be faced with a rather more complex context than what is assumed with the different diagrams given in this article. In the example above, the authors have assumed that Retail operations would be aligned to the Transactional Model. The authors define this model as the one “in which there is a low degree of both interdependence and complexity. Work is typically routine, highly reliant on formal rules, procedures and training, and depends on a workforce that exercises little discretion”.

Indeed, Retail relies on direct transactions with the end-customer. However, the definition above is valid in a mass-market context with low value, low margin, high quantity and relatively low product differentiation. Take instead the luxury market context (and I choose this example because I have 14 years of experience in it) with high value, high margin, low quantity and very high product differentiation. Within the Retail operations of a luxury products (and/or services) organization, you will find:

* A rather low degree of interdependence, so the Transaction model still fits for this dimension.

* There is relatively high degree of complexity. In the jewellery business for example, the expertise in gemmology of the sales-associate can represent the key added-value for the customer in search of a diamond necklace. Experience in how to satisfy very demanding and difficult customers is typically what can make a good sales-associate very good.

* Work is not “routine” to the same degree as in a mass market since each transactions can differ greatly due to the uniqueness of the product sold, the customer’s varying requirements and behaviours and the sales-associate varying level of expertise and experience.

* Work relies on formal rules and procedures but not exclusively. There is also a significant degree of informal relationships between sales-associates or between them and their customers, with whom they build strong relationships over time. Often, a sale is made as a result of this informality.

* Work does initially rely on training - especially for Brand and product knowledge as well as sale-techniques – but the best performers among sales-associates rely even more on their intuitions and experience.

* Work does depend on a workforce capable of making decision on their own, such as proactively contacting customers, deciding on which products to suggest to a customer, or offering/accepting a discount in a responsible manner.

The Expert model would therefore come to the rescue in this context but not as a replacement of the Transaction model. I am suggesting here that a combination of both models is needed to map a luxury business’ Retail operations. In this Retail context, there is still a degree of “routinization” and automation, and there is a definite “productization”. However, there is also a significant need for experienced hiring and capability protection. Capability/skill development is also a concern. (Apprenticeships used to be commonplace in luxury retail businesses some 20/30 years ago, but was replaced by a more individualistic and internal competition-oriented approach. I foresee that it will come back as a result of more knowledge-conscious management – read my earlier post on the subject:

Now what would such a mix of these two models mean in terms of practical solutions? The authors do not provide (for obvious reasons) the list of KM solutions they would implement for each model. However, I can guess one here.

The degree of “routinization” involved in luxury retail operations would demand solutions delivering just-in-time information (as opposed to just-in-case) to the sales staff. This could be in the form of a CRM tool providing a sales-associate specific information about an unfamiliar but regular customer sitting in front of him/her. It could provide the list of all the products the customer purchased so as to enable the sales-associate to suggest matching products among new or older collections. It could also have the anniversary dates such as the customer’s or his/her partner’s birthday, or their wedding date; for the sales-associate to wish him/her and suggest suitable gift ideas. All this customer-specific information is valuable but it can really create significant value when it is associated with context-sensitive information – in effect, offering expert knowledge. This is where our authors’ Expert model comes in: In the situation above, our sales-associate would benefit from the relevant knowledge of a more experienced colleague. More experienced here does not necessarily mean more seniority; it can mean better specific knowledge about the customer being served, or even about the customer’s cultural background. What is needed is therefore an apprenticeship-like solution associated (or better integrated) with the CRM tool. For example, the sales-associate could be informed that the customer is of Indian origin and Hindu, with the “warning” that between August and October, all Hindus purchase gifts (and in particular luxury products) to offer on Diwali (their annual “festival of lights”). The sales-associate could happily then suggest: “Oh Diwali is coming soon isn’t it? Please let me show you this brand new collection of jewellery that should look stunning when worn with a sari” (typical Indian dress).

I have here mixed the Expert and Transaction models but I am sure similar combinations will be needed in other contexts, sometimes involving 3 or even all 4 models. Graphically, this means to allow a work process to be plotted in the middle so as to overlap 2 or more models. The Process mapping diagram shown above seems to allow this (“customer service” work process overlaps the Integration and Transaction models) but it is not clear if it was really intentional (probably they worked it out themselves since then). In any case, it was a promising framework and I would love to learn of its implementation successes.

Peter-Anthony Glick