Exploring the decision process for KM tool selection. Started on 06/22/2000 by DenhamGrey


Is there an optimal way to make this decision?

Selecting the ideal KM tool is no easy task. This is a higly dynamic environment with new vendors, alliances and applications being announced daily. KM itself is a very broad field covering a number of separate areas e.g. document management, helpdesks, business intelligence, relationship management, data mining, collaborative workspaces, distance learning....

Take time to consider exactly what knowledge means to your organization, are you doing this to level the playing field?, using this for strategic advantage?, playing catchup?, how does this link with your strategic directives?, what applications do you already have that can be enhanced by adding knowledge related practices?, should you not start with the cultural and ontological stuff first? Next compile a ranked list of the main representations you will be using, (Expert Choice http://www.expertchoice.com/ and similar decision support software can be very useful here) Then think deeply about your approach to KM, will you favor learning and collaboration or data driven pattern extraction?, do you wish to support internal communities of practice or capture customer innovations and market shifts? will you first priority be knowledge sharing or networking and relationship building?, will you encourage knwoledge emergence via dialog or capture best practices via forms and workflow for approvals?

The classic approach seems to be:

  • conduct a survey / mapping exercise. This helps to orientate the consultant and highlights constraints (integration, time, costs, training, culture) and buys some time to bootstrap the change process.
  • issue a rfi & rfp this gives some facade of due dilligence and fairness when faced with angry vendors
  • select a few (3-5?)vendors based on elimination via constraints, history, relationshipa / comfort levels or 3rd party alliances
  • require an onsite demo using company data this helps with selling the solution, demonstates integration and compatability, provides a chance for the troops to see the vendor in action
  • make the final choice bargain on price, set deadlines and sign a contract.
  • example: http://www.smthacker.co.uk/IS_IT_strategy_software_selection.htm

Few publications or cases state the rationale for application selection or go through a formal rating process. Note selection is very similar to configuration in terms of macro strategy, i.e. satisficing under hard constraints.

Things to consider

  • ASPs a new and emerging market that may greatly reduce risks and training, main benefits are: cost & cashflow, time to run, security, cheaper & faster e-training
  • Existing environment (integration with legacy data repositories, existing expertise & skill sets, avoid delays & costs due to (re)training)
  • Principles & strategy (repository vs. collaboration, reuse vs. innovation, codification vs. personalization, content vs. relationships, internal vs. external)
  • What type of knowledge will you be dealing with? (documents, patents, design rationale, event co-ordination, relationships, newsfeeds, self-publishing, stories, FAQ, problem / solutions, patterns, learning histories, best practices)
  • How and what representations will you be working with? (visual, graphic, text, synchronous asynchronous, integrated, web-based dynamic content) compliance with standards?
  • What functionality do you need essentialities vs. wish lists? (annotation, discussion, annealing, personalization, workflow, diagrams, extent of automated support)


  • Co-design, cultural change, motivation and alignment
  • Introduction sequence, prototyping,
  • Early adopters, chosing a win-win champion
  • Finding 'fit' across groups, departments, stakeholders
  • Judging vendors evolution and promises

Where to start?

  • Look at your approach and philosophy.
  • Consider your leverage position. Are you comfortable being on the cutting edge?, how resistant will your staff be?, do you make a radical or gentle change?, can you add functionality to existing platforms?
  • Look to supporting & enhancing your existing natural knowledge flows.

Decide on a knowledge approach / philosophy

  • Personalisation vs codification
  • Process vs. practice
  • Data driven vs. community
  • Internal vs. external
  • Portal vs. groupware

KM Tool Selection Gurus


An ontology is a way of seeing the world
, an ontology for KM tools needs to start from the heart of knowledge management, i.e. knowledge related practices, it needs to accommodate knowledge types and forms and take account of key knowledge objects & assets An ontology can then be the basis for a tool selection decision,

KM technology ontology links:

Starting from technology, will deliver a product of limited usefullness. For a technology driven ontology see: Alex Goodall 1999 Technologies of Knowledge Management: The technologies covered include:

Agents, Bayesian Nets, Business Rules, Case-Based Reasoning, Constraint Solving, Data Mining, Intranets, Evolutionary Computing, Fuzzy Systems Induction, Knowledge Systems, Languages, Model-Based Reasoning, Groupware Ontologies, Text Processing, Natural Language, Neural Networks, Voice Recognition, Creativity Software

DM - I found their categorisation/profiling lists to be quite interesting too ... trying to do similar things to that which we have done already in WCSN. I am planning on contacting Alex later in the week to talk about collaboration possibilities.

A common way to approach KM tools is to consider two main classes: static publications and active conversations or transactions. This also leads to a push vs. pull mentality yielding little value as ideally both are required in some form or another.

Ruggles, 1997 follows a different breakdown, i.e., generation, codification and transfer which we find to be more useful. http://www.businessinnovation.ey.com/mko/html/toolsrr.html

Key knowledge practices

  • (1)Structuring: linking, classification, indexing, patterns, best practices
  • (2)Sharing: deep dialog, distinctions
  • (3)Creating: inquiry, reflection, synthesis
  • (4)Mapping: boundary objects, flows, dynamics
  • (5)Mining: text and numbers, rules, heuristics
  • (6)Learning: individual & group
  • (7)Storing: corporate memory,
  • (8)Acquiring: searching, elicitation

DM - This grouping sounds somewhat analagous of the Davenport/De Beer article in Sloan (Winter 97 I think) where the authors gave 8 differing categories of KM projects. I later abstracted that lot into a series of slides. Hmnnn - makes me think I should review my existing material on KM and revisit the Ontology. I had not thought of that as a way of categorising the technologies themselves.

Key knowledge types

  • (A)Best practices
  • (B)Lessons learned
  • (C)Heuristics: tips, tricks rules of thumb, intuition
  • (D)FAQ
  • (E)Decision rationale
  • (F)Stakeholder capital: customers, suppliers, partners, employees, owners
  • (G)Patterns

DM - The Knowledge Types categories are interesting ... when I think about how I categorised the existing content in the site I did not look into this way of thinking about the issues/cross-referencing the material. I would not have thought that this section is so relevant for categorising the technologies themselves - more the relevant perhaps for the project descriptions of the individuals.

Key knowledge objects

  • (a)Events:
  • (b)Activities: processes
  • (c)Relationships: personal, links
  • (d)Documents: forms, templates,
  • (e)Stories: cases, exemplars
  • (f)Perceptions: problems, issues, opportunities
  • (g)Profiles: directories, yellow pages

Another look at knowledge structures and activities: http://snow.utoronto.ca/Learn2/schema.htm

DM - In many ways what you have described here are the assumed types of content that we have on the site (and some we don't have). Again, the point is does this list help me categorise the underlying technological approaches of the products or is it that you see this as a way of describing the content that they manipulate, store or assist users in dealing with.

Let's take some examples and see where they shake out: our process is to look at the key knowledge practice which the technology supports first, then to see which knowledge type is targeted and finally to pick a key knowledge object. This yields a 3 level classification.

DM - Perhaps we should talk about the issues associated with categorising these technologies. Work out which ones we should concentrate upon and produce a review or two.

From the [KnowledgeManagement] egroups listserv post 06/27/2000 by Anand Satchit

Tool-set #1 Groupware and/or Project Management Software. Lotus Notes -or MSOutlook, an Intra-net, are critical. Look for PM softwares that link back to Notes or Outlook and that are collaborative in focus.

Tool-set #2 Document Management Software. Bringing source content through the enterprise means converting source documents into electronic form to ease storage access and transmission. Look at tools such as EDMS (Documentum), Datware II Publisher (Dataware), and Panagon (JetForm?).

Tool-set #3 Data Mining/ Cluster Identification/ Retrieval Systems. There's a wealth of product here. Verity and Excalibur are among a number of retrieval tools. There are various bayesian agents and clustering systems that attempt to scrape though the enterprise to identify kindred sets of data and documents, in an attempt to find "implicit" knowledge within the organization on a given area an individual knowledge worker is focused on. I think most of the tools assume an informed user who knows what he/she is looking for, and most of the clustering systems don't really work that well in practice.

Tool-set #4 .External Intelligence Systems I think these are key, and they ought to be wrapped right back into the corporate Intranet or GroupWare?. These consist of news and information aggregators News, reports, and more. The broader the reach the better. Getting useful content can be an issue here. too much stuff can be as big a problem as too little, as it can overwhelm the users. NewsEdge? is great because they provide news and information and then peer review content to make sure its focused an appropriate plus, they add a KM consult to make sure the clients are getting the information they really need to do their jobs. Factiva is less expensive, but it requires someone in-house at the client site to own the process of pulling useful information out of the mix.

Tool-set #5 Intellectual Asset Management Systems. Software that helps track and manage the intellectual assets of an organization is fairly primitive at present, sad to say. Right now these are largely things like legal tracking systems regarding patents, and the like, usually custom and running out of some firms ERP system. Tools in this space should be expanded to systems measuring training effectiveness, and other means of valuing the intellectual assets found between the ears of the employee base. Several of the big six firms are exploring this as are some smaller guys like IXPartners out of Salem NH. NIOSH has developed an assessment model, as has various branches of the military. I'm dubious about all of them. I'd refer you to Ed Riolfi at McGraw? Hill, who has really spent a lot of time developing and assessing metrics in this area