Some things about KM that we now know are wrong

There are a few things that act as talismans for traditional knowledge management. Here’s a couple of blog posts undermining commonly-held KM superstitions.

Superstition 1: We need an expertise directory

This sounds like a great idea. Clearly “know-who” is an essential part of good knowledge management. Without it, how can we justify David Weinberger’s claim that “A knowledge worker is someone whose job entails having really interesting conversations at work.” So what should we do? The obvious answer: get everyone to add their details to an expertise directory.

My instinct is that this approach is doomed to failure. In order for an expertise directory of this kind to work, a couple of things need to converge. First, we need to be able to identify what information might be useful to people in the future. This obligation might fall on the system designer — to build a taxonomy that encompasses all possible future eventualities. Or it might rest with each individual — to describe in free text what they do in a way that includes all the topics that might be relevant. That’s a challenge. The other thing is that the right people (as many people as possible) need to contribute.

My experience, and the reported experience of IBM (over a much larger, and therefore more authoritative, sample) is that this approach fails because neither of these factors is realistically achievable.

After almost 10 years of from-the-executives, repetitive, consistent pressure, only 60% of all IBM profiles are kept updated.(Note that Lotus Connections Profiles is the productized version of IBM BluePages, which has been around since 1998.) And that’s even with an automated email sent out every 3 months to remind people to update their profiles, plus a visual progress bar indicating how complete or incomplete a user’s profile is, plus people’s first-line managers constantly reminding them to update their profile.

So what should go in its place?

Once we gave Contributors the choice about how to share their knowledge and experience, we found that they were more likely to contribute using these social options, since they realized that the result would be fewer emails, IMs and phone calls asking for their basic expertise.

“Read my blog.”… “Check out my bookmarks.”… “Look at my activity templates.”… “Read my community forum.”

…became the new ‘RTFM‘, if you will.

Now, once Seekers find an expert via Profiles, they are able to consume some of their knowledge and expertise without disrupting them. The nature of the remaining email/IM/phone requests from Seekers were about their deeper experience, their knowledge that will always remain tacit.

In practice social bookmarking, internal blogging, communities and activity tracking (all “in the flow”) beats voluntary confession of expertise (“above the flow”). The tools? For (and by) IBM: Dogear for social bookmarking and Connections for blogging, communities and activities. Surely law firms (even those without social networking tools) should have a head start in this area. There is huge scope for leveraging the information about people’s work in existing databases: document management systems, billing and time-recording databases, CRM systems. If we get our systems to talk to each other, we can enable real human conversations.

(For those who prefer a visual approach, there is a video.)

Superstition 2: KM efforts need incentives

I think I have said before that I am not a fan of knowledge repositories and the Field of Dreams triumph of hope over experience. Received wisdom says that in order for such know-how systems to work well, people need to be encouraged to use them. Neil Richards was sceptical, and asked for people’s experiences. An unscientific approach, to be sure, but the anecdotal evidence is unequivocal. Incentives don’t work. Some quotes:

While an initial advocate of incentive programs for fee earner participation in KM programs, over time I found it tended to be the same fee earners participating each time and, in most cases, these fee earners informed me they would have participated in the program regardless of whether or not there had been an incentive program.

They decided to offer a bottle of wine to the person who made the most contributions. At the next annual meeting of the group, one of the team members indeed received a bottle for having made four or five contributions over the year. (The firm’s target was four a year.) And that was the end of the program! Never revived or spoken of again. The contribution rate, which was always fairly low, didn’t change, either during or after the contest.


We have tried incentives for KM participation, and I don’t want to go there again.  Our worst mistakes were done when we deployed our global Knowledge Management program for Customer support back in 2000.  One country unit decided to give away a Swiss army knife to every engineer that wrote 10 knowledge objects. This was one of our larger Country units, so we got >1000 knowledge objects written (and very armed and dangerous engineers…). Why did this fail: There was no incentive on writing anything useful, or to adhere to any of the internal format guidelines. These poor knowledge objects polluted the search for ALL country units for years.

I am looking forward to Neil’s promised further thoughts on incentives, because I think one of the real challenges for knowledge management is to embed good knowledge-related behaviours in the organisation.

(A footnote to the expertise directory issue: a comment on the blog post refers to people’s use of profiles on Myspace and Facebook. I have entries in Facebook and LinkedIn, amongst others, and I find it hard to keep them up to date. However, I also catalogue my library on Librarything, and iTunes synchronises my listening habits to These information flows combine in Facebook to give people a picture of my interests without me having to lift a finger.)

7 thoughts on “Some things about KM that we now know are wrong”

  1. Fully utilising the collective expertise of staff in a way that benefits the whole business is certainly challenging, but it is well worth the effort, as the gains to be had from making specialist knowledge widely accessible are simply too great for companies to overlook.

    People are too busy to regularly update their profile, and the employees whose expertise is perhaps most valuable (primarily the Partners) are invariably the ones with least time available to commit to such tasks, not to mention having the lowest incentive to do so.

    Technology can help businesses automate this process and keep the ‘expertise directory’ up to date. ‘Expertise Location’ solutions are readily available, but few place the onus on the technology itself to automatically categorise and relevancy rank the firm’s experts in relation to a specific query. Intelligent expertise location solutions determine this based not only on the information held in HR/personnel systems, but also on the work they have done and the matters they are involved in – information housed in document management, time and billing and CRM systems. This approach frees up valuable staff time and enables businesses to collate and present all knowledge – no matter where it lives – in a community-based, easily accessible fashion.

  2. The main problem we have is expert locators fade away when the data is not current, and the other aspect is they are usually boring to use, that is they are static.

    That’s why it’s important for an expert locator to allow you to interact, ie. it enables you to find people and add them as a contact (connection/friend). This way when we visit profiles we can see who these people are connected to, and for personal reasons each person can a mass a list of their contacts.

    In this respect it becomes a tool that engages an individual, it becomes your contact list, you can see other people’s contact lists, etc…And as you say, the profile points to their content like blogs, and bookmarks, where in the end you may not need to contact that person as one of their blog posts answered your query.

    Now that people have some self-interest to use the tool, they are more prone to keep it updated.

    You could also do searches like: show me a list of all the experts in “Excel” from within my network, or from within my network and my friends networks.

    What is happening here is that Expert Locators (or even the IM buddy list idea) are becoming Social Networks, so rather than a centralised seek and find database that you use on ocassion, we have a more engaging and personally useful daily tool, where the people create the information and connections from a bottom-up approach (more self-organising), where it does most of the work itself as a collective.

  3. Oops, more to add…I’m cutting and pasting from an Internal blog post.

    I think my main point was that an expert locater is more like the GAL in Outlook, but a social network is this plus interaction eg. sending messages (similar to email)

    That is you use an expert locater occasionally, but you use a social network every day as it’s a read/write tool.

    When you are after some help on an issue, or seeking information or a person, you can get around your vertical hierarchy, by blasting a message to your connections. Currently we save time and money from using a “back channel” in getting things done, and networks augment this in an online scenario.
    Rather than waiting on your boss, to ask their boss in another office, who has to ask someone else, you can immediately be aware and connect to this person that is usually 3 degree’s or separation apart from you. I’d add to this we are reducing opportunity cost by more accurately hooking up with the best available person for the job at that time, and as a result we create new contacts and relationships.

    And then there is the strength of weak ties

  4. Nice post, I agree with most of it. The part of expertise locator’s is understandable from the perspective of traditional Yellow Pages, having employees fill in their profile and keep that uptodate. Which, as you say, not very many employees do. But what is you set up an expertise location system based on peering. We did this in the company I work for. You can find more info here: This tool puts social networks and expertise together. And it works: you can easily map the expertise of a company, keep that up to date, etc.

Comments are closed.