Identifying the Genuine Expert: The Three Stages of Expertise Surfacing
There are three stages in the sharing of expertise information through collaboration software: declared expertise, deduced expertise and discerned expertise. I believe it is time for organizations to demand vendors to deliver the higher stages. Let’s review each in turn.
Declared Expertise
In the first and most elementary stage, expertise is declared. An individual is asked to state what she believes she is an expert in. This is done via an electronic form—and is often part of her profile in the system—of the business areas in which she has built up expertise. The idea is sound because if others can discover where your expertise lies, they can ask you to be involved or to contribute your expertise within related business projects. There are, however, at least three major problems that lead to the perpetual failure of declared-expertise systems. First, people hold inaccurate views on where their expertise lies. Second, individuals can’t describe their expertise in a way that other people need it described, when they are searching for it. Third, people do not keep their expertise lists up-to-date. Ultimately, the system has inaccurate, unfindable and out-of-date information. It is no wonder these systems have failed. What’s the latest example of a declared expertise system? The My Profile portion of Microsoft SharePoint My Site.
Deduced Expertise
A better way is needed of expressing expertise in a social group, and the deduction of expertise by software is that better way. The software-based deduction of expertise requires a system that trawls through an individual’s email messages, authored document collection, blog posts, instant messaging transcripts, discussion list contributions—and even read Web pages—to find the common themes and express those as areas of expertise. In stage two of expertise surfacing, we take the view that expertise is enacted, and that if you want to understand what someone is really an expert in, you have to look at how he acts, what he does, and how he contributes to knowledge and knowing within a certain setting.
Deduced expertise gets over the three problems of declared expertise, assuming the system is good at what it does. The system can create accurate profiles of the expertise that people have, it can describe expertise in a variety of ways, so as to more closely align with what people are searching for, and it keeps each person’s expertise list up-to-date. As individuals change what they write about over time, the system can deduce that a change is taking place, and can shift their statements of expertise in lock-step with what’s happening in line with people’s work. One company that has been actively working in the area of discerned expertise is Tacit Software, with its ActiveNet software for enterprise customers.
Discerned Expertise
The third and final stage of expertise surfacing is discerned expertise—where others discern an individual’s expertise in certain areas. Another way of saying it is that your expertise is socially recognized, within the boundaries of a specific social group. There are two sides of this. On the one side, you have the social recognition of expertise in areas that people have declared themselves to be expert, or in areas where the software has deduced someone to have expertise in. This is, therefore, an additional level of validation of their expertise, where people rank or assess expertise in other people, and where those rankings are shared in summary form with everyone in the organization. In products like GroupSwim, expertise must be gifted to others by the people who appreciate and value their work. Within GroupSwim, expertise cannot be declared—it has to be socially recognized.
On the second side of discerned expertise, is the possibility of people discerning examples of expertise that neither the other person can declare, nor the software can deduce. In such a case, people call out the areas of expertise in others based upon what they see as being their main contributions, whether or not the people in question can elaborate those concepts.
So What?
While some of us find it interesting to pontificate about such things at a high level, there are some specific and concrete actions that organizations should be looking at. Be very wary of systems that rely on people to declare their expertise. These systems are fraught with sociological problems, and in recent history, most of these types of systems have failed. Also, start looking into systems that provide capabilities to automatically reason and deduce expertise, and then publish their findings explicitly or implicitly. The alignment between deduced expertise and real expertise should be much higher than between declared and real. And finally, your expertise system will only work in your social setting if the people in your social setting can rank the people that they work with. Any system that you look at installing going forward needs to bring in the social aspect of expertise.
For Your Reference
GroupSwim: www.groupswim.com
Microsoft SharePoint My Site: www.microsoft.com/sharepoint/default.mspx
Tacit Software, Inc.: www.tacit.com/products/activenet/technology.html
Messaging News writer Michael Sampson helps organizations improve the capability of teams that can’t be together, to work together. He writes at: www.michaelsampson.net

Widgets & RSS Feeds