posted 17 May 2002 in Volume 5 Issue 8
5 minutes with… Mobistar
Jacquie Bran, project manager for the Knowledge Management events team, talks to Dominique Foucart, corporate programme manager with MobiStar, one of Belgium’s leading wireless operators, about the lessons the company has learnt through building and applying a taxonomy to its intranet, K Village.
When and why did you first consider implementing a taxonomy?
I am not sure we initially had a very systematic approach towards taxonomy. Our initial intranet was in itself our ‘quality manual’, and as such included a very formal taxonomy. When we decided to expand it to what is now K Village (our knowledge repository), we decided on rules that can now be perceived as a taxonomy.
What have you done to encourage and promote knowledge sharing, and what barriers have you faced?
Our approach is twofold: generate traffic (make sure people come and search for knowledge through the intranet) and generate content (make sure contributors use the intranet as their main contributing tool). The first objective is answered through classical communication-based approaches: the homepage of our intranet is the news page of the company where people can find all the professional and social news of interest. They are then placed in front of the multiple categories that have been built after analysis of their needs through both interviews and ergonomical studies. Content contribution requires the permanent animation of the ‘publishers’. This is the main role of our dedicated intranet team, which currently consists of two people: one of them focuses on getting the people together and challenging the quality of the content proposed, while the other acts as a support to publishing by developing templates and graphics.
How did you progress to implementing an infrastructure to support a taxonomy system, and what changes were necessary to ensure its success?
We haven’t purchased any sophisticated or expensive system to support our taxonomy approach. Running Microsoft Site Server, we are simply using the capabilities of the system as much as possible.
How did you arrive at a decision with regards to implementing a manual rather than an automated taxonomy system?
Because of our limited size (1,800 people in the organisation) we did not perceive a need for an automated taxonomy system. Once a year, we run a detailed needs and satisfaction survey on our customer base and use it as a bottom-up trigger, to combine with our top-down objectives to provide an actual knowledge sharing system.
How useful has your taxonomy been in retrospect?
Manual or automated, you have to monitor the satisfaction level of your customers closely. A key questions to ask is how easy it is for them to find the information they are looking for. The more we increase the amount of information available, the more critical the taxonomy becomes. Our customers keep telling us year after year that access to the right information can be improved. They also show us that they use the intranet more and more as a knowledge repository. This means that we need to continue to improve the organisation of information to facilitate its access.
How have you incorporated provision for your taxonomy to be regularly updated and maintained? If you haven’t, what plans do you have with regards to this aspect of taxonomy building?
Category management is happening every two weeks as part of the remit of the Intranet Reaction Committee, and we have a number of projects running to develop specific access methods by category of knowledge, made transparent through common interfaces. We still do not plan to develop automated systems in the near future, although our IS department is developing technical intelligence that will allow it to propose alternatives as they become affordable.
What are the main lessons you have learnt?
The main lesson we have learned is: assess your actual needs, and don’t invest heavily in expensive automated systems (and their even more expensive associated integration and consulting costs) without a clear evaluation of the payback. When we did the calculation, we discovered that we could probably live with a very decent system for a fraction of the cost, while still extracting the main benefits of a more sophisticated system.