Monday, 20 February 2017

Knowledge Management, eliminating the Archaeology from projects.

A client said to me last week, a propos of Information Architecture, that "up to 40% of Architecture is Archaeology".  Knowledge Management can help address that 40%.

Image from wikimedia commons

By the statement "up to 40% of Architecture is Archaeology" he meant that up to 40% of Information Architecture projects is spent digging around trying to find out why systems are built the way they are.

IT systems are often built up piece by piece, and revision by revision. Each revision project is fully documented and those documents can often be filed somewhere, but often nobody keeps track of the overall design and the overall rationale behind the system. That high level knowledge is lost; buried below layers on layers of revisions. Hence the need for archaeology. Each systems architect working on a new revision has to dig through the files of multiple projects to understand how everything hangs together, like an archaeologist sifting through the artifacts of a lost civilisation.  

And then, in the absence of a Knowledge Management framework, the knowledge they have compiled through this archaeology is lost again; as if the archaeologist discards her findings and buries them again.

It is not just Information Architecture that suffers from this problem. If knowledge is not stored and shared, then every project needs to do some digging to find out what has been done before, and why.

A simple knowledge management solution, perhaps a wiki, would allow the rationale behind the system to be recorded, and then updated with each iteration. All key documents could be linked to the wiki, and the wiki can evolve over time as the system evolves.

A simple KM solution such as this could save up to 40% of project time. It just requires people to work in a different way, and to share the results of their archaeology so the next person does not have to start again from the beginning.

FInd out how much archaeology your projects need to do - this could form a good value proposition for Knowledge Management.

No comments:

Blog Archive