At the core of the law firm knowledge management (KM) challenge is the problem of unstructured data. Since KM is storing useful information such that it can be easily retrieved when needed, unstructured data presents significant challenges. To clarify, ‘unstructured’ means you don’t know what it means. For instance you don’t know if ‘brown’ is a name or a color.

So why does this matter?

Law firms have never worried much about structuring their information well, so it isn’t. This means it’s difficult to search and retrieve law firm knowledge. Thus the KM challenge. But even more challenging is getting lawyers to change what they do so that data is captured in structured form. As previously noted on this blog, lawyers resist change, especially change that competes with the billable hour.

Fortunately on the horizon are semantic search engines. These tools bring structure to unstructured data – which makes them magic. These tools can apply meaning to words based on their context and usage. But these tools only go so far.

Looking to the future – what makes sense is a balancing act between technology and humans. We need to make some efforts to get lawyers to change their processes. And at the same time we should apply new technology effectively.

I think finding this balance will be the key to successful KM for law firms.

  • Anonymous

    After you have finished trying to drag your mule up the hill you’ll eventually give up the utopian dream of attorneys doing anything other than billing hours.

    People are lazy by nature and will perform actions only when they can perceive clearly positive benefits or negative consequences.

    This epic battle of “search vs. sort” was waged between Google and Yahoo early on in the history of the internets. Google indexed everyone’s otherwise messy junk pages and used keywords and algorhythms to make the user experience stupidly-easy. Yahoo did a great job at classifying pages, but couldn’t get everyone to do it, and those who did classified it wrong, and there was too much volume anyway.

    You can structure anything that you can get control of and funnel through a single source. Lexis and West attempt to do this with information that is limited in quantity. For all other types of information, they use searches and algos.