This is another response to a post about the “shared drive problem.” Shiv Singh of Avenue A- Razorfish commented that “Every document in an organization is not necessarily important enough to tag. Some organizations address this problem by first determining what knowledge/information/data is worth capturing for retrieval and then putting KM mechanisms in place to capture, codify and distribute it.”
My thought is that there is a continuum of value of documents. On one end of the spectrum, news feeds, unmoderated discussion, etc. Chaotic but useful in terms of creativity and problem solving – ongoing discussions like this one. At the other end of the spectrum might be best practices, templates, methodologies – structured, scrubbed, edited and tagged. Higher value knowledge is more costly to vet, tag, file and maintain. A vast majority of documents fall somewhere in between. Many (perhaps most) are intermediary products. Since the value is context dependant (as others have mentioned) and may have value as a need arises, it’s very difficult to organize them without some judgment about current and future value. I’ve seen environments where documents were nominated to be example deliverables – someone thought the document would be useful to others. There was a process in place to measure submissions and people were somewhat incentivized to do so.
For those documents that were simply intermediary work product, they were organized with the specific project (this was a consulting firm) so people could browse through the documents if they came across a similar engagement. At some point, someone still needs to make a judgment about the documents and clean them up or reorganize them. People are notoriously bad at filing things, applying metadata or even naming documents in a meaningful way. So someone was responsible for reviewing and organizing documents on the file share. Overall records policies should dictate retention schedules beyond that. But people are loathe to delete documents so there is a huge amorphous mass of relatively useless content. So there is no free lunch. High value information requires an investment of time, energy, money.
Many organizations are looking at this challenge of “retrospective indexing” – tagging documents after the fact. There are tools that are getting better at entity extraction, clustering and so on, but tagging requires judgment and context so these solutions have drawbacks. As Shiv says “tagging never works if it is an afterthought”…
Filed under: Indexing |