Over at the Infovark site Dean has an interesting post on the semantic web including and analysis on why he thinks there are some faulty concepts particularly around the idea of having a search-bot search for you.
Which got me thinking about content management systems and searching.(1)
There are two basic concepts behind every content management system.
- Putting stuff away
- Finding stuff
In the early days it was finding stuff that was hard, searches weren’t very accurate, and finding things that had just been tossed in the “bit bucket” was almost impossible.
If you think about the Web for a minute, and how Google’s ability to find relevant information completely changed the world you’ll get a bit of an idea.
Because the Web had always been a bit of a free for all - the only way to find things was to make better search engines, because, with the best will in the world no-one was going to voluntarily fill in pages of meta-data if they didn’t have to.
In enterprises it was different. You could *make* people fill in meta-data, so you didn’t have to build such sophisticated searches. Make everyone fill in a “subject” field and you don’t have to try an infer the subject from the document.
The problem is - almost universally people hate filing, and filling in meta-data is filing by another name. On the flipside, nowhere near as many people (particularly those of us who have grown up in the web world) hate searching.
So why make people file - when they don’t mind searching to find stuff?
For some reason amongst the enterprise content management crowd there seems to be a belief that people don’t like looking for things. When I worked in the field we used to toss around spurious statistics like “Workers spend 80%of their time looking for information and only 20% of their time using it” to justify why you’d want to implement a filing system that would reduce the search overload.
Except that people hate filing, so they won’t do it properly, and they’ll still have to spend 80% of their time searching, only now - because their search assumes things aren’t mis-filed they find it *harder* to find the stuff they need.(2)
Which made me think - why is searching the internet for information “fun” and searching the “Corporate Document Repository” for information “not fun” even when you’re looking for exactly the same information.
I think it’s serendipty. I mean, I’m pretty good at using “Dominant Internet Search Engine”TM I can construct a query that will return one or maybe two results that perfectly give me the information I was looking for.
But I hardly ever do that - because it’s all the other odd things I discover while doing the search, the things I learn, the answers to questions I didn’t know I had, that making searching the internet fun…
And you almost never get that from the “Corporate Document Repository”
Well - not without running the risk of coming across information that might get you into serious trouble
Which is a whole different kind of fun - and a post for another day.
(1) Disclaimer: I used to work for an Electronic Document and Records Management Software developer who tried to change themselves into an Enterprise Content Management Development.
(2) Mis-filing is caused by laziness - Try going to ebay and see how many hits you get for palystation or Plam. In this case people have a really good reason to try and get it right - if buyers can’t find your stuff, they can’t bid on it and you may not sell it, or get the money you want. But people still make mistakes..
