Data, data everywhere but not a drop to drink

I spoke a while ago about the curse of email as a chaotic repository for corporate information. Although the use of email in organisations is woeful and counter-productive it pales into insignificance compared to the original curse of the networked PC – the file server!

We’ve been dumping things in our H colons and I colons for many years now and the ability to use those things effectively in the future is nigh on impossible. The lack of structure and random use of file servers is sad testament to our ability, as a race, to classify information effectively. This poor usage has also meant that most attempts to formalise this process fail miserably. I’ve always seen “document management” projects as one of the great IT graveyards where money is poured into a solution that never really works. Reason being is that their imposition usually means that users have to change what they do – which is usually nothing – and start doing things when they save documents – adding meta tags etc.

I was once involved with a project that was required to define the owners of the data on the corporate file servers. Now this sounds like a pretty innocuous task doesn’t it? It wasn’t – drilling down into those file systems is like looking back into time but with no reference to a clock or calendar! Having to trawl through the sub-directories, guess where they came from and then approach the various business managers was a woeful task. Those business managers hardly have time to keep up with what’s going on today never mind worry about what their predecessors were doing 5 years ago. Of course this lack of concern in no way means that you can delete it - "We might need it some day"!

Why does this happen? Well most organisations of a decent size (say over 500 people) are useless at maintaining a map of their organisation. OK there may be up to date org charts in HR but as for pushing this structure into the file systems there is little hope. Most have been mapping access to files in their file servers by users not groups so when an organisation changes structure their data doesn’t automatically go with them.

If you’ve ever joined a company and been asked who in the current organisation your role matches then trust me there is no structure to the file server data. The reason the admin staff copy users is because they have no idea what file locations people have access to and therefore copying somebody who already works there is the easiest option.

The sad thing is it’s really easy to map an organisation within a directory merely by updating the field in the users record which says “Reports To”. If this field is correct for everyone then an automatic structure is built within the directory. By the way you are in an company that has done this I bet that the CEO reports to his secretary! ;-)

Once we have that structure the groups and owners would be really easy to track and our data would have more meaning. The fact that most organisations haven’t been doing it means that they have a project time bomb ticking quietly in the background. That project is either the file server consolidation or worse the automated administration project – I’ll talk about that next.