Why bother with filing?

Simon dropped something of a megaton bomb on the Tower Software blog the other day. For those of you who can't be bothered with the link, the (heavily paraphrased) summary goes like this:

"If I mis-file something, it takes me all day to find it. It also takes me all day to organize my filing cabinet. So, in a world where Google can give me stuff no matter where I file it, Why should I bother filing anything at all?"

ECM Systems, like TRIM Context, are often heavily based on information science, and encourage a kind of "virtual filing".

It reminds me a bit of my own post in 2005, where I was musing on exactly the same thing, perhaps unsurprisingly, after having to search through my personal piles of mess, just like Simon...

A few years have passed, and I've cultivated some partial answers...

• Retrieval - Filing things properly and retrieving them is still the single most efficient way to manage data. Finding a document that's been filed properly will always be faster. It sounds to me like Simon organizes things much better than I do- if he'd been able to remember where he filed his passport, he would have been in and out of his study in no time at all. Sure, an unstructured mega-sort and index is getting more practical as hardware and processing power increases, but it will never be as efficient as finding something precisely where it's expected. High speed hunt and pick is always going to cost more, and take longer. There's a reason Google only completes a full internet crawl every 3 months or so...

• Business Topology - a structured repository is worth a lot more in terms of business value, because it can teach you about your business, how it really works (as opposed to how you think it works) and can also give you clues as to what you need to do to in order to optimize it. Being able to see exactly where people are doing their work, and tracking and reporting on it is they key to operational efficiency. Nowadays, your information IS your organization. Is your organization a big vault filled with random stuff?
  

• Context - Just like I was saying about Pandora - sometimes it's not what you're looking for, but what's filed next to it that counts. If I can find an email about a given project, all by itself it's not very valuable. Within a folder containing everything else that ever occured during the life of that project, it's business context becomes apparent. Now, instead of looking for an email with the word 'Strategic Plan' in it, I have every single business document that contributed to the Strategic Plan being formed. That's powerful stuff.
  

• Control - If nothing was filed, every single piece of controlled information would need to be manually assigned at the document level. Having a structured and laid out filing plan allows you to apportion security and business logic in a much less granular, more sensible way - it's part of what allows you to scale. Applying security policy to a non-definitive search: 'show me all the things that contain the word missile' is hardly responsible. You might miss stuff. And what's worse, you'll never really know for sure...

So, there are more reasons to file in a modern Google World than you might think. Unstructured Search is all about making sense of chaos - systems like TRIM are more about organizing data when it comes in, and keeping the chaos to a minimum to encourage optimal knowledge retrieval.

Although it pains me somewhat to say it, I think that responsible, truly effective information management is really more about files and folders and policy than it is about giant supercomputer AI robots indexing a massive vault of random assorted business confetti looking for patterns.

(That doesn't mean that giant supercomputer AI robots aren't cool though. They are. )