Wednesday, March 30, 2011

Business Intelligence & Analytics

Today I attended the Business Intelligence & Analytics Forum in the Hong Kong Convention & Exhibition Centre. That's already quite a lot of ampersands. I'll try to use no more.

Business Intelligence &1 Analytics is a bit of a mouthful. For those not conversant with the jargon of the IT industry, this is the sector that handles things like reporting how much money the bank has made, or the return rate at Marks &2 Spencers. Or (apocryphally) determining that most Friday night purchases of disposable nappies are accompanied by a purchase of a six-pack of lager, and therefore if a supermarket wants to increase the revenue from young fathers dispatched to buy nappies, it should put them as far as possible from the beer so the father will have to walk further and buy something else on the way. That is, BIA combines prosaic things you'd never imagine to be difficult with mindboggling surprises about human behaviour made visible through the application of industrial volumes of data and statistics.

Or to put it another way, a potent mix of the boring and the potentially evil.

This was the first such forum in Hong Kong, and it was a mix of IT vendors (IBM, Microsoft, SAP) and users from the real world. The man from IBM wheeled out the death by powerpoint presentation that every IBM employee always uses, the man from Microsoft showed something beautiful, engaging and fun involving spreadsheets (and no, I'm not being facetious) and the man from SAP made us watch a video of somebody using an iPad. Draw your own conclusions about the corporate culture of each company.

More interesting were the people who actually had to implement BIA. While a vendor can show you something stunning and wonderful, there's always the nagging thought that it only looks so good because they carefully curated the data and spent months tuning the tool they built, whereas when you try and do it for real you'll have something clunky, slow and ugly. Whereas the people who have to build something for a bank or a firm making trousers or an airline have to actually do something. Thus when we heard about their struggles to get tools to talk to one another, or data not being 100% reliable, or things being expensive and unusable, you trust them a bit more. It's also very interesting to see the similar thought processes people have been through, the same dead ends explored and the similarity in end goals.

What was a little depressing was how little some things have changed. Yes, in 2000 there were no Facebook conversations to analyse, but every mention of a new and amazing way to make a database run quickly, or a better interface to query it, or a way to gather data in just reminded me that I'd heard exactly the same things when I was working in London more than ten years ago. Some of it looks a bit prettier, but not ten years' worth of prettiness. I wasn't sure whether to feel proud that at my first company we'd made something that was so state-of-the-art, or jaded that nothing really new had come along since. Sure, we have bigger boxes to hold more data in now, but it didn't look like anything qualitatively different to before. Perhaps it isn't anything to be depressed about. Maybe I should be happy that we were ahead of the curve.

Maybe I was living in the future for so long that when it caught up with us, I could only be disappointed. But still, it made me feel something that's either regret that this takes so long, or anger that we haven't achieved more, and that perhaps all this time I've been spinning my wheels in the wrong company when I could have achieved more somewhere else. Is it the environment? Are the problems themselves intractable? Am I just not up to snuff? At least it ignited some fire in me, to remind me this is something I do have some passion for, in a way that just producing a set of data for somebody else to mull over does not.

The final presentations were great: one from a company veteran who'd had a BIA installation that had manifestly improved the whole way the company worked, and the other a more visionary one on what we should aspire to deliver in the next decade, actually using data and analytics to provide novelty, excitement and value to people's lives, but looking around the room at the end I was worried. Out of several hundred people, a handful weren't from IT departments. Thus everything was preaching to the converted. BIA is not something for the geeks in IT to build; it's something that businesses need to own, a tool to achieve greatness. If they weren't there to see it, how were they going to be persuaded it was worth pursuing?

1 Oops. One extra ampersand already.
2 Sorry. It really is harder than it looks to keep these damn things out.


Post a Comment