Is BI BS?
Gallant: Why do you think when you talk to customers there's this lingering perception that analytics and BI haven't lived up to the promise? It's as if this stuff is always on the verge of really transforming business, but we never really quite get there. Why is that?
Chen: Number one, it's the culture of IT customers. They want things to be very easy and turnkey, but anything easy and turnkey is basically reporting. So they figure that -- why am I spending all these hundreds of thousands of dollars, if not a million dollars, for a better graphical pie chart? I mean, that's wonderful, but Excel could do that too.
So I think their view of analytics is that it does not live up to expectation to get the meaty, substantive stuff. But in order to get the meaty, substantive stuff, you have to invest heavily in the algorithm you're going to use.
Now some of the industry loves that. It relies on that. Who are they? Insurance people. If they don't have analytics, they die. I mean, that's how they decide how to price things, in what segment of the market, what demographics, how old you are, East coast, West coast. I mean, there's no point in offering jet skiing insurance to people in the middle of Alaska. I mean, they live on those models.
Most other people haven't got that sophistication. All they get is -- they buy from a BI company and say -- oh boy, let's see what I can do. Let's say I buy Informatica, tie it to a database, and then somehow it generates a bunch of data. And then it gets displayed using Hyperion, Business Objects, Cognos, whatever, and it looks great on the screen. The cost is too high for that functionality. It doesn't really tell me a whole lot. And that's the problem.
I think the BI market and the analytics market are attempting to become a quick-deployment proposition. So they made it too simple, they made it too template-driven, they made it too model-driven. Now, the truth of the matter is, people like SPSS, because customers really believe there's a secret sauce. And SAS is not bad; SAS is a great company. So the culture of the market is now maturing. This is not about a spreadsheet displaying on the screen so I can show my boss all the good stuff. This is really about: What is the intelligence on what customers are telling us? What is the market telling us?
Gallant: Is part of that answer that there may be a fair number of companies out there whose business just doesn't really require analytics like an insurance company's business requires it?
Chen: There is some level of that, but I think I could make an argument that every business, because it's competitive, requires analytics. Just know about your customer and know about your market and know more about who will buy what. And that isn't limited to insurance companies.
Next page: Telcos in the cloud