Business Intelligence vs. Predictive Analytics

As a person with an ERP background, and deeply held respect for data, I’m not usually one to rag on Business Intelligence. There is, undeniably, a lot of value in knowing where you’ve been.  But there’s an evolution in thinking, and I think it’s very exciting, and that’s in the realm of predictive analytics. What I’ve seen in my six years of implementations is a lot of ‘stuff’ (data) that is basically gathering cobwebs in the database tables. Unused.  A lot of companies archive this information without even looking at it.  And analysis of all the data might yield interesting results, some facts and figures, but most people don’t have the skills to make real predictions using data, no matter how much of it there is.

Consider, for a moment, the stock market.  I can take any stock on yahoo, and I can get daily, historical returns. I can get the PE multiple, EBITDA, the SEC 10k, and I can learn everything about the management team, including their kids’ names. But does this wealth of information get me anywhere? Not really. Why not? Well, because I’m not a financial analyst. Even with financial analysis experience (which I do have, incidentally), I’m not necessarily able to make conclusions about future performance with certainty. Even if I took all the information available about the market, the competition, and rolled it all around in my brain, the chances I’d come up with an accurate prediction are practically 0.

In the same way, companies aggregate their own floods of information, but the person doing all the number crunching usually 1)doesn’t understand all the information that they are looking at and 2)isn’t trained in predictive analysis.  I think a wealth of tools exists that have enabled organizations to look at the crushing load of data and derive insights from it. It can be described sufficiently in business intelligence to give it meaning to the user. But, that still doesn’t take us all the way to what we need, which is predictive capability.

What would be really incredible would be real-time predictions, with abilities to change certain key parameters. I want to know, with 95% confidence, with an error rate of 3%, what are the chances we will make our numbers for next quarter given our standing as of right this moment. And no, I dont want to schedule a meeting with Bob from Accounting and John from Strategy to tell me this, I want an application to do it.  Sounds awesome, right?  And that’s a financial calculation, mostly. But what about even better predictive analytics, that looks at several different data sets, and joins them all together so that it can tell me if I discount one product, would we beat our financials, and how much can I discount it, and what are other possible solutions?

In short, applications that advise rather than show and tell.

(Sidenote: Hey Mike. I think this should be our next product. Whaddya think?)

There are a lot of players in this space (SAS, Teradata, MuSigma, not to mention the BI integrators, etc.), but there’s plenty of room for more. It’s a universe of data. And I think that the modeling tools , and their hideous interfaces, could definitely use an infusion of creativity. And perhaps also some taste. Whatever, you know what I mean.  There’s a conference in DC in October (details) and I’d love to go but it’s expensive.  What’d be better is if any start ups in the DC area in this field would email me and we’d have a get together. Hint Hint.

  1. Leave a comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: