What is business intelligence?

Source: Wall Street Journal

WSJ: What is business intelligence?

Mr. Davenport: It’s the systematic use of information about your business to understand, report on and predict different aspects of performance. Historically, it has had a number of names. It was called decision support when I started my career…It’s been called business intelligence for about a decade now.

WSJ: Why is it so hot now?

Mr. Davenport: I think it’s some combination of the software and hardware for doing it keep getting better and there is a huge amount of data that most organizations have now. You could argue that for 20 to 30 years organizations were just capturing and refining their data about their business and now they have finally decided maybe they can actually do something with this data to manage their business more effectively.

WSJ: What are the benefits?

Mr. Davenport: I define business intelligence into really two categories. One is just reporting, knowing what is going on in your business.

The benefits of that are that one you stay out of jail. If you are a public company you have to report accurately on what is going on in your business. Two, you get relatively early warning on how you are performing so you can fix the problem, and you can educate mangers about your business.

Analytics, which is the other key category of

business intelligence, is more understanding-oriented in terms of knowing what factors are really driving your business performance, or prediction-oriented, looking forward instead of backward.

WSJ: Can you give some examples of how specific kinds of businesses would use these tools?

Mr. Davenport: In hospitality, pricing and customer loyalty would be the two major areas. Pricing of rooms or pricing of airplane seats. Pricing of airplane seats was a very early use of this stuff.

In health care, you are trying to figure out who is likely, for example, to get a certain disease. Who is likely to get diabetes, a big issue going on in health care right now. So you can do an analysis based on how active physically the person is, how much they weigh, their blood sugar levels are today and you can actually come up with a fairly good assessment of how likely a particular individual is to come down with diabetes.

In financial services, it’s ’should I give this person a mortgage?’ ‘Should I authorize this transaction on a credit card?’ That has become a very analytical decision. ‘Should I pay this insurance claim or is it fraudulent?’

For retail, it’s ‘what price should I charge, in the beginning?’ ‘How many of each clothing type should I have in each store and when should I mark it down and by how much?’

WSJ: What technology does a company need to go down this path?

Mr. Davenport: The most important thing is good data. Now organizations still have lots and lots of it.

The key task is integrating it and making sure it doesn’t have a lot of errors, making sure it’s common, that customer means the same thing across the entire organization so you are not comparing apples to oranges.

WSJ: Is it all about technology?

Mr. Davenport: No, not at all. That was one of the things that I think was an interesting outcome of my research. I’d say the number one factor is really a leadership dimension.

It’s how committed are a company’s senior executives to fact-based and analytical decision-making and the whole idea of experimentation as a way to learn rather than doing it out of gut feel or intuition.

The easiest way is when you have a new CEO who is very quantitatively oriented and who has had experience with this. People like Jeff Bezos at Amazon.com Inc., he was the founder, but he is very mathematically inclined. Gary Loveman, is one of the best examples as the CEO of Harrah’s Entertainment Inc. who has a Ph.D. in quantitative economics from MIT. Reed Hastings at Netflix Inc. was a high-school math teacher.

Sometimes it’s a struggle. I remember some people at Harrah’s saying they did a hotel-room price-optimization system where the business-intelligence system was recommending prices that would be charged for hotel rooms when customers would call and make reservations. Initially, the hotel property managers would override the system, because they didn’t understand it and they thought their expertise and their experience was a better guide to what prices should be.

After some analysis they found out the system was doing a better job, so they had to get a little heavy handing and say if you keep overriding these things you don’t have a future at this company.

WSJ: What do you make of the SAP deal to buy Business Objects?

I think it’s a good idea. I believe that the emphasis in the world has shifted now from compiling data from transaction systems to using the data effectively. SAP had some capabilities for doing that, but certainly I think most people would agree that they were not industry-leading capabilities. Now with Business Objects I think they have a much better story about how you can use your data effectively to run your business.

I think it’s a statement that just having transaction systems is not enough and I think Oracle’s purchase of Hyperion is evidence of the same thing.

WSJ: Will business-intelligence and analytics software disappear as a separate category?

Mr. Davenport: Well that is an interesting question. I do believe that the goal of certainly SAP and Oracle to a slightly lesser degree is to start embedding business intelligence, both reporting and analytics, into the business process itself.

The world is a long way away from doing that, but I believe that is the long-term direction.