dchase on BI

Intelligence and Technology

You can lead a horse to water...

OK the heading's hokey, and I know I have to change it. Still, what I'm going to say here is very, very important. There's some amazing technology out there that does mind-boggling things with data. There's even stuff out there that puts two and two together and creates information (um... four). There are also some incredibly bright people out there building and implementing this stuff too. Still, I've been doing this business intelligence/decision support thing for a few years now and I see this this same thing happening again and again. A lot of people seem to expect technology to replace, rather than augment the abilities of people. A lot of people seem to think that the technology can make us intelligent.

...but you can't make it drink.

Well, technology can't make us more intelligent. We can use technology to help us improve our intelligence, but we're still the driving force in the equation. Technology can put data together to create information. For example, technology can process data (say 2) with an algorithm (maybe x + x) to create something new (like 4). But technology can't take this information and provide meaning. Technology can't know what 4 means. It can only help us know what 4 means. We still have to do a little bit of the work here.

Just because you have the power doesn't mean you should use it.

You don't need a Unimog to do your grocery shopping. There's a series of advertisements on television right now (winter 2001) that show people going overboard with their equipment. My favorite shows a cat stuck in a tree with the national guard, police, fire department and who all else running around with trucks helicopters and other heavy equipment in an all out effort to save kitty. Finally, a man in the crowd witnessing these heroic efforts calls out to the rescuers "Use a ladder." Many business intelligence efforts resemble these ads. But, no one is ever willing to say "use a ladder." Seems to me that the reason for this lay in our desire to have technology be intelligent for us and the fact that we have technology that comes close. If we throw enough technology at the problem, we will solve the problem.

But you need to use the right tools for the right job.

Hey, if you need the powerful stuff, use it. Don't try to tow a semi with a Chevy. If you deal with many millions of customers and operate on razor-thin margins, a one percent shift in market share or supplier cost for a week could mean the difference between profit and loss for the entire quarter. If you have a system that could show current and projected trends based on real-time data, you may be able to react to customer or supplier shifts as they happen. But you ain't gonna get this system from having a dba tinker MS SQL Server.

Make technology easy enough to use

Recently, when researching the business requirements for a company's customer data mart, I asked one of the company's research analysts how comfortable they were with using different types of software. Like many analysts she lived in Excel and Word, knew the basics of Access, and used Notes primarily for e-mail. While talking with her I pictured her using a number of possible analytic front ends that are out there for querying the data mart and playing with the data. I also asked her questions about her work routines, how she obtained data, what bottlenecks she saw and so on. She made it clear that it was difficult to get the data she needed, it took too long, and she couldn't always get data in the format she needed. I'm thinking now, and getting excited. Boy, she's gonna love this data mart, with information organized in the business terms she understood rather than information management terms that caused her eyes to glaze over. And the point-and-click pivot-drill-slice-and-dice interface... wow! I described this vision to her and asked her what she thought. And she answered:

Sounds good, and I'll use it if it's easier than picking up the phone and asking Gloria for the data in a spreadsheet.

OK...@$&! It wouldn't be easier than picking up the phone (or sending an e-mail or fax). No way, no how. It would be faster. It would be more direct and "hands-on" (allowing for iterations of questions at the "speed of thought"). But, it wouldn't be easier. How could any technology we build make the analyst's task easier than using the spoken or written word to communicate with another person?

That's the problem with technology

Technology doesn't improve things - people do. More to the point technology doesn't change what people do, people change what people do. Technology merely offers people ways to do things differently. A recent Wall Street Journal article illustrates this w/r/t B2B "exchanges":

It turns out that that the main assumption behind the exchanges was wrong. The technological workhorses of Mr. Hunt's world -- those fax machines and minicomputers -- are efficient enough that he has little to gain from the Internet. These systems already are integrated into the operations of both the produce company and its clients, and can pretty much match the Internet on essential services like communications and billing. They also offer a crucial flexibility that the Internet can't match, allowing Mr. Hunt to react quickly to market shifts and vary his prices while keeping the variations confidential and preserving goodwill.

On the street, the success and failure of new technology most often rides on the answer to the question: “will using the technology be easier than what's used now?” Using word processing software is easier than a typewriter because you can use the backspace key to change things. Using Microsoft Word is easier for most people than using emacs because it's already on their computer.

Business technology isn't about technology. It's about people.

Skip Weblog Excerpt

From the weblog:

Valid XHTML 1.0!