It's Databases, Stupid has a new capability called a “Purchasing Circle.”

Instead of having its database compare you blindly to other buyers (called “collaborative filtering”), the purchasing circle lets you choose whom to affiliate with and see what people you say are like you are buying.

The system’s new, but it says the people at my old alma mater are going for books on artificial intelligence, corporate finance, style manuals, The Onion, and Arthur Golden’s Memoirs of a Geisha. (That sounds about right.) Amazon has also unified its searches so that instead of looking for books, movies, or videos, you enter something in the search box, and it brings up everything they have.

The point is that Amazon knows its chief asset is its database. Amazon is working hard to tweak and add to the capability so it can learn even more about its customers and, hence, serve them better.

It’s this sophisticated databasing capability that really separates the web’s big boys from the small fry. What a small shop can keep in its head, a big shop can do with its technology, and it will win the sale through heavy advertising and economies of scale. That’s the concept, anyway.

Look across the web, and the story is the same. Yahoo is putting its user database at the service of its advertisers. AT&T wants a cable monopoly so it can ham-handedly build the Excite@Home database into something that anticipates people’s e-commerce needs.

When you look inside the concerns expressed over privacy during the Jupiter Online Advertising Forum (most of us don’t trust web sites), what you see is a fear of databases. They know us too well. The people behind them can’t be trusted, and even if they can, the government or media may get our file and use it to prove we snorted cocaine, say, 25 years ago. That last is raging paranoia. But the fear is real and anyone building a database must recognize it.

It is expensive to build, maintain, host and (most important) get use out of a customer database. Very few sites do all these things well. The New York Times, for instance, has done a good job of building its database, but it has gotten very little use from it. (Compare what they’ve done with the registration data they’ve collected with what Amazon has done with its customer database.)

The real frontier in electronic commerce today (I submit) revolves around databases. How do you get the most value from them? How do you give the most value to your customer from them? And how do you maintain the customer’s trust in what you’re doing? These questions are all related, and not all their answers can be solved with technology.

Here’s a clue to answering them. If you do your technical job right, what do you still need in order to serve someone using the data you have stored on them?

That’s right, you need trust. Most of us still don’t have it, apparently. If you want to get the most from your databases, don’t worry about changing consumers’ views of the market on that question. Change their views on you.

Related reading