I advise a very creative business that, in the last few years, has energetically embraced the promise of database marketing. It's true that its executives came rather late to the technology and, like most companies, had a few engineering stumbles along the way. But now they have a system of sorts that keeps tracks of their customers and their purchases. For a small company, it's been a big step, and they know far more about their market than they used to.

But now that they have all their data, they are so enamored of it that they're paralyzed.

One of their product lines--the one that put the company on the map--is failing; the established customers are no longer buying and new interest has not materialized. So the company needs to make some key decisions: Should it ax the product altogether? Reduce the costs it incurs? Market differently? Or invent something new?

At no time are these simple questions but now that we have a sea of data, nobody can think without it. Any proposal--from revolutionary change to minor adjustment--is greeted with discussions of whether the data would bear it out. Creative ideas are rejected because, given the data structure, they can't really be tested. Unless of course we restructured the databases in which case, in a few months or so, perhaps....

Many of the finest database marketers recognize the dilemma we face. Data will tell you a great deal and help you make endless refinements and adjustments so that you can deliver delight. But big data won't deliver a big new idea, the concatenation of zeitgeist, information, and imagination.

And unless you set up your data systems so that you can experiment with big ideas often, quickly, and cheaply, you may find you know everything but understand nothing. Big data but small ideas.