BY JIM MCGEE

I'm not in the habit of recommending books that talk about hedonic regression analysis. "The Innovator's Dilemma" is a worthy exception.

While the book is sometimes a bit dense, it tackles an important problem and comes to a surprising conclusion. The puzzle is this: Why is it that upstarts, not industry giants, bring industry-transforming innovations to market? Conventional wisdom is that giants are managed poorly, but Innovator's Dilemma states that the real problem is good management.

The author, Harvard Business School professor Clay Christensen, marshals so much data and analysis in support of his position that he makes a powerful case. Worse, he shows that it's hard for the giants to play catch-up once an upstart has reached the market with the innovation.

Mr. Christensen shows that a product doesn't happen in a vacuum. It happens at a particular time, at a particular company, with particular customers, in particular markets. Over time, all those forces shape the idea, and everyone involved winds up wedded to it. A disruptive idea is unlikely to even rear its head in this setting, and certainly won't take hold.

The problem isn't that managers at giants don't listen to customers or do what customers want. The problem is that they do satisfy customers—meaning they miss out as a disruptive new idea takes hold with a new group of customers.

Mr. Christensen makes his point with several examples of companies that developed prototypes that held no appeal for existing customers but later developed into significant new markets. Back in 1985, Seagate engineers developed prototypes of 3.5-inch disk drives. But Seagate's existing customers, which manufactured desktop personal computers, expressed no interest because the 3.5-inch drives couldn't match the storage capacity of existing 5.25-inch drives. Now, Seagate wasn't stupid. It saw the long-term potential of the laptop market and knew that laptop makers needed the smaller drives to conserve space. But any organization has limited resources, and they generally go to keep improving products for existing customers. When laptops took off, Seagate was left behind.

Management's focus on existing customers and markets would not be terribly troubling, except for the little issue of financial returns. Mr. Christensen compares firms that pioneered disruptive technologies with firms that attempted to compete within existing markets and concludes that "firms that sought growth by entering small, emerging markets logged 20 times the revenues of the firms pursing growth in larger markets."

Firms that follow a fast follower strategy fare no better. For example, Kresge revolutionized the retail industry by introducing discount selling in its separately managed Kmart division and by organizing itself to move high volumes rather than focus on customer service. When Woolworth tried to compete by launching Woolco stores, it failed. Woolworth was still Woolworth, and it couldn't operate cost-effectively enough.

Other examples abound. Major airlines have tried to copy cut-rate ones but find they can't. Encyclopedia Britannica couldn't catch up with Microsoft's Encarta CD-Rom product, partly because a company can't afford to sell $80 products door-to-door. Seemingly every manager wants to copy GE Chairman Jack Welch's strategy to be at least No. 2 in every market or to get out—but finds the strategy isn't so easy to make work.

Is major innovation impossible at existing firms? Almost, the author says. He offers a few solutions, among them that dominant companies might try to increase market demand for new features—in other words, stop trying to increase your ability to push innovations into the market; change your customers so they pull more innovations out of you.

He pretty much proves his ideas, too, if you'll trust him on the hedonic regression analysis.