What Money Can't Buy

More than that, Microsoft may be the most striking example ever of the phenomenon that Harvard academic Clayton Christensen famously identified in his 1997 book, The Innovator's Dilemma: When New Technologies Cause Great Firms to Fail (Harvard Business School Press). Good managers, Christensen wrote, tend to direct resources toward protecting established lines of business, usually by investing in incremental improvements that help pad profit margins.

Christensen called these "sustaining innovations." We call it playing defense. It's not that Microsoft lacks creative talent or that it simply has run out of ideas. In fact, the company has an abundance of both. But for most of the last quarter century, it has overwhelmingly devoted these vast resources to the innovative defense of its existing franchises. That's why it has missed opportunity after opportunity to launch important new businesses. And why, in all likelihood, it will miss many, many more.

The Department of Defense

Imagine working at the biggest, richest technology research-and-development operation on the planet. In its fiscal year ended in June, Microsoft spent $6.8 billion on research and development. For perspective, that is roughly equal to the annual R&D budgets of Oracle, Hewlett-Packard, Dell, Apple, and Sun Microsystems combined. Only IBM, with a budget of $5 billion, comes even close.

Microsoft's talent base is staggering, too. Most of its 57,000 employees are engineers, but fully 700 of them work exclusively for Microsoft Research (MSR), the company's quasi-academic think tank that is two-and-a-half times the size of Xerox Corp.'s famed Palo Alto Research Center in its 1970s heyday. With an annual budget of $250 million, MSR can afford to hire the best and brightest: Gary Starkweather, inventor of the laser printer at XeroxPARC; Jim Gray, a seminal developer of the distributed database; and Gordon Bell, known as the "father of the minicomputer" for his work with miniprocessors at Digital Equipment Corp. in the 1960s. Twelve of MSR's researchers have been inducted into the National Academy of Engineering. Two have won the Draper Prize; three more have won the A.M. Turing Award for computer science.

In the past five years, Microsoft has acquired 2,188 patents to protect its researchers' work -- and the quality of those patents is outstanding. Patent Ratings LLC evaluates patents on three criteria: the number of unique technology claims, probable revenue and profits, and the likelihood that they can be defended. Over the past five years, Microsoft's average patent "intellectual property quotient," or IPQ, was 123, well above average. Because software patents are harder to acquire -- and consequently harder to defend -- than hardware patents, it's especially impressive that Microsoft's average IPQ is higher than those of hardware makers IBM and Hewlett-Packard -- the two most prolific patent claimants in technology.

Moreover, Patent Ratings estimates that of those 2,188 patents, 75% are original, as opposed to continuing patents. This means that three-quarters of them were issued to protect entirely new technologies, not, as with continuing patents, to renew claims on old inventions. In the past decade, Microsoft was issued proportionally more original patents than either Intel (71%) or Apple (68%).

But here's the thing: For all of its cash, talent, and intellectual property, for all of its apparent research productivity, Microsoft isn't really set up to invent new stuff. Microsoft's co-chief technology officer, Craig Mundie, volunteers a remarkable statistic: Of every dollar the company spends on R&D, "probably something on the order of 90% is directly in line, or in service of, the existing business groups." Just 10%, he says, "is essentially [invested in] pure research or incubation [of new products]." So yes, Microsoft files many original patents, but if Mundie's assessment is on target, most of those probably represent merely innovative features for existing products.

"We have this interesting economic model, which is we sell a product that never wears out," Bill Gates argues. "So if I don't innovate in Office, I get zero revenue."

There is, of course, a compelling economic rationale for this defensive strategy, as Gates himself takes pains to explain. "We have this interesting economic model, which is we sell a product that never wears out. If you [have] a version of Office, I can't come to you five years later and say, 'Okay, it's worn out and you need a new one.' I have to come to you with innovation that makes it worth it for you [to] move up to [a new version of] the software. So if I don't innovate in Office, no matter what my market share is, I get zero revenue. If I don't innovate in Windows -- you already have a copy, it is right there running on your machine -- I get no revenue."