Microsoft R&D

Fast Company asks the question: "What, exactly, is Microsoft getting for its $6 billion/year in R&D spending?"

They're hardly the first to ask this question, and I've asked similar questions in the past. The paradox is that Microsoft spends more on R&D than anyone else in the industry--heck, practically more than the entire rest of the software industry combined [note: This rhetorical flourish might even be true depending on how you allocate hardware vs. software R&D at places like Apple and IBM] and more than the GDP of some African nations--but seems to have very little to show for it in terms of breakthrough new products.

Fast Company's take is that: 1) Microsoft is too big to develop breakthrough new products, because Microsoft would have to develop a billion-dollar new business every year for it to make a difference, and 2) Microsoft spends much of its R&D money defensively because, well, big established incumbents are naturally defensive. Along the way, FC documents some of the apparent excesses of Microsoft's R&D, like spending $100 million/year on a speech recognition server (an investment which is unlikely to ever pay off) and hiring celebrity researchers who pursue personal projects with little or no business value.

All this is true, and it is made worse by a disconnect between what Microsoft is, and what Microsoft thinks it is.

The reality is that Microsoft is a mature business selling to a saturated marketplace. Nearly every personal computer sold today has both Windows and Office installed, so there's very little room for Microsoft to grow its market share. In fact, the opposite is more likely to happen, if for no other reason than you can't have more than 100% penetration.

Furthermore, in those two products (which generate the vast majority of Microsoft's profits), there simply are not many features left that customers are willing to buy. The most important new feature, keeping out malware, is one which customers view as making the product work the way it was supposed to work in the first place.

So, in reality, in Microsoft's core products, it can't grow any faster than the overall market for PCs, and is stuck making incremental improvements which customers don't think they should have to pay for.

In contrast, Microsoft seems to think it is an innovative young growth company with vast untapped markets still available. So it plows R&D money into all kinds of offbeat projects (Dick Tracy wristwatches? Circular computer monitors?) because that fits with its self-image as an innovator. But because in practice it is impossible for Microsoft to generate any growth with new businesses even if they succeed, there's no pressure to find and fund real new businesses.

I still think, as I've been saying for a while now, that the peak of Microsoft's power and dominance is behind it. Microsoft's peak probably came around the time it introduced Windows XP (and the stock market seems to agree). From this point on, Microsoft is fated to be a mature, and eventually declining, business. The Ma Bell of the computer industry. So how can Microsoft remake itself as a growth company again?

Ironically, splitting Microsoft into three companies (as the original antitrust settlement would have required, which the company fought) is one of the few ways I can see to rejuvenate at least part of the company. Without the safety net of twin monopolies, each company would be forced to compete and innovate to survive--and it is much easier for three $10 billion companies to each grow than a single $30 billion company.

Microsoft has had a very difficult time breaking into industries like mobile phones and cable TV where it has to partner with existing players. Part of this is Microsoft's own doing, since the company's history of aggressive practices have made other companies wary of it.

The other route is the Cisco way, where Microsoft acknowledges that it isn't as effective at developing new products as startups are, so it makes an effort to systematically acquire promising young technology companies. Efforts here have been mostly disappointing (certainly not on par with Cisco's success), and I blame a combination of Not-Invented-Here syndrome, wariness (on the part of other software companies) of being assimilated into The Borg, and halfhearted interest on Microsoft's part in growth through acquisition.

If I were a strategic consultant (which I'm not) hired by Microsoft to come up with a plan to turn the company back into a growth company (which I haven't been), my advice would be the following:

Step 1) Recognize that Windows and Office are mature products and treat them accordingly. R&D effort should be focused on bug fixes and minor enhancements, rather than trying to find the major new features which will convince customers to upgrade. The odds are that most customers will only upgrade when buying a new PC, so R&D aimed at generating an upgrade cycle is largely wasted.

Longer term, I would recommend moving Windows and Office to an open-source platform (with a proprietary layer on top), since it frees up R&D money currently spent on maintaining the core infrastructure (which customers don't care about).

Manage Windows and Office for near- to mid-term cash flow, rather than growth. Growth in these products is limited by PC sales no matter what Microsoft does.

Step 2) Spin off promising young businesses (for example, MSN) as separate companies. This will give employees of those businesses more stake in the outcome, and remove the need for them to become billion-dollar companies overnight in order to have an effect on Microsoft's financial results.

Owning Microsoft stock would become a gateway for investors to develop a portfolio of hot young technology stocks (assuming the spin-off takes the form of a stock distribution to Microsoft shareholders), which benefits Microsoft shareholders more than just sitting on a giant horde of cash.

Step 3) Seek out and acquire very young companies which seem to have promising market positions.

Then, use Microsoft's R&D money to develop those businesses, and spin them out again when they're ready to be public companies.

This three-step strategy would turn Microsoft into something like a cross between a massive technology incubator and a venture fund, financed by the cash cows which are Windows and Office.

This would be a very different company than the Microsoft we know today, but it would unlock the strategic value of Microsoft's huge technology expertise and market dominance, while allowing it to generate returns for investors without having to create billion-dollar markets every year.

The problem Microsoft faces is that it is simply too large to be a growth company any more (and it wastes huge amounts of money trying). Better to put its resources behind creating new growth companies than continuing to attempt the impossible.