I've watched Steve Ballmer in coversations over the years and know he's a smart guy. Everyone has seen his Dance, Monkeyboy, Dance performances, so anyone can guess that his management style borders on the feral. After several hours of reading and rereading the transcript of Ballmer's talk this morning at the Sanford C. Bernstein & Co. Strategic Decisions Conference, I'm stumped as to why Microsoft sent Ballmer to present at this investor conference. He was the feral management bot, defyingWindows and Office are and are not the "core," according to Ballmer. This is emblemattic of a lack of focus. investors need for simple answers to questions about where Microsoft is going, using their money to take the trip.
Ballmer's disjointed presentation about the company, R&D investments, increasing operating expenses and a pile of cash was the subject of several different news reports, each focusing on one aspect of the presentation. But the whole was a perplexing mish-mash. This extract, which takes two sections about Microsoft's "core" investments, is representative:
We have to grow what I called our core. Core doesn't match segments. That kind of threw people off a little bit. But everybody in their heart and soul kind of knows what our core is. It has nothing to do with Windows and Office.... In our core, we've got a lot to do. Windows and Internet Explorer—Windows is a product thats got to be watered every—periodically. We've gone a bigger gap than I would like to go, Windows and IE. There's a lot of innovations still coming in Internet browsing, in hardware. There are so many things we need to do and if we don't keep Windows fresh, Windows will not continue to flourish. The same with our core Office product.
Windows and Office, as you can plainly see, are and are not the "core." This is emblemattic of a lack of focus. Ballmer also started his talk describing how Bill Gates had insisted the company had 60 or 70 things it needs to do within the context of the upcoming year's R&D work. This is alarming to investors, because it doesn't sound like Microsoft has any direction or, rather, way too many directions.
To be fair, Ballmer's describing Microsoft's product development efforts, which are labelled as part of the company's $7 billion in R&D spending but in fact represent product expenses. A very revealing moment in the talk came when Ballmer explained that only a quarter-billion dollars, or about 1/28th of Microsoft's annual R&D spending, is dedicated to pure research that might produce breakthrough technology: "Of course we always have our research group, which I feel super good at. It costs us about $0.25 billion a year roughly and it is the best spent $0.25 billion in our budget perhaps."
I'd be a lot more comfortable with Microsoft's increased spending if more of it were dedicated to advancing computer science, I/O technology, network services and security without regard to specific products, like the MSN/Windows Live software-as-a-service effort, because productization is infinitely more expensive due to the heavy reliance on marketing to take the idea from beta to commercial success. More high concepts from Microsoft Research would yield many more market opportunities.
Marketing for the Windows Vista and the next versions (packaged and "Live") of Office account for most of the additional $2.6 billion in spending Microsoft has slated for the next year. If the company spawned 500 good hacks and put them into the world with their Windows API hooks, they'd drive the adoption of Vista far faster than through advertising.
Ballmer's business logic likewise ran to the impenetrable. His answers to questions about building or buying technology, R&D vs. product development and "big products or little products" were consistently "both." He said being first to market was "more fun," which may be true but doesn't comfort investors who are used to Microsoft's fast-follower approach to software. For 30 years, Microsoft has capitalized on being second and ubiquitous, not first and fun.
The really cogent argument Ballmer did offer for the reliance on internally developed technologies and products was XBox. He pointed out that, if Microsoft had acquired Nintendo (whose U.S. headquarters is just down the street and now surrounded by Microsoft's campus), it would have spent three to four times as much as it will lose in the early years of XBox and "we would have had a business we didn't understand as well, wasn't as good, wasn't as well-positioned and essentially would have cost our shareholders an additional $6 billion, $8 billion." All true, yet it doesn't seem to inform Ballmer's strategic approach to extending Microsoft's business.
If the build-it-don't-buy-it case is proven by XBox, Ballmer should be much stronger in his defense of Microsoft's $35 billion cash reserve, more of which should be going into basic research now to put the company at an advantage later. He called having that much cash a "first-class" problem, but he didn't offer any proof that it was being used better by the company. Instead, Ballmer merely insisted that there was a lot to do. His comments about "role-based productivity" software are interesting, but the fact that is a "big investment" doesn't translate into projectable spending and return on investment.
Of course, that would mean making a promise to deliver product on time, so, well, maybe there is a reason for Ballmer's vague statements.
All in all, the event doesn't give me much confidence that the reorganizations and realignments at Microsoft are translating into better use of assets. There's a lot to do, Steve, but maybe you'd get more mileage from a fanatical focus on finishing one or two "core" products or, even, getting out of some segments that would be better served by third-party ISVs who use Windows technology but could feasibly support cross-platform enterprise environments with a straight face. It seems to me that most of the changes at Microsoft are still to come, and that the fortress office of the executive may be where change is needed. If only Ballmer and Gates do more embrace change than to reinforce Windows, it would be a win.