Does Microsoft really need a chief software architect?

Ray Ozzie announced yesterday that he plans to step down as Microsoft's Chief Software Architect, after five years at the company. The question for Microsoft now is not "Can Ozzie be replaced?" Instead, they should be asking whether the company needs a chief architect at all.
Written by Ed Bott, Senior Contributing Editor on

Ray Ozzie announced yesterday that he plans to step down as Microsoft's Chief Software Architect, after five years at the company.

The question for Microsoft now is not "Can Ozzie be replaced?" Instead, they should be asking whether the company needs a chief architect at all.

Trying to assess Ozzie's impact on Microsoft is tricky. He joined Microsoft in 2005 as the designated successor for Bill Gates. The company made it official the next year, giving him the Chief Software Architect title when Gates stepped down from full-time duties at Microsoft. But as my colleague Mary Jo Foley noted last year, Ozzie wasn't exactly a drop-in replacement for Gates. He "hasn’t found it easy fitting in culturally in Microsoft’s dog-eat-dog culture," she noted, accurately. Aside from a few keynote appearances, Ozzie has been almost invisible as a public face of Microsoft. And behind the scenes, I cannot remember ever talking to a Microsoft developer or manager who brought up Ozzie's name as a source of inspiration or ideas.

There's no question that Ozzie is supremely gifted intellectually. Ironically, though, he's best associated with products that don't exactly enjoy a reputation for excellence. His chief accomplishment before joining Microsoft was designing the product that eventually became Lotus Notes (which in turn was purchased by IBM in 1995). Although it was filled with groundbreaking ideas and continues to have a large user base, Notes was a critical flop. Find any review of Notes written in the past 20 years, and chances are it will include some variation on the following verdict: "Although Notes is an incredibly powerful platform for building collaborative applications, its user interface leaves much to be desired." Most companies deployed Notes as an e-mail program, where it became a source of intense frustration—I don't think I have ever spoken to a happy Notes user.

After Notes, Ozzie's next big accomplishment was Groove, which Microsoft acquired along with Ozzie himself. Groove's collaborative capabilities were added to Microsoft SharePoint, and it was a part of the enterprise version of Office 2007. Most of the Groove users I've met through the years have worked for Microsoft, where it inspired intense loyalty among developers. Among Office customers, it's still a well-kept secret. It's telling that for Office 2010, Microsoft kept the core Groove feature set but streamlined the product and renamed it SharePoint Workspace.

At Microsoft, Ozzie made his mark initially with a sprawling all-hands memo that outlined Microsoft's vision for cloud computing. Ozzie hit the Send button on "The Internet Services Disruption" almost exactly five years ago, on October 28, 2005. Re-reading that memo yesterday, I winced at the very first paragraph: "Next year we have a double barreled release of our two largest products with Windows Vista and Office '12'.  It’s a great time for customers, our partners, and for those at Microsoft who have put so much of themselves into these products." Vista, of course, was probably Microsoft's most spectacular failure of the last decade, and it's noteworthy that Steve Ballmer's memo announcing Ozzie's departure doesn't mention it at all.

There's no question that Microsoft has focused on cloud services since that 2005 memo, and it has delivered a fairly impressive portfolio of cloud-based offerings. On the consumer side, Microsoft now has a full range of Windows Live services, and it's done an impressive job of moving Exchange and SharePoint to hosted services that aren't just for enterprises anymore. Under Ozzie's watch, Microsoft introduced the Windows Azure platform, which has tremendous potential. But those individual products gloss over a failure of vision that's particularly acute for Microsoft. At the November 2009 Professional Developers Conference, Ozzie outlined Microsoft's "three screens and a cloud" vision. Microsoft is still playing catch-up on two of those three screens—phones and TVs—and it's telling that Ozzie announced his departure just before the launch of Microsoft's Windows Phone 7 platform.

Ironically, Microsoft has just finished one of its most successful years ever in terms of shipping products, led by Windows 7 and its online companion, Windows Live Essentials 2011. Although Ozzie's big ideas about cloud computing were the focus of that 2005 memo, his real contribution to Microsoft might have been more subtle and more lasting. This section, near the end of the memo, is a nearly perfect description of what went wrong with Windows Vista:

Complexity kills.  It sucks the life out of developers, it makes products difficult to plan, build and test, it introduces security challenges, and it causes end-user and administrator frustration.  Moving forward, within all parts of the organization, each of us should ask “What’s different?”, and explore and embrace techniques to reduce complexity.

Some problems are inherently complex; there is surely no silver bullet to reducing complexity in extant systems.  But when tackling new problems, I’ve found it useful to dip into a toolbox of simplification approaches and methods.  One such tool is the use of extensive end-to-end scenario-based design and implementation.  Another is that of utilizing loosely-coupled design of systems by introducing constraints at key junctures – using standards as a tool to force quick agreement on interfaces.  Many such tools are not rocket science: for example, by forcing a change in practices to increase the frequency of release cycles, scope and complexity of any given release by necessity is greatly reduced.

Indeed, the Microsoft that produced Windows 7 and Office 2010 has embraced many of those concepts. The post-Vista Microsoft products I've seen over the past three years have paid much more attention to detail and user experience than their predecessors.

Steve Ballmer says Microsoft has no plans to fill the Chief Software Architect role when Ozzie retires. Maybe that's a good thing. Microsoft has never been short of big ideas. Ironically, most of its disappointments historically have come about because the company focused too much on those big ideas, with an overemphasis on architecture and not enough attention paid to the process of actually building the product. Maybe Microsoft needs to spend the next few years dusting off some of those old blueprints and drastically reducing their complexity. In short, maybe it needs more skilled builders, not another architect.

Editorial standards