Fixing the flaws in Microsoft's big picture

Project Constellation was the code name for the now-aborted secret takeover bid by Microsoft for SAP -- or Mensa and Sagittarius as the document coyly calls them. Revealed in court as part of the ongoing Oracle antitrust trial, the presentation has been heavily edited to remove figures and other specifics. But as a window onto Microsoft's highest levels of corporate consciousness, it's as fascinating as slipping Bill Gates two grains of sodium Pentothal during a Vulcan mind-meld festival.

Let's dispose of the code names first. It doesn't take a genius to spot the true identity of a large company called Mensa that makes a product called Office, nor the real names of the major players in the enterprise applications space: Sagittarius, Ophiuchus, Sirius and Pegasus. Presumably in place so that company officers could discuss the deal outside the boardroom without giving the game away if overheard, the effect is to give the document an air of Golden Age science fiction.

The structure is as you'd expect, but only up to a point. It discusses what a great match SAP and Microsoft would be, as they do such different things and address different parts of the market, and how much scope there is to make economies by removing duplicate R&D and marketing. Passing swiftly over those inherent contradictions, it settles down to the boardroom-pleasing predictions of huge gains from flogging Office on the back of SAP, getting "increased monetisation in small and medium business segments" from "more integrated and complete solutions" -- in other words, putting all business IT into one package and charging what it likes -- a "seamless enterprise application roadmap" (see above) and so on.

It's not until page 20 that any downside is mentioned, under "Issues and Risks". The top issue is that the Unix/Linux side of SAP's business is likely to flee, at speed. Then comes the fact that the deal is likely to take up to a year to conclude, with 'distracting' PR all the while. Mm. Then there's an interesting line -- "Mensa has historically had difficulty managing remote R&D": about the only profound observation in the pack.

But something's still missing -- or at least, hidden away. That 'distracting' PR -- well, you can guess what it is. If Bill Gates buys so much as a can of Coke these days, people pick over the regulatory and monopolistic implications: imagine what howls would rise if Microsoft took aim at the biggest supplier of enterprise applications on the planet. There would be top-level governmental investigations in every market. Microsoft works in a global environment where antitrust is part of their business like disappointment is part of being an England football supporter. But mention is there none -- antitrust is an annoyance that can be wished away and presented as bad PR.

It's absolutely true that Microsoft believes wholeheartedly in innovation -- on its terms. The Constellation Project presentation listed among the many benefits for the takeover multiple chances to innovate, mostly in the way customers get at their back-end data and in business process customisation. That the company could do this anyway by actually supporting the open standards and cooperation that it publicly claims to love, doesn't seem to have occurred to it. To innovate means to own and control.

You can see this every day in what the company does. Steve Ballmer proclaims that antitrust lawsuits stifle innovation on the same day that the company unveils a facelift for MSN Search -- a face lifted from Google. That's not what most people would call innovation, but it's the sort that Microsoft believes in most wholeheartedly, and why antitrust is considered so dangerous.

Can this change? It can and it will, and the change will come from an unexpected quarter.

The Microsoft mindset comes in part from its almost cult-like self-belief and hermetic self-containment: it only recruits people who it knows will sign up to the Way, and it puts them on campuses where there's no need to leave. Individual and organisational goals are identical. This leads to personal identification with the company and a fundamentalist attitude to outside criticism: if it's against the company, it's against me and ipso facto wrong.

Cultic thinking rarely survives prolonged contact with the real world -- remember Microsoft's 'difficulties in managing remote R&D'? But that contact is increasing exponentially; there are hundreds of MS employees at all levels engaged in blogging, More start each day. That means feedback, and contact with outside perceptions where before there was almost none. Instinctively, the company is nervous about this -- there have been some rather public slappings -- but equally instinctively it knows that it cannot control it.

The big leaks -- the Constellation memos -- may be the headline grabbers. The small leaks are more important, because they go both ways. Each new MS blogger is introducing one small dollop of reality back into the company, setting the groundwork for the inevitable revolution. Microsoft is its people, and its people are learning to be free.