Windows 2003: Could the crusade backfire?

Peter Judge: Microsoft's religion is one where products are good and services are a sin. But a big server product launch could alienate the very souls it wants to convert

With the launch of Windows Server 2003 only days away, Microsoft is building itself up to a head of steam in its efforts to sell to the enterprise.

There is a great deal riding on this product launch. It may not be changing the kernel the way XP did on the desktop, but it is a significant update to the server part of Microsoft's story. And let's face it, as far as Microsoft is concerned, the server, and the high end of the enterprise, is where it faces significant opposition, and sees potential opportunity.

If past Windows launches are any guide, Microsoft's launch will not be quiet. It will pour on the marketing budget and hold forth with a lot of claims about how different things are going to be from now on, thanks to the new product.

All of which means that now is not necessarily the best time to talk to someone from Microsoft with the word "evangelist" in their job title, but it is certainly a time when such people are very much in evidence.

I've said before that it will take nothing less than a religious conversion to persuade chief information officers to rely on Microsoft, and it seems like the company has already taken this to heart. In the run-up to the launch, it is fielding evangelists.

Charles Fitzgerald is a .Net evangelist at Microsoft. His full title is "general manager of platform strategy, Microsoft's developers and platform evangelism division", a job description which neatly combines verbosity, religious zeal and a sense of a massive quasi-military organisation dedicated to spreading the Microsoft religion to its assigned demographic.

I met him, and came away with a deeper impression -- if that were possible -- of the corporate emotion that Microsoft is investing in its bid for the souls of CIOs.

It's actually a year or two since I met someone with "evangelist" on his card. The professional tech evangelist dates to the bubble years before 2000, when people spread the gospel was of how "cool" stuff like Java was.

Professional technology evangelists used to be gently spoken, enthusiastic and ironic about their job titles. Simon Phipps was certainly that way as Java/XML evangelist at IBM; you can judge from this column about open source whether he has kept a sense of proportion since he "converted" to become a technology evangelist at IBM. Tech evangelists are different now -- at least at Microsoft. Perhaps the stakes are higher. Money doesn't grow on trees any more. The company's future depends on the harvest these guys can reap. Whatever the reason, Fitzgerald is deeply engaged. He preaches a gospel of salvation through products, not services. When Fitzgerald talks about saving CIOs, he is talking about saving them money. Products are tangible things that IT departments can buy and make work, whereas other vendors get their cashflow by hooking the customer onto a steady stream of consultancy. When he promotes the Microsoft faith, he does so by condemning the sins of other faiths. And the biggest sin is services. In his world, Microsoft is a citadel of virtue, promoting products that do the job. Others have products that don't work and -- even more heinous -- they rely on services. IBM is the big demon here: "We sell technology, (IBM sells) busloads of consultants," he says. "I wouldn't want to change positions." Despite this, Microsoft and IBM have collaborated on BPEL, a proposed standard for orchestrating services -- though both have bypassed a formal standard on the subject at the W3C group. It seems that standards committees are like inter-faith discussion groups; all very well, but not to be allowed to distract from the true path. Microsoft and its partners have promised that BPEL will eventually be offered, royalty-free, to a suitable standards group, though there is no indication yet what that group will be. Fitzgerald defends this "trust us" approach and damns Sun's standards machination: "We can't make promises absolutely or forever, that would be either stupid or a lie," he says. "Other companies are not particularly royalty-free. Java is not royalty-free. Sun pulls off hypocrisy to a degree that boggles my mind." He has a point there, but he wastes little time on Sun. Sun is, after all, in Microsoft's view, a "mere" hardware company. Given Sun's deficiency in services, Sun's pitch to the enterprise is surprisingly close to Microsoft's: "Buy our products (in this case Sun N1) and pay less on services." Microsoft's righteous indignation is reserved for IBM, it seems. "For every dollar of (IBM) middleware, you need $18 of consultancy," says Fitzgerald. "We say spend the dollars on getting people to make the stuff work first." Sundry Microsoft users may choke on that one. After all, let us remember that IBM is the one that is currently running vast numbers of datacentres, and Microsoft is the one that wants to run them. But (at least apparently), the anti-consulting bit is a religious thing: "Once you get the consulting habit, you can never go back and be a product company again. You'd have to move from $2,000 per seat, to $99 per seat." IBM's WebSphere, in his view, is a symptom of this problem: "WebSphere is more than 300 random parts, it has no software architecture," he says. "It's the IBM middleware museum -- 30 years of products in one box. However, to tour the museum, you need a couple of busloads of consultants." It's actually a frightening prospect in many ways. But, like all religious movements, Microsoft is not as monolithic as its evangelists would like to think. A few weeks before I met Fitzgerald, I met Microsoft's Consulting Services group in the UK. Now Fitzgerald might give the impression that Microsoft doesn't have a consulting group, but in fact, although it is more about showing solutions than partnering with users, it does exist. "Flogging CDs and showing demos is not the way to talk to enterprise customers," said Peter Cummins, who runs Microsoft business critical consulting in the UK. The group has done a handful of projects, each of around ten people-years. "We don't want 250,000 consultants, like IBM in the US, but we need to operate in the same world. We don't put in teams of 50, like Accenture. We remain a product company." So, whether it is hard-sell, or soft-soap, Microsoft's religion remains at odds with the traditions of the datacentre. Giving Windows Server 2003 a big product launch should play well with the space that already runs Wintel. But in some quarters it could backfire. It may be that Microsoft wants to confront CIOs with a different world-view, in order to save their souls (and it assures us, their wallets). But the very nature of a Windows product launch could serve to underline to CIOs just how very different Microsoft's world is to their own.
More enterprise IT news in ZDNet UK's Tech Update Channel.

For a weekly round-up of the enterprise IT news, sign up for the Tech Update newsletter. Let the editors know what you think in the Mailroom.