Bill Gates' recent announcement that Microsoft would soon be in the software-as-a-service (SaaS) business should be taken as a warning sign to the faithful: Something is rotten in Redmond.
In the past, Gates has aimed his message at the consumer, both business and personal. He usually extols the virtues of whatever technology is being unveiled and explains to his audience how it will fundamentally change their lives (for the better, of course). This time he had nothing substantive to offer them.
Instead, Gates seemed to be reassuring the financial community, devoting an inordinate amount of his time to talking about the coming revenue bonanza from online advertising and subscription services.
Of course, in supporting a technology for which he has no worthwhile capabilities yet, Gates is on familiar ground for Microsoft.
And by his own admission he couldn't introduce Live Software as a new technology. True, the giant from Redmond has been in both positions before - Word versus WordPerfect, Excel versus Lotus 1-2-3 - but this time it is different.
High tech historians will look back and see that it all started with the Internet versus the C: drive. The truth is, in making that transition to the Internet, Microsoft has not been a clear winner. It's kept up, but it certainly hasn't dominated the Web. The problem is that, prior to the Web, Microsoft and its competitors all played by the same rules: software licenses. The Internet and now SaaS have a different set of rules, both in terms of business models and technology.
On the business side, there's the issue of cannibalisation. Although Gates claimed that neither the Office productivity applications nor Windows itself would be offered up live, as he called it, he is indeed leading his company down a slippery slope.
Given the advent of Web services and SOA architecture, who says the same plug-ins Microsoft entices the developer community to design for Office won't also work with StarOffice? Gates' move seems likely to make his own company's products less relevant.
Brent Arslaner, vice-president of marketing at Jamcracker, a services company that helps ISVs transition applications to an on-demand world, said that a traditional ISV used a direct and indirect sales force built around on-premises solutions. But when they moved to a SaaS model, companies must count on advertising or subscriptions to supplant licensing revenue, which could fly in the face of Wall Street's expectations.
As we saw with Siebel, a company that was architected to make money by licensing its software with million-dollar installation deals will have a tough time transitioning to a model that charges $US90 per month, per user. Arslaner also said it is unclear what Microsoft will become. Is it a service provider or does it rely on solutions partners? Who is going to provide the SLAs and the 24/7 support?
Architecting applications for on-demand use is also fundamentally different from building ones that live on the hard drive. It raises issues around secure access, provisioning, and, of course, the notion of multi-tenant administration. In a multi-tenant environment there is one instance of a database that has to scale to thousands of users, each having their own virtual, secure environment.
That would indeed be a very large shift for Microsoft - which, if you think about it, never really got off the C: drive.