Mashup or meltdown for Microsoft?

Microsoft's mashup strategy is an ugly, squidgy mess. None of us would want to take it on. But legacy is often like that, and Bill Gates and his lieutenants have no choice but to deal with it. They have to patiently marshal and mold their products and platforms until some kind of recognizable strategy finally emerges from the sludge.

The biggest problem they face is that different parts of the organization are progressing at different speeds, and from vastly different starting points.This is not where you would have chosen to start from If you had set out to develop a product strategy that's fit for the mashup age, this is not where you would have chosen to start from. There's no knowing whether some business units are ever going to catch up. Microsoft's leadership is just going to have to push and prod and hope for the best.

These are the trickiest mashup challenges they face:

Dynamics faces a huge struggle, like all the leading business software suites, in moving to a services architecture. Gates has been setting out some of the things that need to happen in interviews and talks at the division's user conference this week. The complexity of the roadmap sums up Microsoft's problems in a nutshell. To start off with, the entire suite has to be completely rewritten to adopt a services-based architecture founded on .NET. But it also has to tie into initiatives happening elsewhere in Microsoft, in product ranges as diverse as collaboration, business intelligence, development tools and a new generation of mashup services — as well as in the underlying Windows operating system. Keeping such a disconnected roadmap in sync will be tough (Oracle and SAP of course have similar problems although probably not quite as dissipated as in Microsoft's case).

MSN/Live just makes matters worse, because as Mary Jo Foley pointed out yesterday, there's a strong case for delivering at least some existing or additional product capabilities as hosted services. So the Dynamics group, for example, has got to make sure that its products mashup well, but not so well that it upsets its powerful partner base by doing mashups with Microsoft-hosted Live services that undercut those partners. Then from the other point of view, do the MSN/Live developers really want to spend valuable time waiting for the Dynamics guys to catch up with them? Microsoft will find it desperately hard to make Live simultaneously successful with both business and consumers because of the hugely different pace of adoption in the two camps.

Office and Windows meanwhile have to carve out a new destiny as the default on-ramps to the Web for, respectively, office workers and consumers. This is the most mission-critical element of the strategy. Microsoft desperately needs users to discover all kinds of benefits to accessing Web-based resources directly from within the Office environment and the Windows desktop, otherwise it's finished as the world's most powerful software company. Is that a mission it can achieve while simultaneously keeping other product families such as Dynamics and Exchange/SharePoint up to speed? Something may have to give.

Exchange/Sharepoint and related products are a conundrum. From one point of view, they're the closest Microsoft has to fulfilling its 'server=service' vision — the notion that the same product can perform either role — but on the 'service' side of the equation their functionality is dire, with limited capability for multi-tenancy and highly idiosyncratic mashup interfaces. Some radical surgery is required, but anything too drastic wouldn't go down well with the installed base. With a lot of Web-based competition breathing down Microsoft's neck in this space, there isn't much time to spare.

Server solutions, finally, present a particular headache that could wreck the whole party: licensing. How do you move from a machine-specific licensing regime to one that caters for service providers operating a grid of servers delivering services to a multitude of sporadic users? (And which supports a partner channel, too). Microsoft hasn't worked out the answer to that yet, and the clock is ticking here too.

None of these are easy problems on their own. Microsoft has got to try and deal with them all at once, without alienating three constituencies that often have conflicting demands: enterprise customers, consumers or channel partners. This is a mashup challenge that I'm sure Bill Gates will relish, but if he and his executive team can't find a way to juggle all these competing claims, it could mean meltdown for Microsoft.