On SOA: Why is Microsoft Pushing The Rock Uphill?

I wrote recently about IBM dominating the SOA market. But that’s SOA as defined through a largely top-down lens. Many of the analysts here at TechEd have fostered the frame. So I found it interesting that there is a potentially really solid Microsoft SOA narrative, but Microsoft is somewhat wilfully refusing to just tell it. Microsoft currently talks to SOA under its own rubric of “Service enablement”.

Why not just have a differentiated SOA story and tell it? Some of the issue is historical – and some is now because of IBM’s current success. Microsoft likes to define the terms of engagement, but that’s not always the right strategy.

During a session this afternoon Jon Perera, general manager, application platform strategy, began to add snow to a snowball when he talked to the concept of “Real World SOA”.

What did he mean?

“We’re not saying start at the top and create a perfect SOA infrastructure and map to that. Rather we’re saying each new project look at how service orientation might help you get value from that.”

Amen, brother. One of my major concerns with SOA is when does modeling and consulting stop and SOA start?

On Incremental SOA

Perera talked to an incremental SOA strategy, an approach RedMonk thoroughly agrees with. Its crucial to get the people that do stuff to buy into SOA. It can’t be an entirely top down phenomenon. This is all goodness.

Jon talked to Directory Services and SOA, for example- still one of the best, clearest and easiest ways to get people to grok what SOA is.

So that much was good. If someone asks what Microsoft’s SOA story is, there is a story to tell. Unfortunately I know that in the field Microsoft is not telling that story. I know of two recent customer proposals in the UK- one to a public sector body and one to the risk management department of a major financial services, both of which pitched Office System when Microsoft was invited to pitch SOA.

If Microsoft doesn’t have a canned SOA story its salespeople can’t pitch it.

As we moved to Q&A problems began to assert themselves. Jon got a little prickly, and started talking to a need to pushback. Analysts were asking questions that Jon didn’t want to hear.

He made some good points about Microsoft’s commitments to message-based development, modeling, XML, and WS-I.

And to be fair, analysts and Enterprisey people have not always given Microsoft credit where its due when it comes to enterprise chops. Think how long the “SQLServer doesn’t scale” assumption lasted. Its quite clear many of the analysts here subscribe to Big SOA top down thinking. But that’s not to say that’s the only frame in town.

What’s The Story?

My major criticism of the conversation at TechEd 2007 so far is about a lack of story-telling. While Microsoft wants to avoid vision-oriented hand-waving, and vapourware at the moment, it seems to be pitching feature matrices. But its hard to engage with feature matrices. Microsoft needs to tell stories. But also stories that people are already socialised towards.  

I do think that broadening the notion of services makes sense. I have presented on what I call the Service Mass Convergence myself- given we have SOA, SaaS, IT service management and so on. But there is nothing wrong with just telling a story people get. Why swim against the tide? Why push the rock uphill? Why not just participate in the conversation and help the snowball roll downhill. Microsoft has a good SOA story to tell. But it needs to do so.

Microsoft is a RedMonk client in some areas of its business. The snowball metaphor belongs to Doc Searls.

Picture courtesy of Benimoto.