Enterprise Software

Microsoft: 'We're bringing SOA to the masses'

Microsoft is gunning for the SOA accounts of small to medium businesses or smaller departments of larger enterprises. Joe McKendrick says this type of strategy has been the secret to the company's success.

 

This is a guest post from Joe McKendrick of TechRepublic's sister site ZDNet. You can follow Joe on his ZDNet blog Service Oriented, or subscribe to the RSS feed.

"Microsoft is bringing SOA to the masses," said Steven Martin, senior director of product management in Microsoft's Developer Division, at the vendor's latest SOA and Business Process confab. (Covered here by eWeek's Darryl Taft.)

That's been the secret sauce all along. In recent years, there have been plenty of questions raised about Microsoft's SOA strategy (or lack thereof), and why the vendor seemed to be stuck in 2003 in terms of advancing in this space. But I've contended all along that MS is timing its efforts to mass market disruption -- it wasn't aiming for those high-margin, consultant-intensive implementations we've been hearing about at the Global 1,000 companies. Microsoft doesn't want the SOA accounts of large manufacturers, government agencies, or big banks -- at least not yet. Instead, it's gunning for small to medium businesses or smaller departments of larger enterprises.

MS understands the long-term success that is gained by playing the role of the disruptor that goes in and caters to the underserved or unserved mass market. That's been the secret to its success all along -- commoditize pricing to the point where individuals and small businesses can buy the technology, and make it relatively simple so that a software engineering degree isn't required to run it. That's the strategy we see behind the vendor's Oslo modeling offering for SOA work -- make things simple enough so that non-developers can get involved. (Of course, MS doesn't always accomplish this so handily.)

The question is, can Microsoft get SOA to the masses before the open source or cloud providers -- JBoss/Red Hat, Apache Project, MuleSource, OS2, Iona, Amazon Web Services, Google etc. -- captivate the masses with low-entry priced solutions? Either way, the costs related to the technology that underpins service-oriented architecture efforts will drop significantly. And that's a good thing.

However, the organizational and operational costs around SOA -- that's another thing. We'll keep exploring that in this blogsite.

5 comments
Tony Hopkinson
Tony Hopkinson

That will be about as popular as the chap who suggests re-organise your entire Business IT effort to switch to linux.... It will save you money, once you invested a substantial proportion of your diminishing profit in it. That's aside from the long term business case for a service consumer doing it, is iffy at best.

Justin James
Justin James

Personally, I've always viewed SOA as one of those monolithic projects that large companies like to embark on... that's the main diffference (in my mind, at least) between "SOA" and providing various Web services internally. But I could be wrong. I honestly think that this whole "SOA" thing is completely muddled by terminology. J.Ja

rob mekel
rob mekel

you're right ... SOA looks very much like WEB-services. Even further it looks like OO-development. It's all about multiple(re)use of bits and pieces developed software.

jck
jck

is all about code re-use. That's the foundation of what classes/objects are...one basic template from which you derive like objects, or grow similar (mutated) objects (super/sub-classed objects). SOA is essentially is just...black boxing. You really don't have to OO the SOA, though. To be honest, I think so much is driven by Microsoft in this arena to perpetuate their business model by channelling everything to use a single or narrowed-down paradigm. When in reality, OO (like any other programming architecture) has it's place and use in various times and situations. For instance: would you write an entire class to do the "Hello World" program in one line of code? No re-use ever. Is OO needed in this case? Sometimes, the KISS principle is what people need to be reminded of: Keep It Simple, Stupid At least, that's what my DB Design professor taught me in college...wise man, he was.

rob mekel
rob mekel

SOA, Web-services, OO all relate to re-use of programmed code. KISS is the methode that everlasts. Lots of ppl forget this simple rule for the best products. Guess you learned a lot of your DB-Design-Prof.

Editor's Picks