So SOA (Service Oriented Architecture) was the talk of the town throughout the 2000s, at least by enterprise software vendors. But SOA adoption was more of a pitter patter than a bang.The first major issue was funding. How do you get the business to buy-off on an enterprise-wide refactor without offering them any real new functionality? Sure you could tout the eventual cost savings and interoperability. But that's a hard argument for CFOs to swallow.
The other issue with SOA adoption was the change management challenge, particularly for large organizations. It's hard to get dozens of development teams (often geographically dispersed) singing the same song to make SOA work right.
But there were some successful SOA adoptions in the last 10 years and the most successful were iterative, not big bang. Meaning that over a period of time, enterprise apps would one-by-one, refactor into a SOA model, exposing reusable services across enterprise apps while they incorporated new functionality for the business.
If you're still trying to get your SOA program kickstarted, mobile could be the catalyst you're looking for. The best enterprise mobile apps aren't writing new business services, but integrating with ones that already exist. And not every feature needs to be refactored, just the ones that make sense in a mobile context (i.e. iterative refactoring).
So as you begin your enterprise mobility initiatives, think of it as a way to kickstart your SOA adoption as well. Put some thought into using a consistent set of patterns and technologies when exposing your existing enterprise services to the mobile channel. You just might get your enterprise "service-oriented" quicker than you think.


0 comments:
Post a Comment