In a recent post by Loraine Lawson on ITBusinessEdge, an informal survey was cited that referenced a majority of mid-market cios who “said they had no current business need for SOA.” I was a little surprised by the headline since MuleSoft continues to see tremendous adoption of our open source Mule ESB and subscriptions of our enterprise version among companies I would describe as mid-market. So, I decided to read further and try to learn more.
First, the post doesn't provide a link to the actual survey results, so the only information available is what can be inferred from the conclusions that Lawson provides. I think it would be very helpful to share the full results.
The main takeaway I read was that at least some mid-market CIOs view SOA as a buzz word that vendors use to sell software they may or may not need, and that by doing modular programming they are already reaping the same benefits. To an extent I agree with this. We all remember the debate just over a year ago around Anne Thomas Manes' New Year's obituary for soa. Top down big-bang SOA is something that probably many midmarket companies do not need to be investing in (the same applies to some larger companies). As Ross Mason previously pointed out, “These approaches have failed because they make the average developer's life more difficult, rather than simpler.”
I don't agree that midmarket companies are not taking the next step beyond modular programming. What I have observed in our community and customer base is they are taking pragmatic steps towards service orientation using best of breed tools and architectural best practices that continue the investment they have made in existing systems. If done properly and with proper governance, this better prepares the business to respond to new opportunities, increased competition, or other disruptions in their industry.
Therefore, the responses seen in this survey could actually be the result of a disconnect between what is observed at the cio level and what is actually going on in the organization. This is not the fault of the organization or CIO. CIOs have been conditioned by vendors, analysts, and media to think of SOA as an expensive top-down proposition that requires a huge up-front investment, leading them to think that they need a large project to justify the cost, so they conclude that there is “no current business need for SOA.”
Open source and lightweight tools like Mule ESB completely change the cost equation, as well as the adoption pattern -– development teams are downloading Mule and implementing projects in a bottom-up fashion. In short, a lot of these developers are doing service-oriented development and following SOA principles without knowing (or caring) that they are, in fact, doing SOA.
At MuleSoft, we are seeing strong adoption because we offer best of breed tools that a developer can easily download and start using. No SOA salesman needs to come knocking on your door pitching SOA, and you don't need to justify big checks for perpetual software licenses and training classes (not to mention the new hardware or upgraded development systems that might be required to run the typical vendor stack). You recognize a problem (or an opportunity), discover how others are currently building similar solutions, and get started learning which available tools fit your project the best. It's my experience that this is the best approach, especially in smaller businesses.
