Access and Feeds

SOA: ROI Questioned

By Dick Weisinger

Only 37 percent of enterprises employing SOA are seeing any return on their investment. That’s the conclusion of a recent report done by Nucleus Research and Knowledgestorm. It’s a statistic that knocks the wind out of a lot of recent hype around the technology.

But the numbers shouldn’t really come as a surprise. The same discussion was already happening back in March.
The low number may well reflect that to date many companies have been just kicking the tires of SOA and trying the technology out in small pilot projects. The real benefits of SOA aren’t expected to be seen until the implementation spans the organization. Small pilot projects are, in effect, just more isolated spinning wheels that, rather than simplifying, can just add more complexity to the overall picture.

Short term, adoption of SOA may require more planning, and that in turn can mean higher budget and result in longer schedules. SOA is better viewed as a more strategic move rather than one that immediately contributes to the bottom line.

But some think that ROI metrics relative to SOA are not totally unmeasurable. They advocate an approach of looking at SOA benefits relative to early pilot and departmental projects. Looking at incremental differences in costs between successive projects should begin to show the benefits of SOA as reusability kicks in and new projects can take advantage of a growing SOA infrastructure.

Still others think that the best way to measure the benefits around SOA are to look at developer productivity. The Nucleus Research report found that SOA developers, while only accounting for 4 out of 10 of the enterprises surveyed, had jumps in productivity of 28% after adopting SOA. But then measurements of developer productivity can usually be taken with a grain of salt — getting a handle on how to measure software development productivity is a much debated topic and something which is very hard to do well.

The bottom line is that many feel that SOA is inherently a good, or the right, architectural approach, but that it might take some time before we can really have a good set of metrics that can provide good comparisons between SOA and any alternatives.

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Leave a Reply

Your email address will not be published. Required fields are marked *

*