Java Reference
In-Depth Information
So it may be that SOA is a good idea for some businesses in some contexts, and just because
you can't map an ROI precisely doesn't mean you shouldn't do it. It may also be the case that
there is ROI in SOA, but that attempts to show it will only subject you to the fallacy of False
Precision. One example of False Precision is to suggest that the normal human body temper-
ature is 98.6 degrees. This is simply a traditionally used value, and there is no actual “normal”
temperature that can be identified to a precision of a tenth of a degree. Instead, there is a basic
range of perhaps 97-99 degrees in which different people normally operate.
The point of this recipe is to get you thinking about all of these factors in your organization.
You may have to do some work to “sell” SOA within your organization, and only you can
decide on the best way to determine the ROI for SOA. But consider if you're asking the right
question, or even a fair question.
Finally, just because SOA is an architecture that includes the business within its scope in a
variety of ways (including governance, BPM, and so forth), consider carefully if you are ask-
ing SOA to be responsible for more than it can reasonably be. Consider the well-publicized
fringe benefits Google offers, including onsite dry cleaning and a gourmet chef. What's the
ROI for gourmet food? Could you specify, with anything other than wild speculation, a return
on investment in Kobe beef instead of ground chuck?
I have tried to give reasonable voice to both sides of the argument here. But sometimes, if
solving a certain problem is really, really hard, it's because we're trying to solve the wrong
problem.
One simple and perfectly reasonable way of looking at SOA ROI is to present SOA as a solu-
tion, and compare it with the cost of not getting a solution.
