Wintergreen Research (via Report Buyer) has just released a six-year forecast of the SOA infrastructure industry, calculating that the market will grow to from the current $3.3 billion to $10.3 billion by 2015. (At 17% a year growth.)
You may recall that Research & Markets issued a market estimate last year, putting the SOA infrastructure market as growing from $2 billion in 2007 to more than $9 billion by 2014 -- which appears to be in line with Wintergreen's estimate -- and therefore showing that the market continued to plow ahead through the recent economic rough patch.
But should SOA market sizing be limited to infrastructure spending? AMR Research, for example, sees a much larger market ahead, estimating that the SOA market will grow from the current spending level of $28 billion to $52 billion by 2014. AMR analyst Ian Finley said the AMR estimate is based on 600 companies globally, and covered a broader range of software and services. "Other forecasts were based only on a narrow set of software peculiar to SOA," he explained. And, since the AMR forecast was issued last year, Ian adds the caveat that "we're less optimistic about big growth in 2009 but still think the CAGR will be high relative to other markets."
So, because it's so pervasive and includes so many elements, the SOA market appears to be a difficult one to measure. It's also just as difficult to measure SOA adoption among enterprises, since companies have different ideas of where they're at with SOA. And Ian makes a good point -- since SOA methodologies and practices are being baked into just about every vendor's products, this needs to be included in market sizing estimates.
For example, there are estimates that the entire global software market was at about $203 billion in 2006, to grow to about $272 billion by 2011. So, roughly based on these kinds of numbers, and figuring the SOA market will be somewhere in the neighborhood of $50 billion (based on AMR's methodology), SOA may encompass about a fifth of the total software space in a few years. Given, these are entirely separate estimates with different methodologies conducted at different times -- but this makes more sense to see SOA as a larger piece of the software market, not as a small fraction of three or four percent. (Again, Wintergreen and Research & Markets kept their market definitions narrowed to direct SOA infrastructure.)
The new Wintergreen Research report also gives IBM the honors of dominating the SOA space with 70% of the market, while the rest of market is divided between 12 other participants with measurable market share. Report Buyer credits IBM's horizontal and vertical arrays of infrastructure products.
The report also suggests that "SOA represents a way to decrease IT costs by up to 90%." So SOA can bring IT costs down to 10% of what they formerly were -- sounds great, but I haven't seen evidence of this yet. We need to see examples of how this is possible.
Wintergreen Research also seems fairly bullish on how SOA is progressing:
"SOA reaches into every industry and every segment of the economy and represents a fundamental change in the way automated process is delivered to replace manual process. Service enabling offerings are a response to the fundamental change in IT, where enterprise competitive advantage is gained from having IT flexibility."