If enterprise-software market size, growth trends and market dynamics in general, such as adoption of software-as-a-service (SaaS), matter to an ERP or enterprise-software decision you are making, be cautious about your sources. All this information should be important to you for the reasons I mentioned in my post"Is Your Walk Down Enterprise Software Memory Lane Bittersweet?" But you need to be careful not to depend on statistics about a $250 billion market taken from a few paragraphs in a 20-paragraph article.
Here's an example from CIO magazine about the ERP market, called the "Future of ERP." The CIO article is actually part roundup story (it interviews all the vendors' marketing guys) and part statistics. I link to the first half of the article because that's where the statisitcs are; look near the end of the Part I. The article says
By the end of 2009, Gartner forecasts that global SaaS revenue will reach $7.5 billion, which is an 18 percent increase from 2008 revenue of $6.4 billion.
But the article doesn't tell you that $7.5 billion represents only 3 percent of all software spending. I'm a big fan of SaaS and worked on a lot of the initial market research on Application Service Provision (ASP) in the late 1990s. I have accepted the fact that 3 percent of the pie at this point is not a ringing endorsement of my forecasting abilities. I should have seen the effect on SaaS of the dot-com bust (I did predict the bust back in 1998), but I couldn't have predicted the 9/11-related mini-recession of 2002.
SaaS is growing in 2009 whereas the overall software market is flat to down, so there might be an argument that SaaS makes sense in down times. And good business management theory says: If it makes sense in down times, it makes sense all the time. But it is not the be-all/end-all that the SaaS fanboys are talking about. How come?
The CIO article goes on to point out that Gartner estimated that only $1.17 billion of the $7.5 billion in 2009 is for ERP. The article doesn't say what the rest of the software delivered as a service is, but presumably it is CRM and similar standalone applications for functions such as HR admin, expense management, project management and collaboration. Salesforce.com alone should account for a billion of that, I think. So that should give you a hint about what functions you should be thinking about for SaaS delivery and which you should not. SaaS continues to be in the area of nice-to-haves, not have-to-haves.
That leads to the biggest missing statistic in this and similar articles that set up a non-existent split between SaaS vendors and non-SaaS vendors (or on-demand vendors and ?????). I don't know what Gartner's methodology is, but ask me my estimation of the company that is probably second to Salesforce.com is SaaS. Is it some high flyer getting all kinds of great PR? No, in my opinion, it's that evil company, villified in the CIO article for high maintenance prices and vendor lock-in tactics and other kinds of dastardly deeds, Oracle.
According to CIO according to Gartner:
Adoption of the on-demand deployment model has continued to grow as on-demand vendors have extended their services through alliances, partner offerings, and more recently, by offering and promoting user application development through platform as a service (PaaS) capabilities.
But if my estimate that Oracle is No. 2 with Microsoft and even SAP coming along in the next few years, what's an "on-demand vendor" as opposed to a non-on-demand vendor? Intuit even does more than $300 million in SaaS, although some of that is consumer, of course.
Market researcher Saugatuck Technology is quoted in the article. As I have written earlier in 2009, Saugatuck is a flagrant bunch of SaaS fanboys. It says that by 2014
SaaS will become integral to infrastructure, business systems, operations and development within all aspects of user firms, with variations in status and roles based on region and business culture.
I guess that depends on how you define "integral," but even if, as I now forecast will happen, SaaS jumps to account for 10 percent of all enterprise software spending in 2014, up from the 3 percent in 2009, I'd hardly call that integral.