SOA is hot, governance is hot. When Mercury asked customers what their priorities were for SOA, "The first thing they were asking about was governance," Mercury's SVP strategy, Zohar Gilad, told me yesterday. It was unexpected, but on consideration it makes sense, he added: "Customers need to chaos-proof their SOA implementations."
Why was Mercury asking about SOA? "We believe that SOA is the next big shift in IT," said Gilad. "It's definitely becoming the next big shift in IT."
So buying Systinet was obvious, really. Perhaps more intriguing than the question I've led with (and which I'll return to in a moment) is the corollary: Why did Systinet sell to Mercury? Readers of October's Loosely Coupled monthly digest on change management may recall we reported a hint from a BEA executive that the vendor might cement its OEM relationship with Systinet by acquiring the company. Systinet also has an OEM relationship with Oracle, and as the de facto leader of the SOA governance market it partners with just about everybody else of any substance in the SOA landscape. So it has probably seen a good number of offers, and yet preferred to stay independent. What was different about Mercury?
The official answer, of course, is that a merger with any one platform vendor might have poisoned those other alliances. "We were a Switzerland of sorts in the way we worked in the SOA ecosystem and we believe it's important to continue that," Systinet's CEO Tom Erickson told me yesterday. "Mercury has a tradition of working with many vendors, including BEA, Oracle and SAP."
It's no use me probing because I wouldn't get any other answer than that, but I can speculate. The first thing that leaps out is the price. According to Loosely Coupled's SOA Management Report 2005, Systinet raised a total of $27.4 million in venture finance, and there's been no word of any further rounds since the company seems to have had solid enough revenues to fund recent growth organically. So the sale price of $105 million in cash is a very tidy return on that investment, and one that further cements founder Roman Stanek's reputation as a very dependable and insightful tech entrepreneur. One factor that may have contributed to getting such a good price is Mercury's recent NASDAQ delisting for missing filing deadlines Mercury needed to have some good news to tell.
The timing worked well for Systinet, too. There have been suggestions that the company even though its dominance of the SOA registry space meant it was well-placed to be a big player in governance was slightly wrong-footed by the speed at which governance emerged as such a major concern for SOA customers. This was especially evident last July, as I argued at the time, when it announced "Blizzard", a product that wouldn't be ready until the end of the year (and which hasn't yet appeared). Certainly it no longer has the governance field to itself, as I'll discuss shortly. It may have sold at its peak.
Finally, there's the question of the deal itself. Being all-cash, there's no incentive (apart from any earnout conditions that may have been negotiated to keep management in place for a year or two) for key figures to remain at Mercury once the deal is done. Gilad assured me that Mercury will keep Systinet's Prague R&D base in operation, and the business will operate as a wholly-owned subsidiary of the parent. But I still wonder whether Roman Stanek and other key personnel will be there in a year or so's time. Perhaps Roman's already planning some other venture. Who knows?
The final factor, I guess, is whether Systinet's people believe they've found a "good home" in Mercury. I suspect they probably have. Certainly Systinet seems to fit well with Mercury's mission statement of being the primary supplier of "Business Technology Optimization" (BTO) software, despite Mercury's irritating boasts of being the leader in BTO (having invented the category, it would be a calamity to be anything but the leader). The adoption of BTO extended Mercury's reach out of its roots in software testing into application management and IT governance. Registry, Gilad's words, is "a fundamental tool. This is a system of record that is foundational." It completes a four-stage lifecycle picture of registry - validation - monitoring - feedback.
I also pressed Gilad on whether Mercury might have another acquisition in the pipeline to bolster its SOA management capabilities, although this is another question I knew I wouldn't get an answer to. My speculation is that Mercury is considering it, and is not the only vendor in the software testing and application management space that's doing so.
I mentioned Systinet has a few more rivals than it appeared to in the first half of last year, when I compared it with Infravio and could only name a couple other minor players. But the advent of governance has changed the emphasis from registry technology to repository expertise, and that has brought in a number of new players from arenas such as asset management and even software testing. Most notable among them are WebLayers, LogicLibrary, Flashline and Mindreef. I spoke to the latter two companies late last year and have yet to write up their stories, both of which are very interesting. Now that it's become very topical I'll be doing that next week as part of a series of postings exploring the inter-related topics of SOA governance, registry, repository and lifecycle management.