What was I saying about another week, another deal? Well, it’s another week, so here’s another one – IBM snaps up Ascential software for around a billion dollars. Ascential, you may remember, was also the origin of IBM’s Informix technologies.
Ascential essentially adds ETL (Extract, Transform and Load) capabilities to the IBM data management portfolio, and as others have noted it will be interesting to see what the partner fallout is; if I were Informatica, for example, I’d be concerned. Anyway, on to the Q&A:
Q: Before we get to Ascential, why add this functionality in the first place?
A: In many respects, I see this deal as similar to the acquisition of Aptrix by IBM for web content management in ’03. Web content management and ETL are different disciplines, of course, but both were treated for a long time as partner plays rather than IBM opportunities. In both cases, however, it became obvious that delivering the technologies separately was becoming more and more difficult. In the case of ETL, it’s an obvious need in engagements ranging from traditional data warehousing to more cutting edge information integration (and just for the record, I’m not fond of the EII term). That, combined with the fact that competitors from Microsoft to Oracle are adding similar functionality, made such an acquisition logical in hindsight.
Think of it this way: the growing reality of many enterprises today is that they have data – some of it critical – spread all over the enterprise, in a variety of formats, repositories and systems. Ideally, you’d like to have some centralized control over that – i.e. in a datawarehouse – but the scope of the problem is simply too broad. Hence the need for the information integration technologies. Crucial to either approach, however, is the ability to perform some normalization on the data so that it’s actually usable. And that’s where Ascential comes in.
Q: Ok, now why Ascential specifically?
A: Well, there’s of course the prior partner relationship – which seems to be crucial to any IBM acquisition these days – but obviously there were other partners to pick from. Cost of acquisition was likely one factor, but of course most of IBM’s acquisitions come down to technology rather than traditional concerns like marketshare or installed base. Ascential also has some very good people to pick from, but all in all I’d say it’s a multi-facted decision rather than anything clear-cut.
Q: What does this mean for other IBM partners?
A: That’s the wildcard to this deal, in my opinion. Informatica and its brethren are important partners to IBM, and this certainly isn’t going to make them blissfully happy. But just as its acquisition of Aptrix didn’t kill relationships – at least from a services perspective – with partners like Interwoven and Vignette, I don’t see many if any partners walking away because of this. Certainly they’ll have to step back and reevaluate their available options – one of them being acquisition – but in the end I expect the tactical implications of this deal to be relatively minor. The integration work here is substantial, so it’s not as if competitors will be seeing any combined Ascential/IBM product in their accounts tomorrow, or next week, or next month.
Q: What does this acquisition say about the importance of EII to IBM?
A: Well, actually I think it was the outbound messaging and dialogue on this morning’s analyst call that says the most on the topic: EII’s clearly the fair haired child in the DB2 family at the moment. With good growth behind it, and the aformentioned continuing enterprise difficulties with EII type problems, IBM’s clearly committed to attacking the problem with everything it has to bear – including its checkbook. On the call this morning, EII was mentioned seemingly every 30 seconds, and it’s not really a surprise.
Q: Ok, any closing thoughts?
A: Just two. First, this isn’t the last acquisition we’ll see the DB2 group make – not by a long shot. Second, watch for potential synergies between this acquisition and IBM’s rather unheralded SRD acquisition. Ascential exists to centralize and normalize data from disparate sources; SRD existed to identify relationships in that type of data. Food for thought.
