IBM acquires Netezza

Philip Howard

Written By:
Published: 22nd September, 2010
Content Copyright © 2010 Bloor. All Rights Reserved.

My immediate reaction when I hear any important news is to immediately write a relevant article. In this case, circumstances meant that I have been unable to. In this case, I am glad that I didn’t, firstly because I have had time to reflect on the implications of this acquisition and, secondly, in light of Bob Evans’ piece in Information Week (thanks for the reference Bob) suggesting that Oracle may yet come in with a higher bid. This article is therefore written on the assumption that that doesn’t happen.

My immediate reaction on hearing the news was that IBM has negotiated a much better deal than Microsoft did when it bought DATAllegro. $1.78bn compared to $275m but around 100 times as many customers and a system that runs on IBM platforms (x series) as opposed to one that didn’t run on either Windows or SQL Server. Anyway, that’s an aside.

Most commentators will write about how Netezza TwinFin fits into IBM’s workload optimised systems strategy but I want to drill a little deeper and discuss the synergies, opportunities and issues that this acquisition raises, especially those that may be less obvious (i.e. I am not going to discuss synergies between warehousing and data integration or data quality, these being obvious).

The biggest issue will be strategic: positioning Netezza versus the IBM Smart Analytics System. This shouldn’t be too difficult. I imagine that the latter will be positioned as the company’s general-purpose enterprise data warehouse and the former for specific analytic functions, data marts and edge environments.

The other two issues will arise around Netezza’s lesser known capabilities: Netezza Mantra on the one hand, and its data mining/advanced analytics arm that it gained when it acquired NuTech technologies on the other. The expertise in the latter group will no doubt be used to augment what IBM already has in this area, with respect both to its own capabilities and those of SPSS, so I don’t see this as much of a problem either.

However, Mantra may be more tricky as it is a direct competitor to Guardium, which IBM acquired earlier this year. While Guardium is a much bigger beast within the database activity monitoring (DAM) world, Netezza has a rather different spin on this particular market: notably that simply monitoring activity is one thing but that what you really want to do is to be able to perform predictive analytics against monitored data so that you can recognise patterns of potential threatening behaviour. This, of course, requires storing historic data and then performing analytics against that information, which is precisely why Mantra runs on the Netezza platform. The market has not yet really woken up to this possibility yet but I believe that it should do so, and will do so in due course. Ultimately, what this may mean is that the front capabilities of Guardium and Mantra may get merged with both running on TwinFin at the back-end.

Which brings me onto the first opportunity that IBM/Netezza brings, which is with respect to Tivoli SIEM (security information and event management). I have been arguing for some time that SIEM products are largely built on antiquated architectures: they are processing large volumes of events but don’t generally use event processing to recognise patterns of events in real-time, and they typically store historic data for forensic purposes in file systems or merchant databases, which are not best suited to doing serious fraud analysis. Again, this is partly because relatively few companies are doing that sort of analysis but, similarly, I expect it to become a growing requirement. So, there is an opportunity for Tivoli to embed Netezza as its database (as opposed to the file system it currently uses) at the back-end and, as an aside to the subject of this article, either InfoSphere Streams or WebSphere Business Events (previously AptSoft) at the front-end.

There is also direct synergy between Netezza and these event processing products in capital markets (especially hedge funds, which like to be able to combine analysis of historic data with current stock tick information) and other environments, including the government sector.

The one other significant advantage that Netezza brings to the table is Netezza Spatial. Historically, this market has been dominated by Oracle, but Netezza offers a real competitor to Oracle in this area and, as this is an area that IBM has never been particularly strong in, this adds significantly to its portfolio. Which brings me onto Oracle itself.

In an article such as this I usually comment on winners and losers but in the case of the latter I will confine myself to Oracle. I see this acquisition as a big threat to Oracle, which is why Bob Evans’ suggestion of a competitive bid is plausible. Assuming the IBM acquisition goes ahead, this is going to be very difficult to compete with: Netezza already takes business away from Oracle and with IBM’s branding behind it can take significantly more. Note that, just as there is an almost fully automated facility for migrating from Oracle to DB2, there is also a similar capability offered by Netezza. The only possible fly in the ointment will be if the Netezza sales team concentrate too much on selling into the IBM user base as opposed to competitive and Greenfield implementations: IBM will need to get its incentive structure right.

Finally, I should comment on Netezza’s partners. Most of these should not be affected by this and, indeed, many can expect extended sales opportunities. However, Composite Software is a special case. Composite Software has a solution for data virtualisation (query federation) that has been specifically developed for Netezza appliances (it takes advantage of specific Netezza features). I rate this company as the leading vendor in its field and while IBM has its own technology in this area it has languished somewhat compared to Composite over the last few years. If I was Composite I would extend its Netezza specific offering to cover DB2 (which it already supports) and I think it could do very well out of this acquisition and it offers a real opportunity for the company. Indeed, I can see it becoming an IBM acquisition target in its own right.

Of course, all of this discussion will be blown out of the water if Bob Evans is right. Do I think he is? I have no idea: who can fathom the mind of a man who wants to race the America’s Cup using wing-sailed catamarans? Assuming Bob is wrong there is a lot going for this acquisition and that’s precisely why he might just be right.

Post a comment?

We welcome constructive criticism on all of our published content. Your name will be published against this comment after it has been moderated. We reserve the right to contact you by email if needed.

If you don't want to see the security question, please register and login.