Clarity begins to emerge

Over the last few years we have seen the emergence of “big data”, the increasing adoption of the Internet of Things (IoT), growth in NoSQL platforms such as Hadoop, a similar increase in the number of providers of business intelligence, and the emergence of new markets for tools such as data preparation, data cataloguing and data lake management. This has left the whole market in a state of uncertainty. There have been so many trees that it has been difficult to see the wood.

However, I think that is now changing. There are some clear directions emerging within the market across the whole of what we at Bloor Research call the Data Layer. I won’t say that NoSQL or Cloud or data preparation or any of these other technologies are mature. Not in the way that relational databases or ETL (extract, transform and load) are mature, but I will say that they are leaving adolescence. You can see the potential for maturity and, more importantly, you can begin to see which technologies and, especially, which vendors, are likely to reach maturity, and which are not.

And this clarity is starting to drive “consolidation”. Just to be clear – because we have hadn’t had a significant wave of consolidation in the Data Layer for some time – I should explain what I mean by consolidation. I mean companies going out of business or being acquired or divesting themselves of parts of their portfolio, product development being stopped, companies re-focusing their products on niche markets rather than aiming to be broadly competitive and, in this incarnation of consolidation, public companies going private.

Let me give you some examples. In the graph database market there has been a significant shake-out over the last 18 months: Aurelius has been acquired, SPARQL City is out of business, Titan looks to be dying (there is virtually nobody contributing to its code base), InfiniteGraph is no longer being developed and neither is FlockDB, and there is a new graph niche emerging for graph databases in the data virtualisation space to compete with the likes of Denodo (which is by no means a bad idea). OK, that’s specific to graph databases but I cite this example simply because I happen to be researching this area at present: I can provide plenty of other examples. For instance, Harte-Hanks is looking to offload Trillium Software, word is that Actian is re-focusing on its Ingres cash cow at the expense of Vector and Matrix. Again, Dell has offloaded its software business (excluding Boomi) and also gone private, as has Informatica.

I am not saying that these are bad things, just that they are indications of consolidation, and I expect to see a lot more mergers, acquisitions and “gone out of business” notices. The most recent example is that HPE (Hewlett Packard Enterprise) is selling off its software division too. So the company most famous for keeping COBOL alive is the one acquiring the company where software (mostly) goes to die. I find that delightfully ironic.