Content Copyright © 2015 Bloor. All Rights Reserved.
This blog was originally posted under: The IM Blog
Rocket Software has a track record of acquiring no longer fashionable products that still have significant user bases or technologies. But it also develops new applications as well, and Research Director Philip Howard looks at how they all fit together.
Rocket Software is a well-kept secret. And for a company that is over 20 years old with annual revenues of more than $300m that’s some secret. From my perspective the company is interesting because it has made a practice of acquiring no longer fashionable products but which have significant user bases or technologies. For example, on the database side it owns the multi-valued UniVerse and Unidata (so-called U2) databases as well as D3. In addition, it owns Model 204 which, despite claims from other ill-informed database vendors, was the first to introduce bit-mapped indexes. Another relevant acquisition was CorVu, the enterprise performance management vendor. The company also acquired the Shadow mainframe products from Progress (previously Neon).
One other major point is that Rocket has a close relationship with IBM. It has over 100 products that it has developed and which are white labelled by IBM.
That’s just background. What is interesting today is that the company has just announced Rocket Discover. This is a full-blown combination of a self-service data preparation platform together with a business intelligence/analytics platform. While the product is in its first release on the business intelligence side of things it is intended to offer comparable capabilities to products like Tableau or Qlik but not to compete with them.
That last statement may seem odd. However, a core competency of Rocket Software is its data access capabilities. So, for example, Rocket Discover will work with the U2 and other multi-valued databases, which its more illustrious competitors do not. Similarly, Discover works with TM1, which they don’t, and it integrates with IBM RAVE (rapid application visualisation engine). Finally, Discover runs on Linux as well as Windows (as well as various cloud platforms including IBM Softlayer) and works with mainframe-based data such as VSAM (DB2 on the mainframe in a later release). Add this all up and what you get is a product that reaches parts of the data infrastructure that other potential competitors cannot address. This is how Rocket aims to be a competitor without competing: it will stick to its knitting.
Of course the other element is the built-in data preparation so that you’d have to combine Tableau with, say, Paxata, to get a genuine competitor.
As far as features are concerned, Discover has a scale-out architecture and supports a hybrid approach to data retrieval: in-memory (where data is compressed) or there is a live query capability with support for streaming. Simple reporting is included in the first release – no separate product is required – and more advanced features will be released later this year. One of the advantages of having a combined data preparation and BI platform is that you get data lineage right back to data sources and another is that the sharing and collaboration you get with data preparation also applies to the business intelligence environment. Finally, there is a RESTful API so that the product supports the blending and analysis of social media and other data as well as conventional in-house data.
This is a clever play. There are tens of thousands of companies (at least) that rely on a variety of technologies that are not currently fashionable and meeting their needs makes a lot of sense. In fact, it amazes me that more suppliers don’t take cognisance of this. I’ve run into the same issue in other markets as well, usually with the glib remark that “we haven’t seen any demand”, which is pretty obvious: why would potential clients talk to you if you don’t support their environment? Anyway, good business for Rocket.