As Big as Big Data Can Get

As an analyst, I’m often asked to make predictions as if I have a crystal ball on my desk or know some special magic. Magic seems to feature heavily in the analyst part of the IT ecosystem. But that’s another story… I do feel confident in making this prediction though: That the big news from IBM Insight will be the partnership with Twitter. Big announcements like that tend to catch the attention of journalists, pundits, and other opinionated people. It is a big deal but only as part of an even bigger deal. Think of the IBM-Twitter announcement as the natural progression of what IBM has been up to all along: Building one of the most comprehensive big data and analytics platforms, both on-premises and in the cloud.

At the highest level, IBM can bring to bear traditional business intelligence tools, big data infrastructure of both the Hadoop and in-memory database varieties, hardcore statistical analytical tools, user friendly analytics, stream data capabilities, data integration and governance, and cutting edge cognitive computing. Try saying all that without taking a breath. They have tried and true products with extensive expert followers such as Cognos and SPSS, and newer edgier tools such as Watson Foundation. Some products are clearly for geeks such as BlueMix and SPSS while others are for the rest of us especially Watson Analytics and InfoSphere BigSheets.

In other words, IBM has everything you could want in your big data and analytics toolbox. Sometimes it’s hard to see the breadth of their product since they are spread around a bunch of different product lines. If you think about InfoSphere BigSheets and the other products in that line – mostly information governance and Hadoop products – you have to wonder why more customers aren’t confused. I almost wish they would gather all the big data products up under the Watson moniker.

What’s the value of such a broad offering? Can’t you buy a bunch of point products and assemble your own big data version of Marvel’s Avengers? Sure but, like the real Avengers, getting them to work together isn’t always easy. There’s always a Hulk in the group that doesn’t want to play nice with the others. For companies that plan to lean heavily on comprehensive analytics, trying to get all the parts to work together might keep them from achieving their project goals. Pre-integration is especially important in hybrid, cloud/on-premises environments. Getting cloud and on-premises systems to work together is tough enough. Adding multiple vendors into the mix makes that a much more difficult task.

If big data needs are simple, for example only crunching social media data to get a sentiment score, then by all means go ahead and choose from the dozens of big data-analytics companies out there. If, on the other hand, the plan is for pervasive big data and analytics, then it’s worth considering a broad, comprehensive solution set and IBM has one worth looking at.