Business analytics may be trending. The question is whether growth threatens to outstrip capacity.
Yesterday, Tableau released an interesting breakdown of the specific ways in which business analytics is trending right now. Of course, Tableau is a business analytics vendor, so any trend looks like good news to the company. But its outlook is persuasive.
It emphasizes the importance of "self-reliance" -- users being empowered to generate their own analytical queries and reports rather than relying on IT. It predicts that Hadoop will be big in 2013, which goes hand-in-hand with the growing importance of unstructured data.
"Predictive analytics" and "pervasive analytics" strike me as marketing smoke and mirrors (since when do enterprises use analytics exclusively to understand the past?), but on the whole this is a shrewd take on the directions in which BI is headed.
But there's a potential problem with this. At least, there is if you believe Gartner's latest research on the challenge of scale.
It's not called "big-data" for nothing. IBM estimates that we create 2.5 quintillion bytes of information every day (that's a billion billion, if you're counting). It's safe to assume that no enterprise wants or needs to analyze all that, but simply keeping up with the global conversations on platforms like Facebook and Twitter can stretch capacity -- especially as we're talking about unstructured data here, and not number-crunching.
This dovetails neatly with Tableau's prediction for Hadoop. With its developing reputation as the premium tool for organizing very large sets of unstructured data across multiple servers, Hadoop, and its various modules, Gartner's prediction is hardly surprising:
While IT organisations conduct trials over the next few years, especially with Hadoop-enabled database management system products and appliances, application providers will go one step further and embed purpose-built, Hadoop-based analysis functions within packaged applications. The trend is most noticeable so far with cloud-based packaged application offerings, and this will continue, it said.
Cloud is the other part of the puzzle, of course. Enterprises looking to scale BI in response to the apparently irreversible growth in volume of data are just about compelled to consider cloud-based options.
Hadoop is certainly a solution for distributed data storage within internal datacenters, but internal datacenters will surely be increasingly left behind, not only by the volume, but by the velocity of data requiring processing -- a trend on which the mobile explosion is also bound to have an impact.
The right storage for the right data loads -- which probably means cloud platforms; Hadoop, or comparable software frameworks, to organize the data; and employees with the confidence and understanding to query it. These are elements far-sighted businesses need to have in place, before we go from zettabytes to hellabytes.
— Kim Davis , Community Editor, Internet Evolution