December 11, 2012 – The realities of production and contraction will kill off some big data startups next year, bringing down with them some of the market hype, according to a new industry forecast. But analytics as a whole will build on better big data use cases as well as refined use of cross-channel development and machine learning.

Tom Davenport, International Institute for Analytics Director of Research and renowned author, led a group presentation and discussion Tuesday of the IIA’s eight “2013 Predictions for Analytics.” Analytics was propelled in many ways by interest in big data throughout the current year, though Davenport expressed uncertainty over how the two practice areas will play out in the year to come.

“Big data has certainly been an enormous factor changing the way people view analytics” over the last 12 months, Davenport said, later adding, “It is going to be puzzling a lot of people, how big data and analytics relate to each other. In a lot of organizations, we had parallel silos in 2012, and I hope that’s not the case in 2013.”

On its own, big data faces a major “chasm” in the year ahead. IIA faculty member Ravi Kalakota said there are about 316 big data startups since 2011, and he estimated that more than half of those will be acquired, shuttered or merged with other operations by 2013. The hype cycle for this analytic sector will “meet organizational reality,” running out of customers or funding, stalling out as they migrate from pilot projects to production and losing ground on their more established competition, Kalakota said. Still, based on interest and ongoing projects, IIA expects more use cases and newfound capabilities for big data analytics at the enterprise level in 2013.

Hot on IIA’s list of expectations for 2013 was the expansion of specialized use of analytics. 2013 will be marked by new crossover relationships based on competitive importance of newly aggregated data, according to IIA. For instance, a retailer will approve credit cards to shoppers based on their buying information as well as their credit score, which has been the primary basis for bank approval of those same cards. There will also be more personalization of product-driven analytics, especially in customer-facing industries. IIA faculty points to the connection between a customer’s Wi-Fi wired smart device getting notices on deals in the aisle of a store they are in at the moment. On the enterprise side, IIA noted that the challenge with these cross-channel analytics capabilities comes in closing the decision management loop. Thirdly, more companies will be able to pick out applications of machine learning that show business returns. IIA cited IBM’s work with Watson in the health care industry as an early example of this.

Two other predictions for the coming year touched on that in-demand but hard-to-define role of data scientist. Davenport expects a continued shortage of data scientists to the point that businesses compensate by pooling deeper data projects into small analytic teams. Also, the “mystique” of the data scientist position – with promises of high pay and independent work – will persist even as the lines between its functions and those of other analytics professionals start to blur.

Other noted expectations related to analytics in 2013 by IIA included the growing but still not commonplace use of new visualization tools and escalating numbers of high-profile data breaches that prompt more sophisticated analytic approaches. 

IIA also recapped its predictions last year for 2012. Davenport said areas where predictions missed the mark for the current year – more analytics for corporate performance management and expectations of more attention to privacy – could largely be attributed to forecasts made outside of the “mainstream” flow of business analytics. However, Davenport said that predictions surrounding big data began to unfold and touch on various aspects of emerging analytics and data management during this year.