Through a special arrangement, presented here for discussion is an excerpt from a current article from the Joel Rubinson on Marketing Research Consulting blog.
With people increasingly living their lives online, an enormous amount of information about what people do, what they think, and what they say, is already "out there," replacing much of the traditional need for surveys, according to Larry Friedman, the recently-retired chief research officer for TNS in North America.
"We've seen accelerating interest and experimentation among forward-looking researchers in making use of data in the social and digital spheres to answer marketing questions," said Mr. Friedman in an interview with Joel Rubinson on Marketing Research Consulting's blog.
The big benefits are not so much expense and time, but taking market research beyond the world of "insights" and bring it into the world of "action."
"Tracking, for instance," said Mr. Friedman, "can be transformed from telling us what happened last quarter to telling us what is going to happen next quarter, and what specific actions the business should take given those predictions."
Moreover, while survey-based segmentation can be insightful about defining targets for a brand, in many cases it is difficult to buy media to specifically reach those attitudinal targets.
"TNS has pioneered ways to make use of lookalike modeling based on the digital behavior of those attitudinal targets, and then partner with digital agencies to focus on delivering ads to millions of people who are like those survey-based attitudinal targets," said Mr. Friedman. "This has enormously increased the efficiency and effectiveness of digital campaigns where it has been used, with real business results."
One of the challenges is the questionable correlations that can result when dealing with large digital/social data sets with thousands of variables. Said Mr. Friedman, "Market researchers will need to approach these large data sets with real hypothesis testing in mind, more like social scientists than I think many are used to. In other words, theory is going to have to start catching up to quantitative capabilities."
Yet success using social media discussion data to predict business results is already being found in some academic work from Wendy Moe at the University of Maryland and David A. Schweidel of Emory University as well as the commercial work led by TNS.
"Based on the progress to date, I have hopes that in five years we get to a place where the industry mindset is no longer 'What questions should I ask in my survey?' but 'What is the right type of data and where do I get it?,'" said Mr. Friedman. "And yes, sometimes surveys will be part of the solution, but they will be shorter, more focused, and integrated with social and digital data."