Creating An Innovation Engine For Market Research
Innovation may be the hottest topic in market research these days. Conference agendas promise the latest and greatest, and clients continually ask their agency partners to show them “what’s new.” Almost all of the conversation is around “content”—what the innovation is. For the most part this was true at the Insight Innovation Exchange (IIeX) held in Atlanta in June, which I wrote about in my last Insight Landscape post . Amidst the buzz about new methods it can be difficult to separate fad from true innovation and true innovation from technology substitution (e.g., replacing a manual process with an automated one). And despite the creativity of some of the new solutions we heard about at IIeX, the market research industry lacks a true innovation engine like those found in many of our client industries (of which the Lowe’s Innovation Lab described by Kyle Nel at IIeX is a great example). I was invited to share some thoughts at IIeX about how we might drive more innovation in market research. What follows is a brief recap of my presentation.
Innovation in insight generation over the last twenty years has been driven occasionally by new theory (e.g., behavioral economics) but more often by new “instrumentation” (e.g., computerized eye tracking, fMRI and EEG, hierarchical Bayesian statistics) or by technology changes outside of market research (the Internet, social media, big data, and mobile devices). Although there is some R&D activity in the industry (often around statistical methods), in my experience most innovation in market research has been opportunistic, often starting with a one-off solution to a specific problem.
There have been a handful of game-changing innovations in MR since the advent of survey research in the first half of the twentieth century: the introduction of telephone surveys with representative samples; random digit dialing for telephone surveys; computer-assisted interviewing; and online data collection immediately come to mind. Conjoint analysis is, in retrospect, another game changer, especially when the impact of Bayesian statistical methods (that have come into MR largely through choice-based conjoint) is factored in. Finally, we’re in the midst of a wave of innovation driven by advances in behavioral science and neuroscience.
Each of these game-changing innovations reflects a different innovation trajectory. Comparing those trajectories may point the way to a more systematic approach to innovation in market research.
Consider the transitions from door-to-door surveys to random digit dialing (RDD) and from RDD to online interviewing. Both of these game changers resulted from technology changes outside of market research. In the case of online survey research we could go back as far as 1969, when ARPANET, a precursor of the Internet, was created. And, clearly, we would not be collecting data from respondents sitting at home using computers linked over a network today if some company had not introduced the “personal computer” as IBM did in launching the PC in 1981. The same transformation is happening now with surveys delivered through mobile devices. In each case, a new external technology has been adopted by market researchers as a substitute or replacement for the existing technology.
The invention of choice-based conjoint (CBC) analysis provided a solution for a particular problem faced by marketers and product managers and while some elements of CBC originated in transportation studies, the fundamental problem—understanding how people make choices when faced with multi-attributed alternatives—is the same as that faced by companies making decisions about which products to bring to market. Like many need- or problem-driven innovations, CBC emerged from a combination of theory, new instrumentation (the conjoint experiment) and “prior art”—earlier innovations that may have addressed different problems but that can be adapted or modified.
I offered two approaches—borrowed from outside market research—that could be a platform for innovation in market research. The first is known as TRIZ (an acronym for a difficult to pronounce Russian phrase that translates roughly as “theory of inventive problem-solving”). TRIZ consists of a set of principles for resolving technical contradictions. For example, the move to online survey research presented a technical contradiction with respect to obtaining representative samples of consumers because there was no natural sampling frame for the Internet. Two approaches to resolving this contradiction quickly emerged. One solution was intrinsic: build large panels with sufficient numbers of willing online respondents that would, with proper sample weighting, provide representative results. The other was extrinsic: use an offline sampling frame (RDD) to recruit a smaller representative sample of respondents for online surveys, equipping them with a way to access the Internet, if necessary.
The second approach to innovation that makes sense for market research is design thinking. Design thinking has emerged from the world of industrial design as a sort of all-purpose problem-solving method. Design thinking can be traced to the introduction of participatory design in the 1970’s. Back then, users “participated” in the design process by testing prototypes and providing feedback. Users did not really gain access to the design process itself until much later, as participatory design morphed into “user-centered” and, ultimately, “human-centered” design (with ISO guidelines on user-involvement in the design process). At its core, design thinking is a bottom-up approach to problem-solving, starting with the people who have most at stake—the “users.” For market research, design-thinking can drive innovation that improves both the client experience and the respondent’s experience.
Market research can only benefit by treating innovation as something to be managed actively rather than as a series of ad hoc and one-off responses to a disparate set of problems.
– David Bakken, PhD, Chief Insight Officer