By John Kearon and Tom Ewing
Something old, something new, something borrowed, something blue.
The ARF’s re:Think conference last week showed that our industry finally seems to have taken to heart that familiar refrain, “Market research needs to adapt or die”. Research companies old and new are reacting to changes in technology and the entry of new players into their market. They are adapting and adopting new technology.
That’s the good news. The bad news is that more often than not the new research technology is being used to peddle old research ideas. Just as the first television programs were little more than radio with pictures, breakthrough research tech is hampered by old models. Here are three examples of research innovation using technology on show in New York last week. See if you can spot what they have in common.
First up is Google Customer Surveys, who are revolutionizing the idea of sample sourcing in research by creating “surveywalls”, asking single questions of participants looking for content. Among their projects, they announced, is a 150-question tracking survey for an unknown client, broken into bite-size chunks for later reintegration.
Second we have Millward Brown, who have partnered with facial recognition firm Affectiva to try and measure real-time emotion in advertising. They deliver facial coding at scale, used as an adjunct diagnostic alongside Millward Brown’s unchanged, persuasion-based model of ad testing.
And finally there’s Locately, a start-up acquired last year by SMG, who specialize in location analytics – tracking GPS data and using it to serve micro-surveys via mobile. Locately uses this to ask about recall of particular promotions in-store..
So what do these have in common? They’re all using exciting new technology to service research ideas long past their sell-by date.
Is a 150-question survey really the best use of Google’s tools? The company themselves are scrupulously neutral about uses of their surveys. But even if technology means that a long survey is no longer boring individual customers, it still speaks to a lack of focus by researchers – an obsessive desire to collect data rather than a clear view on the issues that matter. Google customers would be far better off using simple but revealing questions like, ‘how do you feel about XYZ?’ or ‘to what extent would you recommend XYZ to a friend’?
Then there’s facial measurement. Study after study has shown how crucial emotional advertising is to business success – check out the most recent IPA work by Les Binet and Peter Field on the long-term benefits of emotional ads.. But if put in service of a persuasion-based model of advertising centered on metrics Binet and Field show do not correlate with the biggest business effects, facial coding risks being merely decorative. Measuring emotion should be at the heart of any modern predictive advertising testing tool, not merely an adjacent diagnostic.
And finally, using location analytics to serve up prompted recall questions relies on our famously unreliable memories. Much better surely to use the amazing potential of GPS-triggered surveys to capture action and emotion, not just recall.
The marriage of new tech and old techniques is an emotionally tempting solution – enough innovation to feel like change is happening, wrapped around a fundamentally conservative core.
So the risk is that we see a repeat of the shift from telephone to online surveys in the early 00s, where the main concern was to adapt the old methods as precisely as possible. It led to a “lost decade” of overlong surveys, falling response rates and a loss of faith in research data. It’s only really been in the last few years that the industry has begun to explore the potential of online-only methods – there’s no technical reason that MROCs couldn’t have reached their 2013 position by 2003, for instance.
What I realized at the ARF is that we’re at a similar inflection point now. The research industry is adapting to new technology – it has no choice. The future involves faster research, more DIY research, more passively collected data, and new entrants with new money and new business models. These are good things. But there is a real risk that this new world will bake in some of the failures of the old – like its faith in self-reported data and the power of rational messaging.
Notable by its absence at ReThink 13 was any mention of behavioral science and the new understanding of how people really make decisions. Technology and psychology must go hand in hand as research evolves. If they don’t, the question ought not to be “Will research adapt to technology?” but “Can researchers be trusted with it?”
Perhaps it’s inevitable the first uses of new technology are in the service of existing approaches, just as early television and the first internet surveys were. Let’s hope the generative cycle of experiment-test-fail-and-learn works quickly to turn the likelihood of a quickie-divorce into a sustained love affair with the business advantage these new research technologies can bring to our clients.