The Automation of Behavioral Science and the Future of the Market Research

In 2017 we saw the burgeoning trend of automation in market research hit its stride. Enterprises embraced automated solutions that allowed them to gain insights in days (e.g. Zappistore). Small to mid-sized research agencies moved mounds of sample sourcing to automated solutions (e.g. PureSpectrum, Cint, Lucid) in pursuit of greater control, more efficiency and lower CPIs. And the big Market Research firms started to follow suit by providing “cut-down” versions of their flagship products for lower prices and two-day turnarounds.

These advances were driven by pressure on insights divisions to deliver actionable intelligence at a speed that could actually impact business decisions. The industry was not blind to these demands and thus was focused on investing in the technology that could accelerate automation. These trends are encouraging, and the acceleration of delivery is essential for our industry.
However, in 2018, as investment continues to pour into technology that facilitates automation, there is an important point of caution to be heeded:
Speeding up the delivery of inaccurate insights will only put us on a fast track to irrelevance.
Yes, we need automation, but we don’t need automated delivery of information that doesn’t actually predict behavior. We still need to take a long view of how the industry can remain viable in a world of real-time analytics of behavioral data. In 2017, we saw an increase in budgets being stripped from traditional market research and moved into the behavioral analytics space. For many, this trend is alarming and raises questions about the usefulness of traditional MR in a world overflowing with data on consumer behaviors.
However, as useful as behavioral data is on its own, it has two primary weaknesses: it is always looking backward, and it lacks fundamental explanatory power on the “whys” behind behavior.
Fortunately, behavioral science based market research offers the the solution to both of those problems: robust models of behavior that predict future behavior, and deep insights on the whys behind that behavior. These real advantages are why the industry has seen increased use of advanced measurement techniques with the simultaneous decline in reliance on traditional measures. So much so that the phrase “claimed response” has become a pejorative in our industry.
Given the confluence of these two demands, we see 2018 as a year where investment increases in the automation of behavioral science.
It’s funny to write that phases. I used to write the phrase “quantifying emotion” and people thought it was an oxymoron.
Now, as I write the “Automation of Behavioral Science”, I anticipate the same initial reaction: “those two things are the antithesis of each other” and “you can’t do that”. But they’re not, and we already are.
In 2018, behavioral science based research firms will increasingly focus on automating processes, data analysis, experimental design and the automatic visualization of insights through dashboards and norms. As the industry embraces the reality of sound science delivered in days, we will see not only a resurgence of budget investment in insights divisions, but a reliance on our insights for critical business decisions.
Automating the delivery of behavioral science based insights will put us on the track of long-term success.