Surprise! We have a new President…and it’s not who many thought would win when we all headed to the cast our votes yesterday morning. With all the money invested by political candidates and major news outlets in political polling to predict the outcome, they didn’t even come close. Even Frank Luntz, Republican pollster tweeted:
Yet, he was wrong…they were all wrong. The volume of interviews, the extensive modeling, the vast number of statistical gurus did little to improve a process everyone has historically found flawed. And, the pundits were forced to retract their predictions overnight:
What bothers me most about this isn’t that they were wrong, but that many see this as a failure of the market research process. Yet, polling doesn’t even employ the best elements of our craft. Relying heavily on “big data,” pollsters use elaborate modeling schemes based on assumptions of the voting population. These assumptions are the basis of the error, one wrong assumption and the entire model is thrown. Obviously, some incorrect assumptions were being made in the months leading up to this election.
This “big data” approach to predicting outcomes has also found its way into consumer research where behavioral data is being used to predict consumer sentiment. The risk here is no different than in political polling. What isn’t the data telling us and what holes are being filled by “assumptions?”
In the current research landscape, I am starting to feel a bit “old school” as I preach the benefit of not totally abandoning the basics of our craft. Yes, there is definitely a science to what we do, but there is also a bit of finesse in uncovering meaningful insights. The failure of the political polling process only makes me more resolute in justifying my belief in recommending a solution that not only uncovers the “how many” but the “why” behind why consumers do what they do.
VP of Research & Insights
Capitalizing on her broad research experience over the past 20 years, Kathy demonstrates an in-depth understanding of industry trends and consumer needs. With an emphasis on how consumers use media, she has conducted a wide variety of projects including advertising messaging, programming evaluation, brand positioning, market segmentation, customer satisfaction, pricing and packaging, and new product development.