Putting Emotion into the Equation

At The Market Research Event in Orlando, Aaron Reid and Stephen Springfield of Sentient Decision Science discussed the need for integrating affective measurement into market research to complement and strengthen calculative measurement. “We all know that reason alone can neither sufficiently explain nor predict behavior,” Stephen said. “We used to be told to ‘give a rational message wrapped in an emotional envelope,’ but that’s not enough.”

Sentient tested women’s reactions to diamond rings, contrasting galvanic skin response to a .83 caret diamond ring vs. response to a 1.05 caret ring. While the second diamond is only 27% larger, it generates 175% more emotional arousal. This disproportionate response is because the larger ring activates associations such as expense, status, appreciation, and commitment. The 1-caret diamond crosses an irrational threshold.

“We think very rationally in the market research industry,” Stephen said, “but the fundamental premise of marketing is to generate emotional reactions. It’s why marketing and design exist.” It’s often why consumers pay more for a product than it costs to make.

The rational mind does not have access to all the information that informs decisions, yet it feels compelled to explain 100% of what we do. Traditional surveys uncover rationalizations, not reasons. “We cannot reliably evaluate, anticipate, nor communicate the impact of emotions on our behaviors,” said Stephen. “Thus much of the gap between research and results exists because most firms don’t measure the one thing marketers are striving to deliver: emotion. We need to measure micro-emotions to understand decision making.”

Aaron explained how the firm can quantify emotional response and add it to the rational response. Galvanic skin response won’t work through web browsers and smartphones. However, there are other ways to measure implicit associations. “Sentient Prime is a gamified web-based app that quantifies the strength of implicit associations by measuring changes in mental processing ability to determine what happens in your mind without passing through your conscious filter.”

Stimuli, called “primes”, are shown for 500 milliseconds, while you are asked to subsequently sort emoticons into positive and negative emotions. When the emotion contradicts how you feel about the brand, product or packaging you were just shown, you experience cognitive dissonance and it takes you longer to respond and you make more errors. On mobile devices, Sentient also calculates the velocity of the swipe as a measure of aversion. This assessment quantifies the emotional push and pull of the stimuli. You can demo the approach for yourself at http://bit.ly/implicitemotion.

Aaron said, “The title of our talk is ‘Putting Emotion into the Equation’ because we need both emotion and reason to sufficiently explain and predict behavior. We are each a combination of System 1 and System 2 processing: we have emotional responses that we reflect on. As a result, the best models combine System 1 and System 2 data.” (See Aaron’s paper, “Emotion as a tradeable quality“.) “Since emotion Is a weight on our decision making, let’s put it as a weight on our conjoint models.”

For instance, Sentient found that the rational measure predicted 41% of the shared variance of product sales at Walmart the quarter after the survey was conducted, while the affective measure predicted 51%. Combining the two, however, predicted 80% of actual sales. “When emotional appeal is added to traditional choice methods, the results provide a more accurate reflection of how consumers will behave.”

In another case study, for Macy’s, Sentient found that “combining reason and emotion predicts the sell-through of seasonal inventory with a high degree of accuracy.” In fact, buyers’ stocking decisions predicted only 28% of actual sales, while choice-based conjoint predicted 69%, and the emotional and reason model predicted 94%. They replicated this work at Macy’s four times, in every case with a variance of 81% or better. They’ve conducted extensive research on research to replicate the work in many other categories and circumstances.

“Human behavior is no longer a puzzle, once we know how the pieces fit together,” Stephen concludes. “Universally when we add emotion to the category the correlation between test results and sales increases. Emotion reveals hidden truths. It’s a new age of emotional measurement.”

Jeffrey Henning, PRC volunteers as the president-elect of the MRII. He is explicitly associated with Researchscape International.


One thought on “Putting Emotion into the Equation

Leave a Reply

Your email address will not be published. Required fields are marked *