CMB’s Empact Model of Emotions

Emotions have a place in customer experience management

It’s about the emotions, or is it? On June 11, 2015 Erica Carranza, PhD and Brant Cruz, both of Chadwick, Martin, Bailey, highlighted their new model for measuring emotions known as Empact. They led off with a fundamental marketing truth – we act based on how we feel. Our choice of brands is largely dependent based on the expectation they will deliver positive emotional payoffs.

The example Dr. Carranza led off with is one many of us are faced with – where to go grocery shopping? She brought out her two primary options Trader Joe’s versus Whole Foods. If you are familiar with these chains you know they have an overlapping appeal to many people. Trader Joe’s prides itself on providing good quality food with numerous organic options. They also employ a general sense of lighthearted humor in their approach to business. In the other corner is Whole Foods. They are a leader in providing options for those seeking organic and for other dietary needs such as gluten-sensitivity.

The emotional battle expressed here is a familiar one – happy with Trader Joe’s especially because of their mix of quality at a reasonable price versus a sense of security that Whole Foods would not allow additives into the products they sell but at a substantially higher price. Multiple emotions, often in disagreement with each other, can easily occur during a shopping experience.

We are used to qualitative research providing the emotional recap. CMB wanted to expand on that via Empact which is CMB’s platform for quantitatively measuring the effects of emotion in relationship to a brand. As can be seen below emotions are only one part of the equation, but heretofore have been a more difficult piece of the puzzle to measure effectively. Knowing what consumers think is one thing, but knowing how the brand makes them feel takes it to the next level.


According to Dr. Carranza during her days on the client side she received numerous pitches from agencies that touted their ability to measure emotions without having to ask the respondent. She likened this approach to Superman’s x-ray vision. Why ask about emotions when you can simply peer into the depths of their psyche?

Biometric methods, such as skin conductants or EEG, can be leveraged to measure valence (level of good or bad) without having to ask someone, but they cannot measure specific feelings, such as security, pride or sense of connection. Facial coding, as a technique, allows the researcher to capture the seven basic emotions. Of the seven only happiness is one most marketers would want associated with their brand.

Dr. Carranza states there is very little consensus in the scientific community as to the best methods for measuring unconscious phenomena such as emotions. Meaning the biometric approaches do not perform significantly better than self-report approaches. Behavioral economics makes a clear argument that people do not always know why they do something, but more often than not they can tell you how they feel about their actions or the scenario they just went through. Asking them directly is a valid way of assessing those feelings. The exception are circumstances where you believe the respondent is trying to hide their true feelings, such as extremely sensitive topics – in this case facial coding may be the better method.

The Empact method strives to be practical (it is survey-based), comprehensive in the range of emotions covered (especially positive emotions), robust enough to handle large samples, and systematic in a way that allows for brands to be compared over time. The model shines a light on the importance of emotions in the consumer behavior process. In doing so it is useful for brand health tracking, targeting and segmentation, customer experience and NPS studies, and message development. The model was internally funded and derived from 3,000 US adult respondents, weighted to match age and gender.

The model includes the seven basic emotions, social/self-conscious emotions (e.g. pride, embarrassment, etc.) along with general mood states (such as feeling good or bad) to define its affect measures. Underlying all emotion are two core states – valence (good vs. bad) and activation (tired to energetic). These are also captured. Mental states encompassed in the model include: social belonging; cognitive ease; and flow.

The model identified 15 key emotion factors that leverage 38 discrete emotions drawn from an initial pool of 54 discrete emotions and mood states. The end result of the model is a Net Positive Emotion Score which takes into account total average positive and negative emotion facto activation. In testing the model CMB found Empact alone worked nearly as well as traditional measures of functional brand attributes in predicting key outcomes. The results were improved even more when analyzing specific segments such as gender or brand preference.image2

The case study presented focused on jetBlue, which as seen above performed more effectively on the positive emotions (in green) than it did on the negative (red). This can be applied against your brand and key competitors in your space.

Greg Timpany directs the research efforts for Global Knowledge in Cary, North Carolina, and runs Anova Market Research. You can follow him on Twitter @DataDudeGreg.


Speak Your Mind