A happy man wearing a sad face mask says, "No!" I am not happy with that price!"

When it comes to marketers measuring emotions, in the last ten years there have been many false dawns. The origin of measuring emotions partly stems from Charles Darwin in his 1872 work, The Expressions of Emotions in Man and Animals. Darwin hypothesized that some basic emotions were universal regardless of ethnicity or culture. Using facial expressions, American psychologist Paul Ekman’s 1960s work was also heavily relied on to identify seven basic emotions. These were happiness, surprise, fear, anger, disgust, sadness and neutral.

Racing forward to today, marketers have reasoned that these primary emotions are linked to behaviors and eliciting these emotions will lead to market success. The most common use of emotions measurement today, facial coding, still largely relies on Ekman’s 1960s taxonomy. Forethought applies the Laros and Steenkamp (2005) emotions taxonomy comprising surprise, happiness, love, pride, contentment, anger, sadness, anxiety and shame.

Those interested in the literature on emotion will find many taxonomies of emotions. As marketers, we seek to understand just these primary emotions. Primary emotions can be likened to primary colors used in printing. It is unusual to see pure cyan and similarly, it is unusual to see the primary emotion, for example, pure anger. We would expect to see anger accompanied by sadness and surprise, perhaps even love. Just as the four primary printing colors result in hundreds of millions of variations of color, there are almost endless combinations of primary emotions.

Even though a primary emotion does not occur in isolation, as marketers we seek to identify the emotion most prevalent in the behavior that drives choice. For example, love is the dominant primary emotion for first time moms when purchasing diapers (as moms become more experienced the importance of emotion gives way to rational attributes in particular, price). To facilitate affective choice, brands seek to create neural pathways between the brand and the emotion driving choice. Kimberly Clark Australia implicitly cues the primary emotion of love by using imagery and the soundtrack, It Must Be Love.


The false starts

The application of emotion in marketing initially was focused on physiological measures. That is, the mere presence of an emotional response as measured by skin conductance (galvanic skin response), pupil dilation, or a quickening pulse. Next in the evolution was the direct recording of neural activity using techniques such as fMRI or EEG to measure the intensity and valence of emotional activity. The very considerable shortcomings of these approaches were, first and foremost, that they did not measure primary emotions. But rather, they were limited to just capturing a general emotional tone categorised as ‘approach versus avoidance’. Without being able to identify primary emotions, fMRI or EEG were not particularly useful in helping advertisers. Secondly, so crude was the grasp of consumer behavior that back then, there was an underlying assumption that negative valence was bad and positive valence was good. We now understand that in some instances, negative emotions can be more effective in driving behavioral change than positive emotions.

At present, the most popular approach amongst marketers for measuring emotion is facial coding. Indeed, more than one-third of global Fortune 500 companies have used facial coding for ad pre-testing. Whilst facial coding does measure Ekman’s primary emotions, it cannot measure pride. Emotion is not constrained to just those captured by facial expressions. Social emotions such as pride and shame have not been linked to distinct facial expressions and therefore are not captured by facial coding. The trouble is, at Forethought we’ve found when using implicit measurement that pride is the most common emotion driving consumer behavior.


"% Studies where Emotion is in the Top 3 Drivers" bar chart comparing emotions

The most common primary emotion explaining category choice.


If you’re happy and you know it

More troubling is the assumption that by using facial coding, emotions such as happiness may be measured through the presence of a smile. A joke makes you smile but does it make you happy? The outcomes for Beth Harmon in the final episode of The Queen’s Gambit made me happy to witness Beth’s triumph over addiction, but did that feeling activate my zygomaticus major (the smile muscles)?

Positive psychologist, Martin Seligman, describes happiness as a combination of joy and deeper feelings of meaning and purpose. That is a eudaimonic form of happiness. I would hypothesise that eudaimonic happiness is a stronger motivator than momentary laughter. The challenge for facial coding is that it only measures changes to facial muscles. When it comes to emotions there is no hard-stop. Emotions can linger into feelings and be present long after the initial emotional trigger. So even if I did only momentarily smile, the happiness may have been enduring and unless I continue smiling, will go undetected by facial coding.

Compared to using implicit measures such as Prophecy Feelings, Forethought has found that the level of happiness is markedly understated when measured by facial coding. In 2020, Forethought replicated a facial coding study undertaken by one of the major Hollywood studios. Tests were conducted with two genres: comedy and drama and for both genres, facial coding markedly understated the level of emotion elicited.

For the comedy genre, Forethought found that whilst there was a higher level of happiness elicited from comedy, the happiness from comedy relative to drama was more shallow and not as strong in driving behavioral change as the happiness elicited from the drama. It is simply nonsense to assume that people are only happy when they are smiling and until scientists can link specific neurological activity with primary emotions, the most predictive way of measuring emotion is by applying implicit measurement.

"Facial Coding Understates Emotional Elicitation" bar chart comparing facial coding and implicit measurement of the comedy genre and drama genre

Comparison of happiness elicited in comedy and drama using alternative measures.


Moment by moment assessment in pre-testing advertising is used as a diagnostic tool to identify the peaks and the troughs of emotional elicitation and whether a peak-end to the marketing communication is achieved. Creatives derive utility from the moment by moment assessment of physiological and facial coding measures. This practice is analogous to measuring the performance of a sentence as opposed to the paragraph. The most important measure is not any one moment but rather, the aggregate performance of the marketing communication.


The answer is not just emotion

For marketers, the objective is to understand the importance of the primary emotion relative to the rational reasons to believe. The importance of emotion is dramatically different depending on the category and the characteristics of the consumer. Amongst thousands of studies conducted by Forethought, the one most strongly weighted towards emotion was undertaken for Kimberly Clark and the Huggies brand in the USA. In that study, for first-time moms, emotion represented 70% of the brand choice. By far, the single most important emotion was love. Amongst the rational drivers were leakage protection and price. A study in which Forethought has found the least amount of emotion in driving choice was undertaken for the leading Australian telecommunications company, Optus. In that case, the relative importance of emotion in the choice for topping up the credit for a pre-paid mobile phone was a mere 5%.

Advertisements for Optus cell phone and Huggies diapers

Of the many tens of thousands of equations that Forethought has produced for predicting drivers of choice, in every instance emotion has played a role in purchase behavior. This is entirely consistent with Antonio Damasio’s prognosis that we cannot decide without emotion.


The most rational to the most emotional

Once upon a time, creative agencies believed that the mere presence of an emotion would enable a marketing communication to be successful. Some creative agencies still believe this and indeed, rely on Binet and Field’s The Long and the Short of It to justify purely emotional ads where there is no rational reason to believe presented in the marketing communication. They hypothesise that building emotional connections will, in the long run, benefit the brand. Our research has shown that isolating emotional from rational content in communications is risky because it relies on the one consumer to see, recall, and link that marketing communication with the one brand. The astute creative agencies are learning to elicit primary category emotions and integrate in the same marketing communication, rational reasons to believe.


In a nutshell

There is minimal scientific evidence establishing facial coding as a predictor of market performance. Whereas with high degrees of predictive validity, the Prophecy Feelings scale is scientifically validated and provides an explanatory model for the relative importance of nine primary emotions in driving consumer behavior across categories. Marketers must first understand which emotion to communicate. This can be achieved by building a quantitative model determining the hierarchy of importance of primary emotions (and rational attributes – price and quality) driving category consumption. The creative should then be assessed to determine its performance in activating the Communications Triple Play[i]. Prophecy Feelings determines creative efficacy by measuring the degree to which the creative improves or detracts from the brand’s performance on the key emotional consumption drivers.


[i] The Communications Triple Play consists of communications that comprises of the primary emotion, a quality attribute, and a price attribute.


Photo by Ketut Subiyanto from Pexels