Facial Coding – Measuring Emotions without the Guesswork

There is extensive evidence that form a number of scientific fields that suggest that the traditional economists’ view of consumers as ‘rational optimisers’ is not correct.

Behavioural Economics is the field most frequently cited as providing this contrary view. Work here, including that of Nobel-prize winner Daniel Kahneman seeks to inform economic theory with real-world observations of the biases in human decision making, and shows clearly that people rarely weigh up all the options in a dispassionate, objective and through manner.

Similarly work from the fields of psychology in particular the study of heuristics, shows again that most of the time we use shortcuts or rules of thumb in decision making, rather than going to the trouble of thinking things through fully.

Intuitive processes are fast, easy, automatic – they just happen. – You ‘know’ ice cream tastes nice or that you (hopefully) love your family, without thinking. They are also, however malleable and sensitive to context without you realising. The smell of bread or coffee can spark hunger without you really considering why the supermarket may be allowing the smell to waft your way. These processes contribute to your ‘gut feel’.

The Reflective processes are harder work. They are slow, sequential, take effort, and are what we think of when we think of thinking or concentrating. They are, however, much less easily influenced by immediate context and allow us to make abstracted or rule-based choices. They can and do override intuitive responses, but because they require effort to do so, when we are tired or stressed or simply not motivated to ‘think’, intuitive processes can come to the fore in decision making more than we like to believe.

To understand consumer in their decision journeys in everyday life, we need to anticipate both sides of the mental work space.

If you want to read more about Daniel Kahneman and Facial Coding in his field, see below.

In any given decision, both systems will be at play, but the degree to which each one influences the outcome varies from time to time, and from context to context.

Millward Brown survey tools usually actively encourage Reflective thinking. Ad testing e.g. asks people to think harder about the ad, and their response to it, and what it means, than they will do in reality. This still has huge power as Millward Brown extensive validation studies show. However, we do need to ensure that we generate a good read on the intuitive responses that people have as well, so we can moderate our interpretations based on that, and so we can understand when people may be giving us the answer they think they want us to hear.

It is behavioural and biometric methods that can allow us this insight into intuitive processes, as these processes have biological effects that we can read and use, without relying on verbal reports.

Facial Coding is one that allows us to do so with high ppracticality, since integration is crucial.

The science behind this is very strong, and goes back as far as Darwin. He was one of the first to posit that some expressions are both universal, and have their meaning has roots in the biology that we share with our primate ancestors.

In the 1960s and 70’s, Paul Ekman and his co-authors established firmly that there are indeed a set of universal expressions that are interpreted as having similar meanings among all cultures – these are happiness, fear, anger, surprise, disgust & sadness. He also pioneered a system for consistently coding all the potential movements of the face – the Facial Action Coding System, from which characteristic expressions can be built up. Furthermore, Ekman also established that many of these movements are automatically evoked and therefore difficult to suppress.

The breakthrough has been the development of automated facial coding based on the FACS system.  The system allows to automatically recognise key expressions, and thus the cost of using facial expression data has been slashed tremendously, allowing us to use the approach at scale.

Metrics cover both continuous dimensions of response and discrete emotions. They allow us to …

  • Understand real emotional power of spots, and the power of story-telling
  • Highlight the key moments, to inform optimisation & edit
  • Have a greater understanding of comprehension and ‘wear in’ issues
  • Clarify issues highlighted but not explained by survey responses.

As such Facial Coding can improve the predictive power of survey-based advertising research. Key questions addressed by facial coding:

  • Is advertising really engaging viewers emotionally, and in the way we intend?
  • Are we telling a strong enough story?
  • How are people responding to the main creative idea?
  • What are the ‘key moments’ which are driving emotional responses and motivation?

Integrating Facial Coding brings assessment of instinctive emotional responses into mainstream research tools, without losing validated and trusted tools.

For more details about Facial Coding and how to integrate in your advertising testing, please contact Millward Brown Nordic here.

Leave a Comment

Please complete the equation *