A smiley face with a line graph on it overlayed on top of a busy street photograph


You are watching TV when suddenly an advertisement comes on. Screeching! Bombastic! Exhorting you, yet again, to go to a particular shop to spend your hard-earned cash on some low-quality plastic trinket. It seems like the millionth time that you have seen and heard, this cacophony of cheapness!

How do you feel? Does this advertisement impel you leave your comfortable chair and drive down to the store in question? Or does it bring a scowl to your face as your eyes glaze over from boredom?

The next advertisement comes on. It is part of a series. It continues a story. You know you've seen it before, but you don't care. You're secretly hoping that it would be the next instalment, but hey, this one's still entertaining. You watch. It reminds you that you've run out of the product being promoted. Better get some on the way home from work!

The thing is your emotions have totally driven your reactions to the two advertisements. The two promotions have drawn completely different responses from you. The products themselves are irrelevant. Two decidedly different types of ads have resulted in two entirely different reactions.

In today’s world many marketers fear not evoking any reaction at all and consider any emotion valuable. Is any reaction truly better than no reaction?


What is Emotion?


The academic paper “Emotion and cognition in political communication” by Ann Iren Jamtøy, delivered to the 3rd International Conference on Democracy as Idea and Practice, gives a good description of some key emotional terms. Key terms that are important to note and sometimes confused include:

  • Emotions are "specific sets of physiological and mental dispositions triggered by the brain in response to the perceived significance of a situation or object for an individual’s goal."

  • Feelings are “the subjective awareness and experience of emotions"

  • Moods are "diffuse positive or negative states that last for longer periods of time [than emotions]”

  • Affect is wider yet again, and "is an umbrella term referring to an entire class of phenomena that is often taken to include both emotions, feelings and moods, but also pain, pleasure and basic human drives".

People are decidedly complex beings. We have to deal with different emotional states, feelings, moods and affect all the time. Although much of these can be internalised, it is nearly impossible to avoid displaying many of these states, particularly on our faces, as well as in our behavior.


Emotion versus Sentiment


The fact is that most successful marketing campaigns create emotional associations that connect consumers to the focal brand and its products. Whereas some of the least successful campaigns attempt to bash you over the head with their forceful messaging and lack of subtlety.

In Understanding Consumers’ Emotional Fingerprint: Why Sentiment Is Misleading Rick Miller argues "at the end of the day, emotions drive consumers’ purchase-related decisions. And that means brands and marketers need to start taking emotional measurements more seriously when creating content and planning campaigns".

You can gain an awful lot of insight into your consumers' real feelings if you can observe the emotions they display as they interact with your marketing.

The traditional way of doing this is with sentiment analysis. This involves determining the attitude of a speaker or a writer with respect to some topic. In practice, this means analysing whether an expressed opinion is positive, negative or neutral.

In the case of traditional market research, it would typically involve a subject being shown some marketing materials and then they would write down comments (or answer specific questions) about what they had just seen. The market researchers would be able to gauge their sentiment based on how favourable or unfavourable the answers they have written are.

The problem with traditional sentiment analysis, though is that it simply tends to give a broad glimpse into the market’s reaction to a topic, for instance, what consumers feel about a particular advertisement played.

It does not provide many insights into precisely how consumers are feeling, or indeed why they feel the way they do.


Getting Meaning From Emotion


Discovery Research uses several models to measure emotion. One of the models they follow uses a grid of four pairs of emotions; each pair representing opposite extremes of an emotion (for instance anger and joy).

Marketing content related to a particular brand is evaluated based on these axes. This evaluation helps to aid understanding of the emotion expressed about the brand. It becomes a guide to how the particular emotions drive customer action.

Discovery Research believes firms should undertake the following steps:

  • have an emotional philosophy - i.e. pick a model that breaks down emotion

  • operationalize the philosophy - create a way to collect, record and interpret consumers' emotions

  • measure the experience - use the method they have chosen to collect the emotional data

  • correlate the emotional components to actionable behavior - it is hard to determine a clear-cut correlation between emotions and specific actions - you need to find a particular trigger between an emotion and people's actions

  • trigger emotion to drive consumer behavior - once you know how your consumers are affected by particular emotions, tailor your marketing to ensure that it leads to the best emotions for you.


Face Detection and Emotion Analysis in the Future


So, what possibilities does face detection with emotion analysis offer in the future?

For a start, face detection is a great way of tracking people's emotions to differing situations. In particular, it is possible to track microexpressions that often give away a person's genuine feelings. Market researchers have already woken up to the fact that they can keep track of their test subjects' faces and observe how they react to certain situations. There is a real opportunity to catch underlying feelings by using this method of data collection.

Another situation where face detection with emotion analysis has potential is where you might want to judge the truthfulness of a person's statements. A liar could give himself away with his facial reactions, and a computer could detect this. Even the best poker player occasionally displays their real feelings in a fleeting microexpression.

Imagine combining both scenarios. Think of the research that could be undertaken in a casino, simply by focussing on the patrons’ faces!

Indeed this technology could provide great potential to improve the quality of the services offered in the hospitality industry. It would be so much easier to provide the services your customers truly want if you actually know what they are. Already casinos have trialled facial recognition technology searching for any known high-spending high-rolling customers on the premises, and ensuring that these people get the full luxury treatment.

We already have Google Glass. Imagine if this type of technology became ubiquitous; common enough so as not to appear unusual to passers-by. It would be hard to win an argument with a person wearing this technology, with a real-time display of what your emotions and real feelings are. At least it would mean a complete change in the process of buying a second-hand car!

This in turn makes one ponder possibilities for combine wearable technology with market research. In the future could we see market researchers wearing Google Glass while interviewing consumers for passive real-time feedback?

One topic that has been an area of concern for many people has been invasion of their privacy. There have been concerns expressed that the combination of emotions analysis and facial detection provides the possibility of opening up the insides of your head for public scrutiny if there are not adequate privacy protections provided. Our recent article How does Privacy work with facial recognition discusses that vexed topic.




There was a time when computing and emotions were effectively at the opposites ends of a continuum. The television series Star Trek made several episodes focussing on this very point, and the scriptwriters clearly still expected the disconnection to still be a problem in the 24th Century. However rapid progress has been made in bringing the technological reading and analysis of emotions and feelings much closer.

Now that emotional analysis is technically possible, there is a move towards finding practical uses for it. Using it to assist with marketing and determining consumer insights is a clear and obvious move.

Of course the possibilities are much wider than this. Making computers more human in nature opens a whole new world of potential outcomes from software design, to wearables, to smart homes, and any other experience that includes computers and cameras.



Verify people in your apps—Integrate face recognition with our easy-to-code API.



Discover the benefits of Kairos Face Recognition—Let's connect.




Ready to get started with Kairos?