Studying emotion is a pretty tough subject.  Most market researchers really want to understand how customers experience their brand emotionally.  The problem is, on a basic level it's not entirely clear how to define emotion.  Sure, you can look up emotion on and there's a definition.  It will tell you that emotion is "an affective state of consciousness in which joy, sorrow, fear, hate, or the like, is experienced, as distinguished from cognitive and volitional states of consciousness."

Huh?  At least it talks about how emotion is experienced (how we apply meaning to emotion).

Google defines emotion as "a natural instinctive state of mind deriving from one's circumstances, mood, or relationships with others."

This one is easier to understand, however this definition eliminates experience and clearly states that emotion is "a natural instinctive state."  In essence, it stakes the claim that emotion is "nature" rather than "nurture."  

The inability to clearly define emotion complicates things.  Does everyone have emotion?  Does everyone experience emotion in the same way?  Is the meaning of emotion good or bad when someone states, "he's being emotional"?  As with most things, I believe that the meaning and experience of emotion varies on the individual.  So how do you sort it out and find meaning from emotion?

Here's how we do it at Discovery Research.

A number of years back, we started studying emotion as it pertains to market research.  As you'd guess, our emotional modeling and measurements have evolved (and are still evolving).  Discovery is very interested in social media research.  One of the reasons is that it generates so much unstructured text that is filled with emotional expression.  One of the things we do with our social media research is attempt to get past simple sentiment (positive, negative, neutral) to some sort of emotional understanding of the research question at hand.

Here's some output [pictured above] from some work that we did a few years ago on Energy Drinks.

You can see from this example that conversation around this particular energy drink brand can be evaluated on an emotional grid.  The emotional categorizations here are broken down on 8 opposite characteristics.  Brand content can be evaluated based on these axes to understand the emotion expressed around the brand and how these emotions drive customer action.  We continue to use this model often, however, it has posed some questions in its simplicity.  More recently, we've added additional analysis of emotion to include a psychological researcher's (Parrott) tree-based emotional modeling.  This model begins by breaking down emotion into 6 components (Love, Joy, Surprise, Anger, Sadness, Fear) and stakes the claim that all emotion is a subset of one of those six categories.  There are 25 secondary emotions, and an additional 115 or so tertiary emotions.    

These two models are not the only ones for measuring emotion.  Emotion can be measured using qualitative measures (analysis of unstructured conversation using text analytic techniques), and quantitative measures (scales based on how something makes you feel).  Given the complexity of understanding the experience of emotion, how do you proceed?

Here's what we'd recommend:

  • Have an emotional philosophy.  Choose a model.  Find one that you like.  There are many of them that break down emotion in various different ways.  If you don't have a philosophy for handling emotion you won't be able to measure it.  You'll never understand what it means when it's expressed.
  • Operationalize the philosophy.  We often talk about how to "operationally define" things when we conduct research.  When you segment customer categories, you define the characteristics of the segment and then you create models (questions or categories) that make up those characteristics.  You'll have to do the same for emotion so that the experience of emotion is standardized as much as possible.  We've created linguistic models that understand emotional categories and we apply them to unstructured conversation, then we consistently QC the models.  You can do the same with a survey instrument.  
  • Measure the experience.  Use what you've built.  Survey your customers using your model.  Grab that online content where the customer feeling is expressed.  Then follow good scientific techniques for testing your reliability.  Do your emotional models measure what you think they should be measuring?
  • Correlate the emotional components to actionable behavior.  This is the biggest problem with emotional research.  Bridging the gap between emotion and action isn't easy.  Marketing research needs to be actionable.  In order to be actionable, or understand how to drive action, your emotional components must correlate to a specific trigger.
  • Trigger emotion to drive consumer behavior.  Really, at the end of the day, isn't this why you are interested in emotion in the first place, or what you are trying to do in your advertising.  Understanding emotional triggers that result in consumer behavior becomes the holy grail of the marketing world.  You've heard the song "Make them Laugh." There's a reason that so much advertising tries to stimulate the Joy and Surprise emotions.

Finally, it's important to reiterate that you must find a way to be actionable in your research on emotion.  Frankly, without it we're doomed.  Unfortunately, that's the most difficult part of emotion.  In research, we know that consumers believe they make decisions based on a logical stepwise process.  We also know that in 9 out of 10 situations, that belief couldn't be further from the truth.  Consumers base their decisions on emotion not logic.  It's about time that we, as an industry, understand emotion much more thoroughly.

Download the Energy Drinks Conversation Analysis whitepaper.

Originally published on Landmark as "Using Market Research to Define the Meaning in Emotion"