A few people sit down to play a card game. It’s not just any card game though. Some cards are associated with monetary rewards while others are associated with punishments—a game specifically designed to simulate real-life decision making through trial and error of learning short-term consequences. The players are unique as well—a group of people whose limbic region of the brain had been impaired from injury.
The goal? Scientists wanted to understand the link between decision-making and the area of our brain that controls emotion, behavior, motivation, long-term memory, and olfaction (smell). This same region lights up when we have to realize our decisions and act. It has no capacity for language or logic.
In this card-game study, it was discovered that “without emotions, decision-making was impaired. Patients whose limbic systems were malfunctioning due to injury consistently took the wrong decisions and lost money in card games,” reported Katharina Kuehn, director of RDG Insights.
Why is this important to know?
As a marketer, you want a person to choose your solution over another.
As a customer experience professional, you want a person to keep choosing your solution over another.
That’s basic loyalty.
There is a clear link between choosing and the emotions we feel. If we wish to persuade our audiences to choose and rechoose us, we must understand the emotions and motivations behind those choices.
The decision-making process begins and ends with emotion. All the reasoning and logic in between these emotional bookends only validates or devalues our initial emotions until we feel right or wrong about our impressions and research. Only then do we act.
“The consciousness of a consumer is like a government spokesperson who has to announce decisions they a) did not participate in and b) whose true decision-making reasons are not accessible to them,” observed the American neuro-philosopher Daniel Dennett.
How we’ve tried to measure emotions in the past
In the past, data gathered from solicited responses provided insight into the sentiment and tone of how something was perceived or experienced.
However, that data could only offer emotional insight if the questions were geared to receive that type of feedback.
It’s important to distinguish the difference between sentiment analysis vs. emotion analysis. Sentiment analysis reflects the polarity of audience tone on a scale of positive to negative. This is not emotion. You can feel joy with a negative sentiment or exhibit disgust with a positive tone. Emotion analysis delivers insights into how your audience is feeling. Those feelings precede decisions. Those decisions become behavior. Sentiment alone cannot deliver the motivation, desires, nor behavior behind decisions made. However, sentiment + emotion can.
The question remains, if emotion is at the center of measuring effective experiences and brand perception, how can we gather large amounts of useful unsolicited emotional data?
Gathering Unsolicited Emotions Data
A tool with true emotion analysis must comprise 3 key elements:
- Focus on emotions, not sentiment
- Human annotation
- AI-Assistance for scalability
Focus on Emotions, not sentiment.
As we already mentioned, sentiment is tone of voice. Emotion is feeling. Sentiment is incredibly important and should be paired with emotion analysis. To analyze emotions, actual emotions should be the focus of data ingestion and interpretation.
To do this, Plutchik’s wheel of emotions is a great guide for which emotions to measure. Plutchik discovered the 8 core emotions we feel as human beings:
Let humans annotate human emotions
In order to accurately predict emotion data, you must employ actual humans to read and analyze millions of text blurbs to interpret which emotion and the level of emotion expressed in their language. Then, let other humans analyze the same data sets to ensure the interpretation is consistent person to person.
Without the touch of a sentient-being analyzing data from other sentient-beings, you have no basis to analyze further data. This is all meant to create a model to teach computers.
Scalability through AI.
All of this comes together when these emotion analysis models are put through algorithms to amplify and scale the analytical power of a human being. Without the aid of artificial intelligence, it would be impossible to gather the scale of data by which any reliable insights can be gathered.
Don’t worry, we’ve done all the work for you. 😊
To put it simply, Nuvi has done all this work for you. Earlier this year, we launched our proprietary Nuvi Language Engine which became industry-leading in both accuracy and speed. After adding the ability to analyze emotions, we were awarded Martech’s 2020 Breakthrough Award for “Best Social Media Analytics Platform.”
Learn how you can begin to harness the power of emotion analysis to bring greater impact to your customer experience strategy.
I want to learn more!