On October 11th, 2023, I attended a talk entitled Mastering the Emotional User Experience by Bill Albert, SVP of Global Customer Experience at Mach49 in Silicon Valley. Albert began the talk by noting that the emotional user experience was the topic of a recently added chapter in the newest edition of his book, Measuring the User Experience: Collecting, Analyzing, and Presenting Usability Metrics and something his company is now just beginning to touch on.
Why should we care about measuring emotional UX? Albert cites improving products/services beyond usability, achieving KPIs (loyalty, satisfaction, etc.), aligning with brand strategy, and gaining competitive advantage as positive outcomes for considering emotions in UX research.
Albert utilizes scales of “Arousal” and “Valence” to classify emotions. Arousal classifies the degree of excitement or engagement from “calm” to “excited’. Valence measures the “goodness” or “badness” of a response from negative to positive. Albert argues that only a few emotions are truly relevant to UX design. He includes engagement, confidence, trust, frustration, affect, and stress, but also notes that these are highly contextual. As an example, Albert shared a user testing video for a participant who was laughing at the poor functioning of a website. The face tracking software registered her smiles as a “joyful” response, where in reality she was amused at how bad the website was. Albert also notes that UX testing usually elicits a “weak signal”, or low emotional intensity, and is thus not as obvious to measure compared to high emotional intensity events such as a roller coaster ride, a great first date, or a team winning a big game. The example of the user laughing at the poor functioning of the website, and that response being registered as “joy” is known as “noisy data”, or data that has to be sifted through and removed by a person, not AI. Albert points out that we don’t all have the same definition of each emotion, and we have to also consider cultural and language-based differences – this all results in “noisy data”.
As we begin the semester, I am thinking about how I can narrow my research and get closer to my final thesis topic. Bill Albert’s talk was very relevant to my past research, as he exposed the shortcomings of AI in understanding human behavior, and the prevailing importance of having real people test, analyze, and summarize data. This “impulse” was a great way to get the ball rolling for this semester of Design and Research, and will provide “food for thought” moving forward.
