A survey and non-survey evaluation of utilities websites was done during common user tasks. Customer video verbatims were recorded as part of surveys. Face video and eye-tracking were also recorded during labs-based sessions. Face Emotions were assessed as an average percent of emotions expressed across tasks during recorded video per brand and were matched with satisfaction ratings per respondent. Analysis of Variance of emotions indicated statistical significant brand differences for Joy [ F(6, 226) = 2.25, p = .026 ] and also for Frustration, Anger, and Contempt. Correlations showed that lower ratings of website appearance (r = -.49, p < .02) and navigation (r = -.50, p < .02) increased negative emotions for customers. Regression analysis also suggests that face emotions can predict customer satisfaction ratings. Conclusions: Face emotions is a viable assessment and corresponds well with survey ratings. Brand comparisons on emotions provides key customer insights.
- Face emotions from video analysis is a viable assessment of customer satisfaction
- Face emotions correspond well with survey ratings and provides new client deliverables
- Brand comparisons on emotions metrics provides key customer insights.