Facial expressions as predictors of online buying intention
This study attempts to correlate the movement of facial features to consumer purchasing behavior and ultimately detect patterns in facial expressions that will predict behaviors. The facial expression of forty-two undergraduate and graduate students were recorded while they participated in an online shopping experiment where they rated twenty-four products pretested for humor and involvement levels and chose five final items to purchase. Datasets were then created by postprocessing the videos to extract the students' facial features and the movements of these features were calculated using computer software. Finally, collected datasets were analyzed using two learning algorithm classifiers. The results were promising in that classifiers were able to predict buyer intent substantially above chance. Some of the best predictions were made regarding purchase likelihood of subjects, more specifically on humorous and high-involvement product datasets. Theoretical implications and possibilities for practical applications using facial expressions as predictors of online buying intention are discussed based on these results.