The Emotional Effectiveness of Advertisement

F. Javier Otamendi

Dolores Lucia Sutil Martín

Based on cognitive–emotional neuroscience, the effectiveness of advertisement is measured in terms of individuals’ unconscious emotional responses. Using AFFDEX to record and analyze facial expressions, a combination of indicators that track both basic emotions and individual involvement is used to quantitatively determine if a spot causes high levels of ad liking in terms of attention, engagement, valence, and joy. We use as a test case a real campaign, in which a spot composed of 31 scenes (images, text, and the brand logo) is shown to subjects divided into five groups in terms of age and gender. The target group of mature women shows statistically more positive emotions and involvement than the rest of the groups, demonstrating the emotional effectiveness of the spot. Each other experimental groups show specific negative emotions as a function of their age and for certain blocks of scenes.

To carry out the emotional measurements in this study, a software platform for biometric measurements research called iMotions was used (iMotions, 2020). This company indicates that its software can combine “eye tracking, facial expression analysis, EEG, GSR, EMG, ECG, and surveys” (Taggart et al., 2016). The platform is used for various types of academic and business research. Version 7.0 was used in this research.

The software records several raw indicators per frame based on biometric measurements or action units while an experimental subject is watching a stimulus on the computer screen: 34 core facial landmarks (jaw, brows, nose…), interocular distance, and head position (yaw, pitch, and roll).

The recorded values for the raw indicators are then transformed by the software underlying models into Ekman’s seven basic emotions. An indicator for each emotion is provided based on the probability of appearance of the emotion, so the range of values for each of them is from 0 to 100. A value of 50 is proposed by AFFDEX as an initial threshold to determine if an emotional response has been detected.

This publication uses Facial Expression Analysis which is fully integrated into iMotions Lab

Learn more