Ads that evoke emotions keep us entertained and are more memorable, but do they really drive product sales? To answer this question, we use facial coding to quantitatively measure peoples tacit emotional responses to ads, and then model the facial responses with sales data for these ads. We collected data from a 1000+ viewers as they watched 140 ads from the comfort of their homes in the US, the UK, France and Germany. Using dynamic modeling, we identified the facial measures and emotion trajectories that are most predictive of sales effectiveness. In this session, we will share our key findings and show how this approach can be applied to other areas of market research like concept testing and media research.