Models of consumer choices have operated based on the principles of the rationality of individuals. When observing consumers choice, those are supposed to be maximizing their utility of a given good or service. Albeit this assumption has been functioning so far, increased understanding of human decision making has evidenced that consumers choices are not only driven by conscious factors but also result from unconscious factors. Recent technological developments now make it possible to incorporate biometric data into models of consumer choices. Here, we introduce a protocol for a laboratory experiment that aim to understand consumers preferences for labelled olive oil and the impact of information about the health benefits of olive oil on their willingness to pay for a litre of olive oil. The protocol combines an auction mechanism known as the Becker–DeGroot–Marschak procedure with the measurement of biometric data using an eye-tracker and. Several biometric features are measured: eye-movements but also facial expression. We then compare different consumer choice models based on different sets on factors and discuss their predictive power. While the Tobit model deriving consumers’ willingness to pay for olive oil allows good prediction of consumers’ behaviour, the mixed models incorporating biometric data have a higher predictive power.