Are the drivers of liking the same as drivers of purchase intent?


Are the drivers of liking the same as drivers of purchase intent? Bayesian networks give us an answer.


Typically, the primary indicator of product performance is Overall Liking. Consumer research, conducted through blind product testing, identifies the emotional response to the sensory properties of a product. Post consumption, emotional responses to the sensory properties provide an indicator towards repurchase behaviour.


SAM wanted to further explore the correlations and differences between the motivations influencing consumer Overall liking and Purchase Intent.



Category appraisal was performed on 16 different dark chocolates in blind sequential monadic approach by 360 consumers, evaluating overall liking, purchase intent, emotional response, claims perceptions, expected benefits and usage situations.


Through the utilisation of Bayesian networks we observed the probabilistic dependence or independence between several variables grouped into different factors: aromatics organoleptic characteristics, consumption habits and frequency, extrinsic attributes, expectations, etc to obtain a model of joint probability that can best explain the relationships between the variables.



Bayesian Network data revealed a direct relationship between Purchase Intent and Product Claims which was not the case for Overall Liking, only strongly connected to organoleptic items. A non-elusive explanation for the differences observed could be in the distinction of the indicators themselves – that Overall liking is an initial and spontaneous, implicit response to a product; whereas Purchase Intent is a conscious, rational choice relating to a multiple of factors that elicits an intended behaviour response.

Read more