Firms that are able to adapt their marketing instruments to match consumer segments are efficient; those that do not forego potential profit. For example, it is reasonable to expect that websites are more preferred and increase sales if their characteristics (e.g., a more detailed-data website) match a customer’s unique characteristics (e.g., a more analytic cognitive style).
Because consumer segments are often unobserved (e.g., cognitive styles, purchase funnel stages, cultural styles, or propensity to defect), learning and matching them to marketing instruments (e.g., banners ads, promotions, website designs, or price levels) in real time is a formidable challenge. It requires learning the consumer’s segment and the success probabilities for each version of the marketing instrument (e.g., the purchase probability of analytic consumers when they use a data-intensive web design.) This dual learning while maximizing sales makes the problem even harder. “Morphing” involves automatically matching marketing instruments to latent consumer segments while solving, in real time, this classical problem of learn–versus-earn.
Who is Gui?
Gui Liberali is the Endowed Professor of Digital Marketing at the Rotterdam School of Management (RSM) of the Erasmus University. He holds a doctorate in marketing and a B.Sc. in computer science. His work has appeared on Marketing Science, Management Science, International Journal of Marketing Research, Sloan Management Review, and European Journal of Operational Research. His research interests include optimal learning, multi-armed bandits, digital experimentation, natural language processing, morphing theory and applications (e.g., website morphing, ad morphing), dynamic programming, machine learning, and product line optimization.
Personal webpage: www.guiliberali.org