A Dynamic Segmentation Model in a Multichannel Environment | Weatherhead

A Dynamic Segmentation Model in a Multichannel Environment

A Dynamic Segmentation Model in a Multichannel Environment

Authors

Published

International Journal of Research in Marketing

Abstract

On one hand, consumer buying from multiple channels (e.g., Internet and retail store) is linked to an increase in the number of purchases and more profits (Ansari et al. 2008, Venkatesan et al. 2007). However, research also finds that multichannel enthusiasts are less brand-loyal than retail store customers (Konus et al. 2008) and long-term purchase incidence decreases with increase in Internet channel usage (Ansari et al. 2008). These contradictory findings point to the need for more research to disentangle the long-term effects of marketing activities and channel choice across channels on buying behaviour. One possible explanation to these counter-intuitive findings is that prior models in the channel choice literature ignored dynamics inherent in consumer buying behavior over long term. Ignoring dynamics might lead to erroneous parameter estimates when modeling consumer behavior (Fader and Hardie 2010). As such, we explicitly model these dynamics by extending the Poisson hurdle model to 1) allow for a multinomial decision in the first part of the model to account for channel choice, and 2) assess the impact of marketing promotions on a consumer’s propensity to buy by introducing random coefficients driven by a hidden Markov chain. Using data from a multichannel retailer over a nine-year period, we empirically validate our model. Our results indicate how accounting for dynamics changes the conclusions about which type of marketing activities are effective in promoting same-channel and cross-channel purchase behavior.