Beyond the bid amount : an empirical analysis of online upselling of hotel rooms

Abstract

Hotel companies implement online upselling with the motive of increasing revenue per customer. However, several unknowns exist regarding selecting the right customers for online upselling and quantifying the impact of these decisions on incremental revenues. This study applied a two-step analytical approach involving logit regression and propensity score matching to analyze longitudinal online upselling bidding data. The purpose was to uncover factors influencing a hotel’s online upselling decisions and their impact on the multiple revenue-generating points: rooms, food and beverage, and other facilities and services. The results show that upselling bids from repeated and long-haul guests are more likely to be accepted, while upselling bids with Saturday-night stay(s) and longer stay lengths are less likely to be accepted. The bidding time, the number of bids and the average amount of bids also increase the odds of acceptance. Upsell decisions based on these factors positively impacted revenue per customer from rooms, other amenities and services, and total revenue by an average of 48%, 84%, and 42%, respectively. These insights can be leveraged to target valuable customers for upselling.