Length of stay heuristics Do they really make a difference?

Abstract

Most of the original revenue management systems used by hotels assumed one-night stays. In this article, an evaluation is made of a more sophisticated decision rule that incorporates a guest's actual length of stay (whether it be one or more nights). A heuristic approach is presented that could be used by more firms in the hotel industry to achieve improvements in expected contribution of up to 2.94% over Belobaba's expected marginal revenue decision rule (a standard of comparison for decision rules that don't incorporate length of stay). The sensitivity of the difference in expected contribution between these rules is tested for three different hotel properties (actual hotel data) which have differing values for the input parameters (e.g., length of stay profile, no-show rates, average rates for each rate category), and managerial insights are presented.