Pragmatic Estimation of the Elasticity of Demand in Hospitality From Noisy Reservations

In hotels, future prices should be determined based on predictions of future demand and the responses of clients to price changes. Price changes will actively affect future demand. If this effect is neglected, opportunities for promptly and accurately setting prices will be lost, as well as potential increased profits. The role of elasticity of demand in setting optimal prices is well-developed in Economics, but the practical estimation of elasticity for a specific hotel from data about the pickup of reservations is worth investigating. In this paper, we highlight the risk of estimations based on a single A/B test and propose practical rules and pragmatic experiments. The analysis is based on statistics but simplified so that the results can be easily applied in single hotels without excessive disruption of daily operations. After defining rules to derive error bars on the estimates, we experiment in different situations, including estimations from scratch or by gradually tracking abrupt or seasonal changes, that are more realistic for a hotel in operation. Our methods can be useful to hotel managers who wish to decide about prices based on measuring the response of their potential customers, in a data-driven manner and based on statistically significant estimates.