Colorado Springs hotel market compression detection tool

摘要

In this paper we describe the development of a model to assist in hotel room pricing by predicting upcoming market compression with multiple linear regression. Our client is a local 4-Star hotel in Colorado Springs, Colorado with 500 guestrooms. Currently, the hotel receives data on the rates of their competitive set (CompSet) of hotels in the Colorado Springs market on a weekly basis from a contracted market researcher. Additionally, the hotel gathers information about upcoming local events that potentially bring in large groups of guests. Using a heuristic set of rules, the revenue manager utilizes the data to determine the nightly rate of the rooms to maximize revenue. The process is iterated until the stay night; the revenue manager revisits the nightly rates as more data becomes available. We focus on providing a simple to use tool that systematically automates the process of analyzing the weekly market data received. Our decision support tool receives pricing data on the CompSet, analyzes the data using a series of linear models, and projects a market compression percentile for dates up to 90 days in the future, accounting for up to 82% of the variation in compression. This decision support tool allows the revenue manager to base their heuristic guest type allocation and room pricing decisions on precise predictions of market conditions.