What constitutes a good location?

In the mid-1980s, La Quinta Motor Inns was experiencing increased competition from hotel chains. Competitors were seeking to enter La Quinta’s market niche serving frequent business travelers. Among the competitive factors deemed most important by La Quinta’s management was the location of its hotels. The question was especially important as La Quinta was expecting to enter a period of rapid expansion. In 1987, La Quinta operated 151 mature inns. In 2011, the company operated or provided franchise services to more than 800 hotels in the United States, Canada and Mexico.

The decision process: Before the application of analytics, hotel location decisions were based on the combined expertise of many different individuals: the director of marketing research, the vice president for development, the vice president of real estate, four site evaluators and La Quinta’s president. The firm’s president made final decisions and felt obliged to be actively involved in site location decisions. It was a business model that couldn’t easily scale.

Researchers from the School of Hotel Administration at Cornell University and the Department of Management at the University of Texas, Austin, were brought onboard to apply analytics. The question they sought to answer: What location-related factors best predict the success or failure of a hotel?

The project: The project consisted of a regression analysis. Rather than start with a blank slate for choosing explanatory factors, the researchers drew upon the expertise of the decision-makers as well as published literature to arrive at 35 potential factors. Data was gathered for 57 of La Quinta’s mature inns over a three-year business cycle. The model was fitted to predict hotel operating margin. The predicted operating margin of hotel sites under consideration could thus be estimated by supplying the values of explanatory factors to the model (e.g., the amount of office space within four miles of the location).

Beyond predicting operating margin, the modeling efforts determined which explanatory variables were most highly correlated with success. Among the most important factors were those related to brand recognition. For example, inns located in a region with other inns of the same brand tended to fare better than those in new areas. The economic health of a region was a useful predictor, as were signage and traffic flow. In all, the researchers found 16 variables that correlated significantly with hotel operating margin. Just as importantly, the effort weeded out variables that were thought to be explanatory but weren’t. For example, the total number of competitive hotel rooms in close geographic proximity wasn’t a useful predictor.

The results: Sam Barshop, chairman of the board and president of La Quinta Motor Inns at the time of the study, was pleased with the results, noting, “…the researchers were able to come up with a usable and helpful model to assist us in our decision making process. We currently use the model to help us in our site-screening process and have found that it raised the ‘red flag’ on several sites we had under consideration. We plan to continue using and updating this model in the future in our attempt to make La Quinta a leader in the business hotel market.”

Based on “Profitable Hotel Sites at La Quinta Motor Inns,” by S. E. Kimes and J. A. Fitzsimmons, Interfaces, Vol. 2, March-April 1990, pp. 12-20.