Ithaca, NY, February 11, 2016 – Hotel guests often post a review of their stay along with a numerical rating of the hotel, but a new study from the Center for Hospitality Research (CHR) finds that the comments and the ratings don't always match up. Instead, negative comments have a heavier weight on the guests' ratings than positive comments, a finding that points to the importance of consistency in hotel service.

The study, "What Guests Really Think of Your Hotel: Text Analytics of Online Customer Reviews," by Hyun Jeong "Spring" Han, Shawn Mankad, Nagesh Gavirneni, and Rohit Verma, is available from CHR at no charge. The researchers analyzed 5,830 TripAdvisor reviews covering 57 hotels in Moscow by identifying and sorting over 18,000 key words relating to five specific hotel topics, namely, amenities, experience, location, transactions, and value.

Han is an assistant professor at the School of Business Administration at the National University Higher School of Economics in Moscow; Mankad is an assistant professor at the Samuel Curtis Johnson School of Management at Cornell University, where Gavirneni is an associate professor; and Verma is executive director of the Cornell Institute for Healthy Futures and the Singapore Tourism Board Distinguished Professor in Asian Hospitality Management at the Cornell School of Hotel Administration.

Considering the outcome of the text analytics study, Verma explained: "Text analytics of guest reviews provide insights about the reasons for customer satisfaction and dissatisfaction that often are not apparent by just examining the numerical scores. Given that customers of hospitality and healthcare and other industries are increasing providing comments in addition to numerical ratings, it is essential that managers use text analytics to understand the underlying customer sentiments. Such an approach will provide better insights for quality and process improvement."

One particularly noteworthy finding of this research is that negative reviews tended to be relatively long and they focused tightly on a limited number of hotel attributes. In particular, unhappy guests focused on elements relating to the value provided by the hotel and its transactions, that is, the mechanics of the stay. On the other hand, guests who wrote relatively briefer reviews that took a wider view of their hotel stay generally assigned higher ratings.

The authors conclude that sophisticated text analytics can give hotel managers specific information about how to improve their guests' experience—and that information is not available from numerical ratings.