Text mining in hotel reviews
The primary problem most hotels face is in dealing with all of the unstructured content that is generated by their customer listening programs. Many companies collect a great deal of customer information, often in text form, yet few have the wherewithal to actually process all of it.
Online reviews and social media add an entirely new element into the mix. Some text analytics systems focus exclusively on customer feedback data. Unfortunately, this misses some of the most interesting content available. Your customers may tell you one thing and then may say something different in public; or perhaps they don't say anything to you and then they go complain on Twitter or Expedia. The ability to fetch, parse and produce information that is directly comparable to your internally gathered information provides a complete picture of the customer experience.
Hotel reviews represent one of the most fascinating uses of text analytics. Using publicly available customer reviews, a recent analysis was done comparing Bally’s verses Bellagio in Las Vegas using an application that plays to the strengths of sentiment scoring. In this case a group of reviews are rolled together to form a “consensus opinion” of hotels in a narrow geographic area. Automated engines are very accurate in such a use case – possibly more accurate than people – and they can handle a large volume of content. Their accuracy in a narrow vertical like hotel reviews is a result of tuning the engine to the language used in such a narrow vertical, where terms like “rude doorman” or “awesome meal” are commonplace.
The analysis of Bally’s verses Bellagio was done using something called “categories,” which is basically a fancy name for search strings. The important aspect of this analysis was in finding the sentiment associated with different important aspects of the hotel experience – using a known set of categories.
Rather than simply generating a score for each property, the reviews were scored for various features of the hotel, such as location and staff and dining. For this test we used reviews for Bellagio and Bally’s, measuring the following features for each:
- Rooms
- Price
- Facilities
- Location
- Cleanliness
- Service
- Overall
Digging deeper into the results, it is surprising to see that Bally’s had higher scores than Bellagio, since Bellagio is one of the 5-star properties in Vegas. Therefore, it was necessary to dig even deeper to make sure the scoring of the reviews was correct. We focused in on the most positive and most negative reviews and sought to determine why Bellagio was not scoring higher. The chart below shows that the “happy campers” were equally happy with Bellagio and Bally’s. The difference, however, was that the “unhappy visitors” were really unhappy with the Bellagio. By analyzing the reviews, it was discovered that people expected more for their money than they were getting at Bellagio.
Through the simple application of sentiment analysis on publicly available information, it is possible to show that companies can make these comparisons with much higher reliability, at minimal incremental cost and with an unprecedented ability to adjust categories on-the-fly, either based on these results or to test out new hypotheses. In fact, using this technique, we can move beyond the limitations of traditional approaches by running additional analysis to discover new and previously unmeasured categories based on recurring themes within the data.
What does this mean for brands? Those who are able to leverage sentiment analysis will remain at significant advantage over their competitors. They will be able to anticipate and proactively respond to how customers perceive their brand much faster and more comprehensively; and it is significantly cheaper than existing methods.
biography
Jeff Catlin is the CEO of Lexalytics (www.lexalytics.com), a software and services company specializing in text and sentiment analysis.