Reaching Beyond the Stars: How to Mine the Best Insights from your Hotel Reviews.
By Jeff Catlin, CEO - Lexalytics
Social Media has been fully accepted within the travel and leisure industry. Virtually all of the major hotel chains have active social media engagement programs designed to enhance the image or their brands and sniff out minor issues before they become major ones. This article will take a more detailed look at how social engagement should differ for the various brands within a hotel chain. Specifically, it will examine how sentiment is not a universal measure, and that it's not only correct, but appropriate that sentiment for high end hotel chains will have a much larger variance (more really positive and really negative comments).
OK, some people have more clout than others, that has been a truism in marketing for generations, so let's remove the authors from the discussion and assume (incorrectly) that all tweets are created equal. In this utopian world we'd like to believe that the sentiment of a tweet or customer review can be scored on an equal scale regardless of the property being reviewed, but our research has shown that customer expectations play a huge role in people's sentiment about a property. Let's say our sentiment technology is capable of reading reviews and scoring it on a 1 to 10 scale, where 10 is great.
If in fact all tweets are created equal then a review claiming that the front desk service at a Motel 6 was amazing would carry the same weight as a comment on the great front desk at a Mandarin Oriental. Our research shows that the comment for the Motel 6 should carry much more weight than that around the Mandarin Oriental. On the surface, this seems a bit counter intuitive so let's take a deeper look at why strong positive sentiment for Motel 6 carries more weight.
Customer expectations play an important role in understanding the sentiment at a particular property. If a customer books in at a standard run of the mill hotel that does not portray itself as a vacation destination then that customer is in all likelihood is expecting a clean and quiet room and an efficient front desk but not much more. If such a customer bothers to fill out a survey and is strongly positive about the front desk staff, then clearly the employees here have done something that caught the customers attention and drove their satisfaction up, but why am I contending that this is more important than a similarly positive review at a 5 star property like a Mandarin Oriental or a St. Regis. It turns out that user expectations play a huge role in sentiment, particularly when you compare star ratings with written commentary.
If you look at the star ratings of a supposed 5 star property you'll see mostly 4 and 5 star ratings for the different departments (restaurant, front desk, housekeeping, etc…) but if you were to examine the sentiment scores for written comments about the same departments you'd see a much wider variance. Customers have a built in bias for star ratings, even if they are unsatisfied with the front desk at a 5 star property, they are unlikely to give it anything less than 3 stars, but if you were to look at the written comments from that customer about their issues with the front desk you'd quickly see that they had a 1 star experience.
Jeff Catlin is currently the co-founder and CEO of Lexalytics, Inc.The company launched the world’s first commercial sentiment analysis engine and currently processes billions of documents a day for the top social listening and customer experience management companies in the world. Re-investing millions of dollars a year into development and operations has kept Lexalytics at the forefront of innovation and ahead of the competition.More from Jeff Catlin