Unlocking insight with guest feedback text analytics

Written guest comments contain the specifics on how to improve your hotels. Today, there exists significant opportunity for our industry to process this feedback for better operational and marketing insights. That’s the message of a new paper from Cornell’s Center for Hospitality Research: What Guests Really Think of Your Hotel: Text Analytics of Online Customer Reviews. I encourage you to read the full report, but key takeaways include the following: 

Intercepting service failure is key. “Negative comments have a heavier weight in a guest’s rating of a hotel than do positive comments. Guests’ bad feelings from poor service generally will submerge their favorable feelings from good service.”

The key here is to surface dissatisfiers while the guest is still on property with staff training and technology like in-stay surveys, to catch and resolve an issue before it’s posted online.

Know what matters to guests in your segment. “Ratings differ based on hotel segment. For instance, the guest’s experience was mentioned more commonly in reviews of high-tier hotels, while amenities and location came up more frequently for motels in the middle tier compared to hotels in other tiers. Guests at the lower tier properties wrote more commonly about transactions and value than those staying at hotels in the middle and high tiers.” 

Consistency matters. “It is better for hotels to provide guests with a moderately good overall experience than an experience that is extremely good in some regards and terrible in other ways, because in terms of ratings. The weight of the terrible service will swamp the good feelings from the stay’s excellent aspects. Therefore, in addition to the overall quality of service, the consistency with which this service is delivered to a guest is itself an important driver of guest satisfaction.”

Improvement is the goal. The study’s authors talk about using a “managerial” approach to text analytics in order to drive change at the hotel level. This isn’t just an abstract statistics project. We need insight for action to make our guests happier, increase loyalty and drive revenue.

Online “star” ratings of hotels only tell part of the story of a guest’s experience. Typically, guests will leave several paragraphs of text talking about everything from their arrival at the hotel to the view from the room – and everything in between. 

Text analytics to process written feedback for sentiment has existed for a while, but its value for obtaining insight for improvement has increased as the technology has evolved. 

As a hotel executive, you need to ask yourself the following about your current text analysis capabilities:

Do our text analytics provide us with enough detail to make decisions with? Or, perhaps more telling, what decisions have we made recently based on text analytics?

Is our text analytics engine built specifically for the hospitality industry? Dictionaries designed for general use instead of hospitality operations carry inherent bias, note the researchers – while a hotel-specific dictionary enables more precise analysis.

Can we influence category/topic modeling? Using out-of-the-box text analytics that automatically categorize keywords limits usefulness for running a hotel, as each operation is managed differently. You should be able to create your own categories.

Are we processing guest feedback by guest type? The more granular your analysis, the better you can serve each guest.

Are we processing guest feedback in multiple languages? The study’s authors note that multilingual text analytics hold the potential to “yield insights about cultural effects that can further aid hotel managers in improving their customer experiences.” This is especially crucial in global cities. Just processing feedback in English may not be enough.

Can we conduct meaningful longitudinal analysis on the evolution of sentiment over time? Knowing what’s different this year over last year – and what’s trending now – makes it easier to take action.