Ajay Aluri, a professor of hospitality and tourism management at West Virginia University, said he hopes to prove the potential of artificial intelligence in the hotel industry.
WASHINGTON—Artificial intelligence is no longer just the stuff of science fiction and speculation. Various applications of the technology—both business-oriented and consumer-facing—now exist, and one expert said it’s time for the hotel industry to take stock of AI’s potential.
Speaking at Hospitality Technology Next Generation’s North American Insight Summit, Ajay Aluri, an assistant professor of hospitality and tourism management at West Virginia University, said while the technology holds many possibilities, the hotel industry is still “in its infancy stage on how to use it.”
Here are some takeaways from Aluri’s presentation.
1. Accuracy improving
Aluri pointed out that the accuracy of consumer-facing, voice-activated AI technology is getting better and better.
He noted the accuracy of a device’s ability to both recognize a person asking a question and to answer that question correctly varies from device to device. But, he said, some devices are pretty impressive in that regard.
Google Home devices, in particular, are able to recognize 68% of questions and answer 91% of yes-or-no questions correctly. Of course, it helps that those devices are tied to the huge amount of data Google houses, Aluri said.
“There’s a reason why Google is able to do it so well,” he said. “It goes back to what’s in the secret sauce … and it goes back to their deep learning neural network.”
2. Boiling down machine learning
Aluri said hoteliers should be able to use AI and machine learning to help crunch data and ultimately improve the experience for guests. That might come back to tracking tangible things, like if rewards discounts are being used, to see if guests are taking advantage of their opportunities and are likely more fulfilled, he said.
With any sort of machine learning, the output has to be something digestible, Aluri said.
“One of the most important things is people need a simple way to understand it,” he said.
3. Historical data alone isn’t enough
AI systems can do a great job of digesting historical data and outputting models that predict exactly what will happen going forward, Aluri said, assuming nothing comes along to change consumer and guest behaviors. But that can be a pretty large and foolhardy assumption, meaning hoteliers and other businesspeople must learn how to adapt their systems to adjust for the unexpected.
“Most businesses start with historical data to create models, create strategies and all of that,” he said. “That’s great, but the part historical data doesn’t give you is how customers are evolving and what’s behind that. You need real-time data describing how guests are behaving at your hotel.”
4. AI can test the performance of initiatives
Aluri’s research team at WVU is currently working with a national company he couldn’t name to gauge the success of its loyalty platforms and whether discounting is having the intended effect of increasing engagement levels for what he described as season-ticket sales.
He said his team is working to achieve this by giving random discounts at a series of events and measuring the impacts at each. He said the results have been promising.
“So when we look at total engagement, it did seem to improve loyalty and value,” he said. “And we’re able to predict renewals.”
5. AI can assist with operations
One of the most intriguing applications in the hospitality space will be using artificial intelligence to better steer front-desk agents toward relevant conversations with guests, Aluri said. Automated systems can take care of the simplest requests, like answering questions about what hours the pool is open, while providing opportunities to cross-sell.
“You can also tell people about things they might find exciting,” he said. “So you can say, ‘The pool opens at 8 a.m., but we also have a Zumba class.’ Those type of cross-selling opportunities should be exciting for revenue management.”