Integrating artificial intelligence and machine learning into hotel management provides opportunities to improve revenue-management strategy, staffing levels and more.
NASHVILLE, Tennessee—Consumers are sharing more data, and artificial intelligence and machine-learning technologies are increasingly evolving to better process that data.
The question is whether the hotel industry is using those tech tools to best capture the most revenue.
The reality is hotels continue to lag other industries in tech adoption, despite the tantalizing prospects of smarter operations, quicker revenue-management decisions and better personalized marketing.
Andrew Rubinacci, SVP of revenue and distribution at Omni Hotels & Resorts, said hoteliers have been burned on the promises of tech vendors before.
“We’re just unbelievably slow in capital investment technology, because these things take a lot of hard capital to build,” he said on the “Alexa! How should we fill our hotel next month? AI & ML: The ultimate disruptors” panel at the 2019 Hotel Data Conference last month.
“We’ve got to have that ROI there that’s believable,” he said. “There’s a lot of skepticism by a lot of owners out there on technology because, let’s face it, as an industry we’ve been sold a lot of technology over time that just has not panned out and doesn’t bode well for us investing more in technology.”
Mark Biondi, VP of business intelligence at Host Hotels & Resorts, said owners can’t prioritize tech upgrades over more pressing ones, such as property renovations, and that can put them behind the technology goals their brands might have.
“At the end of the day, it’s about where to deploy the capital investment that you need,” he said. “Everyone has alluded to reasons in the past why the hotel industry has been slow to update certain technology because the capital needs of the physical structure tend to trump everything else. There just needs to be a willingness on everyone’s part to recognize the value of the investment in technology.”
Benefits to doing business
Revenue-management systems are increasingly relying on smarter technology—including AI and machine-learning—to improve pricing recommendations. But Tina Meredith, VP of portfolio revenue strategy at PM Hotel Group, said most revenue managers agree their jobs are not at risk.
“The amount of data that can be looked at that we don’t necessarily have time, or really ever had time, to look at is incredible,” Meredith said.
“Maybe 15 years down the road, there could be a system that wouldn’t need human intervention. But I don’t see a case where you don’t need to intervene in some way, shape or form. There’s just so many things those systems can’t know. Yes, it picks up trends and it can do short-term rate changes for you. There are countless things I could cite examples of that this machine is just not going to know, and you really need that human interaction to help it along.”
Biondi noted any smart-pricing technology shouldn’t be fully independent from the guidance of a revenue manager.
“One of the things that’s important to understand: This is a tool, it isn’t meant to be a crystal ball,” he said. “What we’re looking for though is a tool that can be objective and that can perhaps challenge (our) preconceived notions. That’s where the value lies, when it’s providing something that’s maybe a slightly counterintuitive result and it really calls into question (our) thought process and hopefully it can lead to better decisions, because senior management is going to look at that and have to reassess a strategy that’s been long-held.”
Trusting the technology completely is still a long way off, Meredith said.
“There’s still some distrust out there; old habits die hard,” she said. “How could that system know more than I do? But by feeding the right information into it and interacting with it the right way, it can get to the point where you can trust the decisions it’s making.”
The future of operations and guest-facing technology
In the race to run a hotel portfolio more efficiently, Biondi said he’s interested in how AI and machine-learning technology could trim labor costs.
“All the data that’s being gathered by various property management systems, POS and hotel apps—how can that be leveraged to make operations more efficient?” he said. “If you think about something like check-in times, if you look at the history, you should be able to know when the majority of your check-ins occur and if that can be built into your labor management systems. It can really cause a kind of sea change of what you think of as traditional staffing levels at a hotel if you know that the majority of your check-ins are occurring in an hour-and-a-half- to two-hour window every day.”
There’s also room to improve how much hoteliers knows about its guests before they even check in to offer a personalized experience while maximizing ancillary revenue, Rubinacci said.
“I know you travel five to six times (per year), you go to city center, you want a king-size bed and a breakfast, and I can do that for you,” he said. “But … within five years, I want to know so much (more) about you as an individual. I (want to) know that every spring you go away for spring break with your family … you argue about it over Christmas, the first week in January you start to look for it. And I start showing you relevant data. … I might know you have an 8 p.m. flight out on your last day, and I offer you a late check-out at check-in, because that’s when it converts the most.”
Rubinacci added the data required for personalizing the hotel guest experience is definitely attainable.
“The goal would be get all of the relevant data in there so I can offer you exactly what you want when you want it with the highest probability to convert,” he said. “I’ll be able to extract as much revenue from you as possible, but it’s exactly what you want to buy, so you’ll be very happy doing it. All the data points that we have in there in order to understand a person that well, it’s out there, you just have to locate it and find it and then be able to use it.”