Denver — Artificial intelligence could propel travel management past other disciplines on service innovation. According to experts, this is because of its data and people.
American Express Global Business Travel’s Evan Konwiser said business travel firms have the info to make AI useful. Others don’t. Among these data are user profiles, history, preferences and loyalty. Content, rates and negotiated contracts play a part. Big data concepts manage the interplay. Konwiser said all this will “drive incredible innovation in our space.”
He argued corporate travel providers and clients are well-positioned to take on the “Herculean” challenge. Co-panelists at the Global Business Travel Association convention here last month agreed.
In travel, “we kind of give ourselves a hard time about lagging behind,” said Sabre Labs director Mark McSpadden. “Here, travel is uniquely positioned. In consumer electronics, I have no humans to fall back on when that bot doesn’t do well. Travel, I think, will become the poster child for these types of interactions.”
McSpadden had outlined the interactions in a primer on AI. They boil down to three categories: at the point of booking, for service and support, and for merchandizing. He said Pana, Hyper, Lola and other startups exemplify the first category. The second category could appear on platforms like Facebook Messenger. FCM Travel Solutions recently released such a service. The third, said McSpadden, remains untapped because it requires critical mass on one of the first two.
The first category is about making requests “in natural language” by voice or text, McSpadden explained. “The interesting thing we’re seeing here is there’s a ton of context. They’re saying, ‘I’m going to a conference and I need a flight’ or, ‘I need to meet with three clients and do this.’ That’s significantly different from the amount of information we’re getting now from booking tools.”
Panel moderator Tony D’Astolfo of Deem chimed in on the status quo with corporate booking tools. “You go to the second screen to find purpose of trip,” he said.
“Yes, and that’s very general,” agreed McSpadden.
In the second category, itinerary management could adopt “conversational service and support.” Sabre Labs is working on a related proof of concept on Facebook Messenger. “I text in ‘Where is my hotel?’ and it gives me the hotel address,” McSpadden said. Sabre provides a leading itinerary app in TripCase.
McSpadden also went beyond these traveler use cases.
“We’re not spending enough time on this — AI will affect assistance for corporate travel managers as well. To let them know if they’re falling behind on obligations for contracts, or spending more in certain locations. The intelligence piece is a [matter of] when to bubble up information for you. It’s more than basic reporting. It’s knowing ‘I need to send the travel manager a message now.’ That’s the kind of intelligence humans have. We know when to bubble up information to a boss or a peer. That’s what we’re looking at these systems to do for corporate travel managers.”
AI is “in the immediate future,” said BCD Travel director of emerging technology Miriam Moscovici. Mentioning Evature, she noted that natural language parsing for online booking tools is not new and “hasn’t quite picked up.”
“I think we’ll see more of it in what we call bots,” she said. “We all use bots today. Any TMC has the mid-office. We automate things to happen based on the itinerary. You book to a dangerous place and we can send your itinerary to the security department, or remind you to check in with the CDC. I think you’ll first see bots reaching out to warn you on a security issue and engage you right there without having to get a human on the phone.”
Gant Travel president Patrick Linnihan said TMCs will soon exploit a combination.
“We track why people call us,” he said. “You will see AI inside TMCs delivering on that question it seems that 30 percent of people ask at the end of each month: ‘Can I have the invoice for my trip?’ Will it take over? No. But you’ll see the hybrid of the AI agent in the months ahead.”
Konwiser said the “dirty secret” behind the Panas and Hypers is that they use human agents.
“Because these channels exist and we have humans to fall back on, we’ll see applications have an impact in the near term,” Konwiser said. American Express GBT is spending “a lot” of money on this concept, which is “extremely important to the foundation of the business.”
Konwiser, GBT’s VP of digital traveler, continued.
“The real opportunity [is] to change the service experience and perhaps the call-center business model. Then we’ll leapfrog a lot of other industries that don’t have access to that kind of user data. The leisure travel industry [may] know one search you made on their website. They don’t know all your trips, why you want to travel, your preferred suppliers, your policy, your budgets. We have work to do, but the sky will be the limit.
“As it turns out, humans are pretty good at responding to chats. I have 8,000 agents in 40 countries around the world answering the phone, emails, instant messages, texts. Tapping into that network is a huge priority for us, and layering on AI.”
BCD Travel recently held a summit with purveyors of AI travel applications. Pana’s Devon Tivona said there that AI is in one of several historical “investment blooms,” Moscovici reported.
She suggested this bloom will sustain itself because the channel it uses, the smartphone, is ubiquitous.