Booking.com also says the number of homes, apartments, and places it lists is 27 percent higher than last year. The announcement is part of the company’s switch to “reported listings” — from “total properties” — in an effort to emphasize that it competes directly with Airbnb. Until now, Booking.com has been perceived as mainly offering hotel listings, a perception it is aggressively trying to change.
Booking.com is the flagship of Booking Holdings, the largest travel company in the world, by market value.
Bot resolves half of all questions within 5 minutes
Separately, Booking.com said it has hit another milestone with its use of AI to improve customer engagement. The company says its Booking Assistant bot now resolves half of the thousands of daily customer support queries it receives within five minutes. That’s up from 25 percent in July and 30 percent in December, when the company launched the bot out of beta.
The advancement is proof that Booking’s AI efforts are paying off, said James Waters, VP of customer service, in an interview with VentureBeat about the bot’s advancements. Five minutes may sound like a long time for resolving queries with AI, but travel questions can be complex and varied.
Indeed, the task of building out the bot comes as travelers and hosts are finding it necessary to share an increasing amount of information, often without speaking the same language. And now that Booking.com is locked in a battle with upstart Airbnb in the accommodations market, the bot could prove a valuable addition by augmenting customer service.
It’s too early to say whether the Booking.com bot itself is driving increased booking revenue because it only deals with questions that arise after the booking process. For example, travelers may inquire about early check-ins, whether they can bring a pet, or the size of a bed.
It’s difficult to trace and attribute business that directly results from a bot experience: If a user comes back, are they doing so because of the great bot, a competitive price, or for some other reason? But the bot data looks good. When a user of the Booking Assistant clicks “I’m done,” the bot asks them “Was this helpful?” and 85 percent say yes.
The bot has responded to more than a million queries so far, Waters said, and the company plans to increase access to 100 percent of English-language bookings this year. Booking says it has more than 30 million unique users, but it does’t break that down between English and non-English users.
The company’s AI push stems from experimental culture
Booking.com is the flagship property of Booking Holdings, which also owns Priceline.com, Kayak, Rentalcars.com, and OpenTable. Valued at a $100 billion market cap, Booking Holdings has seen explosive growth that has doubled over the past two years. It is many times bigger than the next largest travel competitors, Expedia ($16 billion) and Trip Advisor ($6 billion). And the company increasingly views private company Airbnb as a competitor, especially in the U.S., where Booking.com is trying to boost its profile. Airbnb does not have a comparable bot for users, instead relying on Messenger to help hosts and travelers communicate. Third-party company Burner also offers a bot called Hostbot for Airbnb hosts and users, but it’s not exclusive to Airbnb.
[Update: And the gloves are off. An Airbnb source emailed us after this story was published, with a statement: “Airbnb is growing faster and spending less because we have a community model as opposed to Booking’s commodity model.” When asked to elaborate, the source explained that Airbnb’s business model prioritizes hosts, for example charging them less of a fee than it does for users. Booking.com’s fees are the other way round: It takes a bigger cut from from hosts, none from users.]
As part of a new branding push, last month Booking.com’s parent company changed its name from the Priceline Group to Booking Holdings.
The company has attributed its success to its data-driven culture, which prioritizes rapid iteration through constant testing and optimization — done in small, autonomous units. Booking.com has more than 100 data scientists working on things like messaging, customer service, and other areas of product development. Many of them are in Amsterdam, but the company has opened centers in Tel Aviv and Shanghai.
In 3 years, half of bookings will involve NLP
Booking.com CEO Gillian Tans, in an interview with VentureBeat during a visit to San Francisco in February, forecast that within three years, at least half of all bookings on the site will involve AI in some way. By that, she means advanced forms of AI, including natural language processing, not just basic machine learning. Machine learning is already used at every level of the site, including every search result, she said. “This is moving faster than we thought,” she said of the number of users interacting with the bot.
Indeed, Booking is continually making changes to improve the bot product — conducting over 1,000 A/B tests a day across its offerings, Waters said. The video above demonstrates the bot’s hybrid approach, which uses natural language processing to understand open-ended questions but also offers menu cards with common topics and frequently asked questions and answers to help guide the user.
When it looks like the bot won’t be able to answer a question quickly, Booking routes the query to either its customer service agents or the host property — depending on who the bot thinks is best placed to answer. For example, a traveler who has already checked in might ask about how to restart a hotel’s internet router. The bot may decide to route the question immediately to the hotel owner, since it is so specific to the property.
Notably, the bot now answers about half of all queries by itself, without having to escalate the query, Waters said. And regardless of whether it escalates or not, it can now answer half of all questions within five minutes.
The bot can handle 90 specific subtopics, including everything from payment to transportation, arrival and departure times, date changes, cancellation requests, parking information, extra bed requests, pet policies, and internet availability.
Will a move to messaging drive business?
The bot uses Active Learning, a semi-supervised machine learning approach that allows the algorithm to learn from interactions with the user, says Adrienne Enggist, Booking’s director of product for the bot. The company built the logic and AI entirely in-house because the range of post-booking travel topics the bot deals with is narrow enough that Booking remains the expert. As the bot project expands, at some point Booking.com may experiment with other technologies to accelerate efforts, Enggist said.
More fundamentally, the company is testing whether the messaging format can drive business by consolidating user interactions into one easy, immediate medium. Text is less cumbersome than phone or email, and it offers a personable, human touch if the service has a routing feature like Booking’s. “It’s no secret that more and more consumers are open to (and very interested in) interacting with brands via chat,” said Waters. “For us, it’s also about leveraging the immediacy and simplicity of conversational UI mixed with the ability to tap into huge amounts of information automatically — be it product data, recommendations, help,” he said.
Booking was early to experiment with AI and bots. The chatbot revolution hit the travel industry first, driven by the many questions travelers have in the process of planning trips. But even within the industry, Booking has been a trailblazer. It launched the bot in testing mode two years ago and has experimented with other offerings, such as Destination Experiences and Passion Search, launching these efforts before competitors like Expedia and TripAdvisor joined the fray.