Adam Jay, the president of Hotels.com, an Expedia Group brand, is responsible for overseeing innovation, strategic development and growth of the brand globally. Prior to Hotels.com Jay worked in travel tech at Travelport and Avis Europe (now Avis Budget Group). Jay, who began his career at Boston Consulting Group, has a degree in politics, philosophy and economics from Oxford University and an MBA from INSEAD. The Innovator’s Editor-in-Chief Jennifer L. Schenker is scheduled to do a fireside chat with Jay at the March 6 Enterprise Innovation Forum in London organized by Next World Capital.
Q: Expedia and Hotels.com are both disruptors in their own right. What do disruptors do when they are disrupted by newcomers like Airbnb?
AJ: I have been working in travel tech for 15 years now and so much has changed in that time. Suppliers (airlines, hotel chains, car rental) have moved online; online travel agents like Expedia and Hotels.com have grown up; there has been an explosion in reviews; metasearch emerged quickly as a disruptive force; self-serve corporate booking tools like Egencia were launched; Google has become a huge player in travel; and of course, Airbnb appeared and rapidly created a new category. Its pretty awesome to work in such a dynamic space.
So how do incumbents like us react in a world of constant disruption?First, we need to make sure our target customers know about Hotels.com, love using Hotels.com, and want to come back again and again.Second, where we see opportunity to better serve our customers through a disruptive business model or disruptive technology we often jump right in. This can be through organic investment, like the investments we are making in machine learning and chat bots, and sometimes it is through acquisition. We purchased HomeAway a few years ago to give us a leading position in the rapidly growing vacation rental space, and before that we took a majority share in Trivago, recognizing the disruptive power of metasearch. We have bought a bunch of smaller players too where it helps fill gaps.Third, we are always on the lookout for how Expedia Group assets can better serve the evolving travel ecosystem. We have a very successful travel wholesale division called Expedia Partner Solutions which itself powers many of the so-called ‘disruptors’.
Q: How is Hotel.com making use of machine learning and AI to determine which customers are the most valuable and improve customer experience?
AJ: Everywhere! We even have an internal program called ‘Data Science Everywhere’, where we will progressively replace any rules-based code with machine learning. Last year we launched a very popular ‘Data Science Academy’ so that the whole company can become fluent in machine learning techniques and application.
Machine learning lies at the heart of building an ever more compelling experience for travelers and supply partners. Our marketplace is centered on a complex matching problem. We need to make it easy for travelers to find the right room, in the right property, at the right price, and of course in the right location. And the answer to all of that varies search to search, customer to customer, and trip type to type. Perfect grounds for machine learning.
There are so many applications for machine learning in any business. In terms of customer experience ML drives image display, prioritizing content in real time, and training our chat bots. On the marketing side we use machine learning for channel attribution, and it defines which customers we should target, and what we should bid in performance channels. It also helps with many other internal processes like issue detection and financial forecasting.
But getting machine learning right is not easy. A badly trained model can leave a real mess. The key is to be super clear about the objective you are trying to achieve. You need data that is relevant, clean and organized. You need skilled data scientists. And of course, you need to be able to act on the data and insights that are generated. If you manage an established business (like Hotels.com) you need an active program to drive machine learning into every corner of your business.
Q: Personalization is a big trend. How are you personalizing service for your customers and how do you see personalized travel services evolving in future ?
AJ: I would say we pretty good on core elements of personalization. We use customer click and booking history to provide relevant content and recommendations. We make it easy for users pick up on a previous search, including those made on other devices. Customers can find everything they need to manage their bookings or their free Rewards nights through their online account. In our app we offer a Concierge service where we help customers with anything they might need pre-stay or in-stay.
But we are far from ‘done’ and we will continue investing heavily in all areas of personalization. As one example, we are investing in chat bot technology to serve customer needs in a relevant and personalized way. And this week I have been with our APAC team looking at what we are doing to better serve local needs of customers across the region. It always amazes me how different everything is in Asia. I tried cheese tea for the first time.
Q: How will voice-activated services change the customer experience ? What should businesses be doing to adapt ?
AJ: Voice will provide new ways for customers to interact with businesses in two ways. First, via voice specific access points like Alexa, Google Assistant, and Siri. These ecosystems are still early stage, but we want to make sure that Hotels.com is in frame whenever a traveler is thinking about a trip, booking a trip, or needing support with an existing booking. Second, voice is a new input mechanism for existing web and app products like Hotels.com. Some devices have voice as the only way to respond, which puts limitations on the type of travel transactions that are possible, but increasingly voice devices return images, text and video, which starts to get super interesting for us.
How should your business adapt? Make sure you understand your customer needs and how voice interaction can help solve them. Capture the voice inputs you are already getting and learn from them. If you have a call center, what do customers call about? How do they ask questions? This is great data that you (and your voice bots) can learn from. And don’t be afraid — test and learn. There are lots of tools out there now which enable you to get basic voice capabilities live easier than ever before. Put services live that address your customer’s problems and see how they work.
Q: What advice would you give to others in the room about anticipating disruption and adapting to market changes?
AJ: Pretty obvious, but watch the dynamics in your market, stay close to the start-up world, and understand where investment capital is flowing. As a global business playing in every part of the travel ecosystem Expedia Group is lucky that we get an early view on evolving B2B and B2C dynamics across the world. Our corporate dev and commercial teams get to see dozens of business every week. We also run an accelerator program which gives us more direct access to work with early stage businesses.
Understanding and anticipating disruption is one thing, but much harder is deciding what to react to. Is it a fad? Will it scale? Does it help our target customer? Does it help our supply partners? How will it impact economics through the value chain? What is the best way to test? These are often not easy questions, and you need a coalition of strategy, analytics, product, tech and user insight to figure out priorities in the context of your overall strategy.
At Expedia Group we are lucky to have thousands of engineers and a wide portfolio of brands through which to experiment, but we still have to very thoughtful on where to place our bets.
Once you decide where you want to focus then it is simply an execution problem. Carve out focused teams, have clear KPIs, minimize dependencies, then, execute, execute, execute.
We are constantly investing to create an ever more open and adaptable technology, data and machine learning platform. Well architected and open source code allows us to nimble in reacting to new opportunities. Nothing stands still in travel distribution.