If you’ve ever consulted the web for trip tips, you’ve probably arrived Expedia site. The travel site offers useful tools for travelers such as travel recommendations, smart shopping, and price tracking to ensure you get the best value for your money. You may be shocked to learn that the recommendations are not actually human-made.
“Whether it’s providing insights to our travelers, or driving better personalized experiences for both travelers’ needs as well as those of partners, everything — from search and sorting to our fraud-pushing and risk-protection capabilities — is powered by artificial intelligence and machine learning across the board,” says Rathi Murthy, President of Technology division of Expedia Group.
It turns out that an AI-driven algorithm knows more about your travel preferences than a human ever does.
“There’s really a lot of complexity for the human eye to figure out what our algorithms can do,” says Murthy. “There are many cues from the traveler’s side, as well as the partner side that our models can pick up on.”
The Expedia platform processes more than 600 billion AI predictions annually, backed by 70 petabytes of data. To put the size of this amount of data into perspective, one petabyte equals 20 million file cabinets long, according to terata.
The data that Expedia collects to learn about your personal preferences comes from your profile, your choices, and most importantly your behavior patterns. The company also uses supplier data to help give you the most useful and accurate recommendations.
“We have a lot of data that we can run to drive these experiences for travelers and partners in a way that no one else can,” Murthy says. “And all we’re doing today is really leveraging that data, artificial intelligence and machine learning.”
To transform this data into products and services and develop ML models, Expedia has more than 6000 software developers, scientists, and data designers.
Essentially, every keystroke and mouseover on the site helps Expedia better understand your preferences and offers features like price tracking, forecasts and one of its most recent launches, guest experience results.
The Guest Experience results were revealed in May 2022, and they take cues from customer reviews, experiences and more to give hotel properties a score. This result leads to how the properties are presented to the site visitors and highlight the property.
“This takes into account a range of signals, such as our ratings and reviews for that particular site,” Murthy says. “So a combination of machine learning and math models helps us figure out how good this particular resource is, so we can give travelers more confidence as they book.”
In addition to using artificial intelligence and machine learning to deliver better experiences to its customers, Expedia has also been able to harness the power of artificial intelligence to reduce business costs.
A big part of Expedia’s platform is being able to provide customer service to its clients who may have problems with their travels and need immediate assistance. This was especially true when the pandemic first emerged and people had to cancel flights, flights, or see if they were able to travel.
“During COVID times, we were getting a massive onslaught of call center calls with flights canceled, etc. So we took advantage of this moment to build AI/machine learning capabilities,” Murthy says.
To improve conversation capabilities, Expedia Group is running more than 29 million virtual conversations. This implementation saved the company more than 40% of the agent’s variable costs.
“We’ve already seen multi-million dollar savings just by leveraging automation and artificial intelligence/machine learning on the call center front,” Murthy says.
Concerns about these types of chats usually come down to whether they are effective and actually solve the problem. However, 60% of travelers self-serve using Expedia’s chatbot tool, and the more people use it, the smarter and more useful it gets. This applies to all company algorithms.
“The more you use us, the more you tell us about you, the better we become with our models,” says Murthy. “We have a continuous learning intelligence service that makes our AI and machine learning models better and smarter every day, so we can continue to personalize your experiences.”