Amadeus is experimenting with a branch of artificial intelligence called "reinforcement learning."
The distribution giant is looking at ways to employ the machine learning technique in the travel industry with use cases in revenue management already being trialled.
Rather than having to rely on huge data sets, reinforcement learning works within ongoing experimentation by testing and learning in a live environment with the system rewarded against specific KPIs.
Baptiste Chatrain, manager of data science at Amadeus, says the technique could also work for improving conversion on travel websites.
He says that travelers, at the inspiration stage of a trip, are currently presented with product recommendations that are the best option given the context.
But, by using AI, recommendations could go a step further in terms of personalization.
Subscribe to our newsletter below
Traditionally, travel companies would carry out A/B testing whereby a human would define the experiment with variations on content or functionality delivered to different groups.
He says: "A/B is a way of learning from the environment and making sure you improve conversion but it takes time to set up and it’s human experts setting up the variations.
"What we’re trying to do with continuous experimentation is have machines explore the set up of variations.
"With A/B testing you keep what works for the majority, but at the other end of the spectrum, you serve a version to each individual that comes to your site, that would be the ultimate personalization, unique to you and tailored to your needs but it’s not realistic as of today. So, it’s putting in an AI agent to get the best of both worlds.
"The idea is to use the sophistication of AI to have the machine design the best set of features or content you’re exposing or the recommendations you make. You don’t really need any of the data, you can learn from the environment.”
Disruptive forces
The initiative fits in with "improving conversion," one of six themes that Amadeus has identified as disruptive to travel alongside blockchain, extended content, messaging platforms, operations and performance.
In a report on traveler motivations, the company says travel companies need a deeper understanding of consumers, the emotions at play and the decisions they make, to improve the products it offers to them.
It draws inspiration from 2017 Nobel Prize winner Richard Thaler, an expert in cognitive bias, and stresses the importance of developing rules and algorithms that take into account the influence of psychological, emotional and social factors on consumer behavior.
AI will not only play a significant role the inspiration and shopping stage of travel but also in booking in terms of more dynamic pricing and offers, the in-trip experience with AI assistants and post-trip by analyzing reviews.
On top of AI and continous experimentation techniques, other technologies such as virtual reality and extreme search will also improve the travel research and booking experience in the future.
Chatrain says that current technology does not address that some travelers enjoy the research stage.
“One consumer might be decided and doesn’t want to scroll through options, and others, in the way they interact with tools, need that search element.
"I believe AI will be able to pick that up and suggest more tailored options if you’re the type that enjoys spending time finding the perfect vacation but the only way to know is to experiment."