Much has been written about how powerful and transformative artificial intelligence (AI) can be in the domain of airline pricing or optimizing the allocation of a given carrier’s seat inventory.
These are ideal tasks for a technology capable of providing actionable recommendations based on the analysis of very large data sets.
Not long ago we explained how tech firm Fetcherr has developed a large market model (LMM) which does just that. Much in the same way, if fed the right amount and type of data, AI-driven LMM models can also enhance an airline’s commercial capabilities in the customer-facing side of the business.
A new paradigm in airline-passenger interaction
Over the last couple of decades, the popularization of e-commerce has massively expanded the direct distribution capabilities of airlines and brought about a massive transformation of the industry’s commercial landscape.
Over the course of these years, airlines have got better at selling seats and ancillary services. However, the basic mechanics of the airline-passenger interaction have remained pretty much the same, with travelers interacting with form-like interfaces, which require considerable time and effort from the traveler.
Could AI be the game-changing technology many in the industry have been waiting for?
During the World Aviation Festival, which took place in Amsterdam in October 2024, Dominic Kennedy, Senior Vice President, Revenue Management, Distribution & Holidays at Virgin Atlantic, offered some clues about the future of airline-passenger interactions in the era of AI. The executive explained how the British airline is working with Fetcherr to bring deep, holistic transformation to the airline customer experience.
As one of Fetcherr’s early customers, Virgin Atlantic has already had the chance to experience the power of AI, and of Large Market Models in particular, in the area of pricing and has also started to evaluate the potential of this technology to transform the whole travel journey.
“We’ve been working with Fetcherr for a number of years. The relationship has been great from the start, but we just kept strengthening and deepening it. I think there are a number of reasons for that. First is the fact that Virgin Atlantic has long prided itself as a trailblazer and tried to do things differently,” Kennedy explained in conversation with AeroTime.
He mentioned some of the innovative products and services that gave Virgin Atlantic its enviable reputation, such as the introduction of Premium Economy or onboard bars.
“When we first started engaging with Fetcherr it was really a wonderful meeting of minds, because they were thinking very differently from everyone else about commercial airline planning. Initially this related only to the airline’s pricing function, but their vision is much bigger than that,” he added.
“We are in the process of rolling out Fetcherr’s Generative Pricing Engine (GPE) and we are looking with interest at what they’re doing in the inventory management space,” Kennedy explained. “What we’ve got with the generative pricing engine is effectively a fully automated pricing solution all the way from the conception of the price point to execution in the market.”
He continued: “Historically, we’d need an individual looking at competitive pricing, as well as considering our general pricing policy within a given market, and then finally pressing a button and filing that in the ATPCO database and seeing that through to being available to consumers.”
The automation of processes has transformed how Kennedy’s team performs, with analysts now focusing on more strategic decisions and helping steer the suggestions by the system.
Even a boldly innovative airline like Virgin Atlantic started the roll-out of this technology in a rather controlled way. The airline used one of its routes for experimentation, giving half of the flights on that route to the AI for optimization and doing the other half in the traditional way.
Fetcherr’s AI passed the test with flying colors and was then tested in more and more markets and compared to industry peers, again with positive results.
“There’s also the responsiveness you get from AI, by being constantly plugged into all that external data which sits on top of your own data. That means that you are not reliant anymore on a human spotting something in the market. We are responding in real time to demand fluctuations and other external factors,” Kennedy explained.
Spurred by a common desire to challenge the status quo, both firms have been cooperating in shaping some elements of Fetcherr’s suite of AI solutions, with Virgin Atlantic’s understanding of the airline business complementing Fetcherr’s own AI expertise which is rooted in the world of algorithmic trading.
A holistic approach
One area that appears to be ripe for disruption is the interaction between airline and customer.
From the point of view of the passenger, two essential elements of the AI-driven travel purchase experience stand out.
One is the ability of the system to interact with the customer through natural language, that is by expressing the needs in the same way and using the words and expressions we would use if talking to another human being.
The other is the capability of the system to provide answers and execute actions, including processing bookings and payments, in a multi-variate environment.
What does this mean?
Let’s say you are a Coldplay fan based in the United Kingdom and your favorite band is touring the United States next month, right when you have some days off. Wouldn’t this be a great opportunity to fly to the US to enjoy an unforgettable experience and do some sightseeing and shopping along the way?
Imagine, next, that you head to Virgin Atlantic’s website, and you tell the airline (or rather, the customer-facing AI engine that powers customers’ interactions) exactly what you need. For example:
“Hey, I would like to see Coldplay next month for their upcoming tour, but I don’t want to spend more than, let’s say $1,500 in this trip. What do you suggest?”
In this example, you would expect the system to return a shortlist of travel options either on Virgin Atlantic or on its partner airlines, that will take you to the US cities where Coldplay is playing and where it is still possible to book tickets for the concert.
You would then be able to engage in conversation with the system to put together all the elements of the trip. It may even be the system that proactively suggests things.
For example, it may suggest a shortlist of hotels that match your budget and preferences. Or it may arrange the most convenient transfer options. What about some restaurants near the concert venue?
You could, of course, search for all this on your own, but the AI’s tool is your time-saving machine here. And, for the airline, it’s a revenue-generating opportunity, as well. As you narrow down your choices, the system is also able to book them for you as it arranges your itinerary.
This frictionless model is not yet a reality, but it is a vision that Fetcherr is confident it will be able to deliver in the not-too-distant future.
“It’s a wonderful vision. We definitely believe that airline distribution is moving in a direction that aligns with that, an implementation path,” said Kennedy.
In addition to driving significant revenue increases for Virgin Atlantic, Fetcherr has been able to add some custom elements to its Large Market Model in order to match Virgin Atlantic’s unique profile as a purely long-haul, UK-based airline.
The vision is to have just one point of contact with the airline, which would be able to sell you everything you need to enjoy a trip to the fullest.
“We aspire to build an ecosystem in which people come to us to book a flight, and we are then able to offer them all the different permutations of that trip. This is something that definitely extends beyond travel in the traditional sense. This should be about the experience, about that excursion, sports event, or night at the theater. Our vision at Virgin Atlantic is to create compelling reasons for you to satisfy all those needs while staying within our ecosystem,” Kennedy said.
“The AI is going to know and understand your preferences, your needs, your behaviors and it will be able to draw on the hundreds of thousands of different permutations, not just flights but also hotels, clubs, excursions, experiences…whatever it might be that meets your requirements,” he added.
Kennedy concluded by highlighting the gradual and unstoppable nature of AI adoption.
“Once the AI understands the sorts of things you like to do and when you would like to do them, it won’t take long,” he said. “It will first be exploring your behavior, and it will take a little bit of time, but it won’t be long before it’s able to serve you exactly what you want to do on your next trip to San Francisco, Athens or wherever the journey takes you.”