An artificial intelligence specialist claims it can use machine learning to boost conversion rates for travel websites – and expects chatbots to become personalised travel experts in the near future.
Sentient Technologies uses AI to optimise travel companies’ websites, testing various combinations of minor changes for different users.
It uses data from its testing to see which combination of changes have the best effect on conversion rates.
Vice president Jeremy Miller said: “We really see AI as being able to help travel brands and online booking agents optimise their sites, optimise their bookings and optimise their loyalty programmes. Ultimately, create a better experience for their return visitors.
“We are seeing traction in the travel space because it’s a sophisticated market. AI ends up becoming a very attractive option. It’s extremely useful because it helps people cut through the competition and become user friendly websites that keep people coming back.”
Miller said that the process cannot be undertaken by humans because the tests cover all different potential set ups simultaneously, in every combination imaginable. The AI gets on with it without human interaction and reports back concluding which combination is performing best.
“Some of the brands we are working with today are using technology to improve basic things on their site, but across the entire booking journey,” Miller added.
Sentient says it can, typically, improve travel sellers’ websites conversion rates by between 1-2% which it says, depending on the size of the business, can equate to $100,000 a month.
“If you are looking at a website like Booking.com, it’s already very advanced and the changes may be very subtle,” Miller explained. “If we tested five different changes, it would give tens of thousands of possible combinations.
“If you were to try and do that manually, that would take years. With AI, it takes weeks.”
And instead of having to run the experiment, and implement it, as a human would do, Miller said AI could “implement it on the fly”. Not all traffic goes through the same test, meaning Sentient’s AI can compare and contrast to determine which is working well and not so well.
He said some customers choose to optimise earlier on in the experiments, which can run for as long as the company likes, while others wait longer to make changes based on more data. Between 30-40% of Sentient’s clients optimise within 7-8 weeks.
The value per transaction is not always huge, Miller admitted, but, like all data, its use is enhanced with scale – when Sentient says it can provide “significant gains”.
The company’s AI is also used with companies in the e-commerce and retail sectors, but Sentient says travel is a key market.
Miller added: “Travel is a big priority for us. Online travel agents are very savvy optimisation experts. They know they have to optimise their sites to capture more value.
“A lot of travel companies in general realise that they need loyalty programmes. Anything you can do to make the booking process easier and more efficient is a good thing. It’s not just about the site being more appealing, it’s got to be more usable too. That’s the differential. If your customers are coming to the site and transacting efficiently, that’s good news.”
And airlines are already cottoning on, he says. “The changes are the kind of changes you would expect. We’ve seen airlines change background or border colours and font sizes, or add a call to action.
“It’s these kind of changes, in combination, which make a significant impact. We would never be able to discover that unless you have a system that tests it.”
He suggested online travel agents use more loyalty programmes to boost customer retention, and said AI could help in that respect.
“Loyalty is the best retention scheme that OTAs have,” Miller said – and suggested they offer deals like a free night for every ten trips a customer has booked is an example of “finding different ways to entice customers to keep coming back”.
“Anybody that books travel online has an infinite number of choices,” he said. “Unless there is an extra offer, you are just going to shop on price every time.”
As for the future of AI in the travel industry, Miller believes it will soon become a ‘complete personal travel assistant’.
“We say AI has the opportunity of becoming a personal travel assistant. If you know your customer well, and know their travel habits.”
For example, Miller said AI could prompt notifications to customers if travel plans change, such as alternative hotel check out times or flight delays. And he believes the process “can get much more dramatic” with OTAs able to use machines to tell customers if there are better connecting flights through chat bots which can help customers rebook when their plans change.
“Travel is just like everything else,” he said. “People value service and good experiences and where you can provide that you get more return customers.”