How AI is changing travel companies’ spending habits

Travel companies are dedicating increased spend to artificial intelligence (AI), shifting capital strategies and pulling money out of software-as-a-service (SaaS) tools and outsourced development.

Trivago’s figures illustrate the scale of the shift: Last month, the metasearch engine revealed it spent more than five times as much on AI tokens and tools for its teams in the first seven months of 2026 than in all of 2025.

According to an internal Trivago survey, 93% of employees are using AI daily, up from 63% during the same period last year. And, 86% of respondents said the technology is making a “measurable difference” in their output.

Johannes Thomas, CEO of Trivago, said that right now, AI shows up across more than 100 teams because every talent has an AI budget—spend is following where value is being created.

“The increase comes down to two things, rising adoption among our talents and models that have become more powerful, letting us do today what wasn't possible before,” Thomas said at the time.

And Trivago isn’t alone: Withlocals, Faye, Priceline and Travelier all told PhocusWire they are scaling their investments internally.

The 2025 to 2026 shift

Over the past year, there has been a marked shift in how travel companies spend on AI.

The start of 2025 was all about making early bets, building prototypes, testing and learning about AI models, according to Daniel Green, co-founder and CTO of Faye, a PhocusWire Hot 25 Travel Startup for 2025.

That period spent gaining understanding lent itself to smaller-scale operations with beta users.

“2026 has been about scaling what works and rolling that out to all Faye-ers at the same time that our company rockets forward in growth,” Green said.

Withlocals’ internal AI spend has also seen a meaningful increase, totaling about two to three times what it was in 2025, according to CEO Matthijs Keij.

“[In 2025], our internal AI spend was basically a handful of seats,” Keij said. “It's a real increase, but from a small base, and it's still a modest absolute number for a company our size. The biggest change is more in how we work than how much we spend on AI tools.”

In 2026, the company has moved from most employees using AI as an enhanced search tool to running it inside workflows. 

Externally, the company’s fastest-growing investment is with platforms such as OpenAI, Google and Anthropic.

And the line is also blurring between internal and external investment, according to Keij.

“The same API access powers an internal agent that cleans up operational data and builds a feature a customer is using,” he said. “So we don't really budget them separately anymore.”

Keij added that spending on external help is now decreasing, as AI is enabling Withlocals to handle work they would have otherwise outsourced.

AI investment is also "absolutely increasing" at Priceline, specifically via its AI assistant, Penny, according to CTO Sejal Amin.

"You can see in the market we're doubling down on Penny, and we're investing in those capabilities," Amin said. "Our internal investment is increasing in the tools that are being made available generally to people. Our investment is also increasing in our developer experience."

Capital allocation

Travel companies are approaching AI spend based on their existing funding structure. 

For Withlocals, AI spend is being reallocated from other software costs. 

The company simply stopped paying for tools that only have one purpose, Keij said. Instead, they are building those tools internally. Some of that development has become possible with AI, even for non-engineers. 

“A lot of SaaS platforms are wrappers around one task with a nice design,” Keij said. “And a lot of what you're paying for was priced before building this in-house was possible.”

In a recent CEO Spotlight interview, Mews CEO Matthijs Welle also commented on this, noting that the hospitality technology specialist has a new team of "product builders" that are able to develop product iterations in-house faster and more affordably.

At Priceline, spend it being redirected as the company retires unnecessary tools and those that can be consolidated, Amin said.

But not all of the increased investment is coming from funding previously dedicated to other projects.

"Some of the funding is incremental," she said.

Meanwhile, as a younger company, Faye has been built with AI elements from the start. So, AI has been incorporated in the company’s “day one” budget, according to Green.

Five years later, Green said Faye is seeing two benefits from AI. First, he pointed to reduced reliance on investments in SaaS tools that can be built internally or replaced by capabilities provided by Claude, ChatGPT or Gemini. Second, he said AI has allowed Faye to scale its headcount at a slower pace.

“Often, one person can handle two to three times more work than they could even a year ago without additional strain or compromising on the customer experience,” Green said.

AI ROI

But as AI budgets ramp up, how are travel companies assessing the return on investment?

It’s difficult to calculate what Keij called a “clean ROI” for AI, especially considering that AI tooling is involved with everything the company is doing.

He’s looking at three factors right now when gauging ROI. First is throughput per person, or whether Withlocals is shipping things it wouldn’t have without AI.

“That's the clearest signal, and it shows up as work that would never have made it onto a roadmap because it was not worth a quarter of development time or hidden in a major backlog,” Keij said.

Withlocals is also considering how long it takes for ideas to turn into live products and commercial impact.

Faye looks at AI ROI in terms of operational efficiency and customer experience, Green said. The company also sees returns in how quickly the company is able to improve or build insurance products.

“AI can help us create customized products for partners in hours rather than the months it can take incumbents, while improving underwriting, pricing and personalization by learning from claims outcomes, traveler behavior and evolving risk patterns,” Green said.

Where is AI spend headed?

As travel brands' AI usage increases, the mindset around spending will continue to evolve.

Green anticipates that companies will begin thinking about AI spend as part of the total cost of an employee and part of the cost of a sale.

“Our company will likely think of AI being both a significant part of our spend (in the same way we think of the cost of salaries or rent) and something that will appear throughout our budget,” Green said.

Keij said he expects the total spend on technology overall to look similar a few years from now, but the composition of that investment will be unrecognizable.

“On one hand, usage grows very fast,” he said. “More agents, more automated work and more of it running completely automated, without anyone asking. At the same time, the cost per capability keeps going down. What is expensive today is close to free in 18 months. So they cancel each other out, unless the entire AI market proves to be completely unsustainable, but that's a different topic.”

Keij also expects that the industry will decrease spend on software that “performs a task” versus “capability to perform a task,” and the AI budget will no longer exist as a line item.

“It will just be how the software and our platform works.”