the OTA-led Defence Against an AI Takeover

It’s no secret that the AI revolution has arrived, and the online hotel distribution sector has already adapted to the first ripple effects.

But OpenAI’s Operator, launched earlier this year, revived concerns about the lifespan of long-standing industry players in the wake of new technologies.

The AI agent not only reads data, but can also interact with a browser.

This means in addition to generating travel recommendations and plans, Operator can make bookings, only asking users to take over for tasks that require login and payment details, or to solve CAPTCHAs.

Adding to the momentum, Perplexity recently partnered with fintech platform Selfbook and Tripadvisor to bring hotel bookings directly into its AI-powered ‘answer engine’ – another clear sign that agentic AI is on the rise.

Currently, Online Travel Agents (OTAs) are still the biggest drivers of hotel booking revenue, having withstood and adjusted to decades of technological advances – including mobile, a revolution that many thought would leave OTAs behind and instead has seen them use to further their competitive advantage.

Yet new tools like Operator and Perplexity raise questions.

Does the advent of AI agents signal the beginning of the end for third-party bookings, or are we simply entering a new phase in the evolution of online hotel distribution? And, how can other key industry players, particularly hotels, learn from the action OTAs have already taken?

Glimpse into the AI-led future

As with any industry where multi-layered intelligence is key to achieving the ultimate goal of personalisation, data is king in travel.

For this reason, revenue managers and OTAs analyse a range of past, present and future factors to sell the right rooms at the right times, at the right prices, to the right guests.

So when looking for their next stay, today’s guests are offered broadly personalised options largely based on the type of traveller profile they fall into, according to their demographic’s general preferences.

AI tools like Operator could take things a step further and tailor accommodation offers to suit each individual guest, thanks to the large pool of data at AI’s disposal and the technology’s ability to analyse and act on all of that information quickly.

Even with this potential, Simone Puorto, founder of consulting firm Travel Singularity and AI-powered start-up Rebyū, does not believe AI tools will make OTAs obsolete.

But he can imagine a world where OTAs transform to become data hubs from which AI agents would draw information, leading the hotel industry into a new era of “hyper-personalisation”, led by both OTAs and revenue managers.

“With more information, what you could do is create a rate for [Guest A], then create a rate for [Guest B],” Puorto says. “This will become more than revenue management. I think it will become traveller management – you will really optimise for that specific person rather than just try to create boxes [to sort them into].”

In addition to guest profiles splintering from demographics into individuals, Puorto expects an upcoming explosion of AI-powered tools will further fragment the travel industry’s tech stack offerings in the short-term.

This would compound pre-existing industry issues, with hotels already facing the challenge of trying to connect data across a throng of sometimes-incompatible OTA and other third-party systems.

But this won’t be the case forever.

Puorto predicts the natural accumulation of closures and mergers of AI tools will eventually result in a more centralised travel tech ecosystem.

Founder and CEO of Catala Consulting, Thibault Catala, expects the technologies could even morph into a single form of AI.

“I do believe the opportunity in the next few years will not be about fragmentation, but centralisation, of data,” Catala says. “And AI will allow this because that would be the only [thing able] to handle that amount of data and to be able to identify those different systems, which humanly is currently not possible.”

Slow on the tech uptake

Puorto points out we already live in a hyper-personalised world.

Think of Netflix or Amazon; no two homepages are exactly alike, as the companies target customers with suggestions for products that will interest them most as individuals rather than as members of demographic groups.

But apart from OTAs, which have proven to remain at the forefront of new tech adoption and increased personalisation, the rest of the travel industry lags behind.

Catala estimates the vast majority of hotels, particularly those on the smaller scale of the market, are unaware of the tools and technologies already at their disposal.

“I’m blown away every day by the state of the technology in the hospitality industry,” he says. “Even people reading [this] blog are quite advanced, because they are aware of revenue management.

“To talk about a system, to talk about connection, AI and so on, that seems super-duper advanced compared to the real state of the market. So for me, the biggest opportunity in the next few years, it’s education; increasing awareness of tools, best practices and so on.”

Optimising for AI

As travel hurtles towards an AI-centric environment, OTAs are working fast to accommodate the technology and have a significant head start.

So, how can they improve even further, and how can hotels catch up? First, consider all options with an open mind.

Digital transformation consultant Howard Phung says: “Many hotels, especially traditional operators, are risk-averse and reluctant to cede control to AI, particularly in areas like pricing and guest interactions.”

Compounding the reluctance to move on with new technology is the practical difficulty of adopting that technology, with Phung, like Puorto, emphasising the fragmented state of the industry.

“Many hotels still operate on legacy systems that don’t easily integrate with modern AI solutions,” he says. “Unlike industries with standardised tech stacks, hospitality tech is fragmented, requiring custom APIs and middleware for AI adoption.”

“However, AI doesn’t need a perfect ecosystem to be impactful. It will adapt to the fragmentation by working within existing tools, leveraging cloud-based integrations, and offering modular AI capabilities that don’t require a complete tech overhaul,” Phung adds.

To smooth the road for AI, an Agent Experience (AX) website audit is in order, according to Puorto. He says with the tech moving from search engines to “do engines” as AI agents are now capable of acting on a guest’s behalf, there needs to be greater understanding of how these agents work in order to optimise websites as necessary. Beyond OTAs, this is especially relevant for hotel websites, which remain a significant source of direct booking revenue for properties.

It follows, then, that AX Optimisation (AXO) may eventually surpass SEO in importance.

Catala says it was enough to be among the first 10 results of a Google search page in the past. Looking ahead, an AI agent may only present guests with a single recommendation, customised according to its in-depth knowledge of the guest.

Luckily, AXO and SEO share many common values: just as broken links are not good for SEO, neither are they for AXO, Puorto explains. Likewise, meta titles and descriptions are equally important for human user and AI agent experiences.

But some adjustments still need to be made for both hotel and OTA websites.

For example, Puorto says the use of CAPTCHA may need to be reviewed, as AI agents cannot solve the tasks. Agents also dislike pre-contracted static rates, instead preferring real-time inventory.

And although minimalistic website design is trending right now, AI agents “love to read”, Puorto says.

“It could be a good idea to start getting a little more text on your website,” he says. “I was completely against that before. Nobody’s reading blog pages on hotel websites… but [AI] agents, yeah, why not?

Puorto adds, “Probably we will see a return to a little more ‘90s design, just because [AI] agents, they don’t freaking care if it’s beautifully designed. They want the information.”

Travellers still prefer a human touch

Hotels do not adopt shiny new technology just based on trends, Phung says; they adopt it when it has proven to solve a problem.

In terms of AI, he says hotels are seeking tools that enhance efficiency, reduce costs, and improve the guest experience.

“The industry sentiment is cautiously optimistic – hotels are open to AI but want to ensure it is secure, scalable, and easy to integrate without disrupting operations,” Phung says.

Whether or not hotels adopt up-and-coming tools, the strength of AI’s future grip on online hotel distribution will be determined by the extent to which guests are willing to abandon older technology and platforms to embrace the new.

And this is where OTAs have the advantage.

In 2024, SiteMinder research found OTAs were still the biggest drivers of total revenue for hotels. SiteMinder also found travellers are hesitant to adopt a fully AI-centric hotel model, with only 1 in 10 agreeing the technology should do it all.

OTAs have proven to be resilient and adaptable to decades of technological advances.

“Booking a hotel in 2002 and booking a hotel now, it’s not that different if you want to do it online,” Puorto says.

“Nothing big really changed.”

Much of the enduring popularity of OTAs can be attributed to the simplicity of browsing and booking on their websites. And instead of a favourite hotel or brand, Catala says many travellers are instead loyal to a particular OTA.

But that does not mean accommodation providers can rely solely on OTAs to gain them new and repeat guests.

“That’s very important to think about the way people find [out] about us, but at the end of the day, the relationship, the experience, the connection … that’s human-to-human,” Catala says.

“That’s what will differentiate [you from] someone who is relying only on OTAs and doesn’t care about the guest.

“But that’s up to you to make sure you go to the check in, you go to the reception and you connect with your loyal guests. You talk to them, you have your own communication program, marketing activities to reach out to those guys.”

Evolve now to avoid to avoid future extinction

While AI agents like Operator could theoretically bypass OTAs by offering hyper-personalised booking experiences, Phung, like Puorto, does not see OTAs disappearing anytime soon.

Instead, he says, AI will likely strengthen OTAs in several ways.

“Rather than being a death knell for OTAs, AI is more likely to encourage them to become more innovative, faster, and more automated distribution channels,” Phung says.

“The future isn’t about OTAs disappearing – it’s about AI-powered travel experiences that blur the lines between direct and third-party bookings.”

But in order to be strengthened, not eaten, by AI, Catala says the technology needs to be embraced correctly by OTAs and revenue managers.

He warns those who only focus on the old status quo may disappear, while those who evolve and build more diverse expertise by experimenting with new technologies will stay competitive.

Several major OTAs are fully onboard and already years into investing in an AI-led future. Companies like Booking.com and Priceline have partnered with Operator. They, alongside other OTAs like Kayak, Expedia and Ixigo, have developed onsite AI programs which further personalise services for customers – something most hotels cannot afford to do at the same level.

Meanwhile, revenue managers could be in a more precarious position than OTAs; Catala says some people in the industry expect the role of revenue manager could disappear in the next five years, either being replaced by AI or integrated into other positions.

How revenue managers adapt to new technology will mean everything for their future.

Just like OTAs, Puorto says it could serve revenue managers well to start working hand-in-hand with AI agents. He predicts revenue managers who have a good understanding of the market could become sought-after as consultants and trainers for AI agents.

But even in the age of AI, it is important to remember human connection goes a long way.

“Those kinds of tasks [like reporting on pricing] can be automated by AI with more accuracy,” Catala says.

“But the strategic elements, the experimentation elements, the influence, the communication, that’s something which AI cannot do.

“We are still operating not in [business-to-business], but in [human-to-human].”

This article is part of Revenue Management: A Deeper Conversation, SiteMinder’s thought leadership blog series which harnesses the knowledge of hotel industry experts to spark real conversations and elevate discussions surrounding the practice.

About SiteMinder

SiteMinder Limited (ASX:SDR) is the name behind SiteMinder, the world’s leading hotel distribution and revenue platform, and Little Hotelier, an all-in-one hotel management software that makes the lives of small accommodation providers easier. The global company is headquartered in Sydney with offices in Bangalore, Bangkok, Barcelona, Berlin, Dallas, Galway, London, Manila and Mexico City. Through its technology and the largest partner ecosystem in the global hotel industry, SiteMinder generates more than 125 million reservations worth over US$50 billion in revenue for its hotel customers each year. For more information, visit siteminder.com.