Why Google can’t go all in on AI travel just yet

So, it wouldn’t have come as any great surprise when Google began ramping up its AI-powered trip planning via Gemini. Google has had an on/off relationship with travel over the years and now it seems things are very much on again – only this time, powered by AI.

Could this be the start of Google becoming a full-blown OTA? And if so, what would that mean for the companies spending huge amounts of money advertising on its platform every year?

The short TL:DR answer is: it probably won’t be a full-blown OTA anytime soon. Not while it could jeopardise its precious ad revenue cash cow.

But let’s look into it more closely.

A relationship worth preserving

Travel has always been one of Google’s most valuable sectors. In 2018, it was estimated that travel accounted for around 12% of its total ad revenue. That number might be a tad lower today in percentage terms, given growth in other categories such as retail and finance, but travel spend in absolute terms has certainly grown.

Last year, the ‘Big Four’ of online travel – Expedia group, Booking Holdings, Airbnb and Trip.com – spent a combined $18 billion globally on marketing. Unsurprisingly, a hefty chunk of that went directly into Google’s coffers. If we assume Google retains its typical share of approximately 39% of digital marketing digital ad market, that’s somewhere between $7 billion and $8 billion – or 5% of its TOTAL GLOBAL REVENUE from those four companies alone.

And that’s before we include the rest of the travel industry, i.e. mid-sized OTAs, tour operators, cruise specialists and the long line of suppliers that rely on Google Ads for discovery and direct bookings.

It’s understandable that any move which diverts users from clicking on paid ads would go down like a lead balloon with the brands that rely on Google to drive traffic and conversions. And let’s say travel companies pulled 30-40% of their spend, that would be a 5% drop in overall revenue - or the equivalent of Google’s own Brexit.

Would Expedia or Booking immediately turn off their spend? I doubt it – Google’s reach is too valuable. But they could reduce budgets, redirect spend to other channels or double down on direct brand building, resulting in a sizable dent to Google’s revenue.

It wouldn’t just be about the Big Four either. Smaller brands already feel squeezed by increased competition and rising CPCs. If these businesses feel that Google is going to become a competitor rather than a marketing channel, they may pull their spend too.

Would Google work as an OTA?

Google has been gradually bringing everything within its own ecosystem for years. It already offers flight and hotel search tools and could feasibly push its own travel packages by pulling together itineraries via Gemini and then allowing users to purchase directly through Google Pay. The experience would likely be smooth and familiar to the end user. And while many may moan about Google’s global dominance, it's that lure of familiarity and simplicity that keeps people coming back.

But travel is a complex beast. Unlike buying cinema tickets or a new jumper, travel purchases are typically higher value, multi-step and full of variables. There’s a lot more at stake. When I think about booking travel, the following tweet always springs to mind:

While the conversation might've moved on from mobile v laptop (not for me – still won’t spend more than £100 on mobile purchases), it’s likely that booking an entire holiday via AI might face a greater level of resistance, for longer. Perhaps not for younger, more tech-savvy users, but certainly for the wider market (and let’s not forget it’s the older demographics who have the greatest spending power). They want/expect to see recognisable branding, read detailed terms and conditions, compare prices, check reviews and have someone at the end of a phone/email if something goes wrong.

Trust, in this context, is about more than a slick, familiar user experience.

Snared in red tape

Perhaps the biggest reason Google won’t rush headlong into direct travel sales is that it would instantly find itself trapped in a web of consumer protection law. The UK’s Package Travel and Linked Travel Arrangements Regulations 2018 state that any company organising and selling package holidays is responsible for:

• Ensuring the performance of those services

• Providing assistance if something goes wrong

• Offering full refunds in certain cancellation scenarios

• Protecting customers’ money through financial safeguards

Nick Parkinson from Techlaw, who advises travel businesses on exactly these issues, explained more about the challenges Google would face:

“If they (Google) are selling a package (say flight + accommodation / a cruise) in the UK/EU then:

• They become liable for any accidents/claims which are the fault of the individual supplier (usually the hotel but it could be an airline or excursion provider)

• They need to have insolvency protection to protect customer monies (specific / limited options available)

• Customers would have a high level of refund rights in the event of changes or cancellations (including the customer wanting to cancel in certain scenarios)

For less complex processes such as single destination bookings, there are still challenges if payment was taken directly:

“It is correct that selling ‘just a flight or hotel’ does not make a package but, in that case, we still have to determine whether Gemini are acting as agent or the contracting party. If they are acting as agent, they have no real liability or obligations. If they are acting as principal, however, there are other UK/EU consumer laws (e.g. the Consumer Rights Act) for Gemini to consider which could result in: (a) liability for any accidents/claims which are the fault of the individual supplier, and (b) refunds in the event of changes or cancellations”.

The obligations are extensive, and they exist to protect consumers from the kind of chaos that happens when intermediaries go silent. As Nick points out “It’s not just about what’s being sold, but how. Gemini either acts as an agent (with minimal liability) or as the provider (with all the risk). There’s no middle ground”.

Competition red flags

If Gemini starts favouring Google’s travel products over that of third-party providers, this could step on the toes of competition laws, similar to when Microsoft bundled IE into Windows as the default and primary browser. At a time when Google is already under scrutiny in the UK over search dominance, it wouldn’t take much to spark an investigation.

And the argument that Google isn’t ‘targeting’ UK users just because its app is based in the US won’t wash. Nick was clear on this point: "If the service is available in the UK and promoted in a way that reaches UK users, the Regulations apply. Trying to wriggle out of them won’t work in court".

The data problem

There’s also a more practical, technical issue. AI needs access to accurate, real-time data to work well. In travel, that means availability, pricing, blackout dates, service changes and booking rules - the kind of data that comes from complex supplier relationships. Many of those suppliers - especially airlines - are highly protective of their APIs. Ryanair, for instance, has a history of taking swift legal action against companies using its data without permission.

Unless Google negotiates individual deals with each supplier, Gemini could end up producing inaccurate itineraries, e.g. hotels or restaurants that are permanently closed, flight routes that don’t exist, ski resorts in summer. That’s not just a bad experience – it’s a reputational risk.

Right now, travel companies absorb that risk and invest in the infrastructure to manage it. Google would have to do the same if it wanted to play a more active role in the value chain.

“My assumption is that Google would prefer to just let others do all the work and get some sort of commission as a referrer to existing travel companies".

That, after all, has worked very well so far.

So where does that leave us?

Google won’t rest on its laurels. It will most likely continue expanding Gemini’s capabilities - offering smarter suggestions and improving the user journey across travel research.

It may also look to monetise this in more subtle ways: affiliate links, premium placement in AI results or deeper partnerships with selected suppliers. That would allow it to boost revenue without taking on the operational and legal weight of becoming an agent.

There’s certainly a financial upside. If even a fraction of users book through Google via affiliate links, it could represent billions in incremental revenue without cannibalising its core ad business.

But there’s a fine line between helpful automation and anti-competitive behaviour. And Google will need to be cautious – not just for legal reasons (they have had the money and power to bat most lawsuits away), but to maintain its position as top dog rather than becoming a competitor to the very businesses that fuel its profits.

Thanks to Nick Parkinson at Techlaw for providing expert input on the legal side of this article.