Your favorite assistant will go beyond simple recommendations but will be able to guarantee you availability, or even book online for you.
When AI becomes your travel agent
“Chatgpt, reserve me for a weekend in Venice for two people, budget 800 euros, with a view of the channels.” This sentence, which was of science fiction six months ago, became reality with the rise of the MCP (Model Context Protocol) protocol, exchange protocol between agent IA on the one hand and external data and sources of data on the other.
Your favorite assistant will go beyond simple recommendations but will be able to guarantee you availability, or even book online for you. A silent revolution that redistributes e-commerce cards and will sooner or later come to send the world of tourism (booking of flights, hotels, stays).
The tilting is underway
The figures do not lie. The Talan-IFOP barometer published in April 2025 reveals that 45% of French people now use generative AI, compared to 32% the previous year. This 40% increase in one year is a massive change in behavior. Even more revealing: among those under 35, the adoption rate reaches 91%.
In the travel sector, the mutation is already noticeable. Chatgpt, Gemini and Perplexity today represent 2.6% of SEO traffic in tourism, against only 1% a year ago. An increase of 160% which announces a heavy trend: 15 to 20% of your web traffic already comes from AI agents, but remains invisible in Google Analytics 4.
The phenomenon takes on a planetary dimension with 1 billion weekly “Search” requests on Chatgpt in the world. Admittedly, this remains modest in the face of 90 billion Google requests, but the growth curve is exponential.
MCP: the technical vault key
The Context Protocol model represents much more than a simple technical protocol. This is the solution that allows AI agents to go from recommendation to action. Concretely, your generative AI becomes capable of accessing in real time to dynamic inventories and prices, personalizing its recommendations and finalizing a reservation in complete autonomy.
The stake is considerable: to offer agents IA a means of acting in the web world. Google, Openai, Anthropic and Perplexity have agreed on this open source protocol, guaranteeing secure and bidirectional connections. This technical convergence announces a rapid standardization of the market.
The profits for the tourism industry are multiple: real time access to proprietary inventories, increased personalization of recommendations, interoperability facilitated between actors, and above all, fluid and automated reservation process.
The end of the SEO, the advent of the Geo
This technical development upsets the foundations of digital marketing. We are witnessing the announced death of the SEO (Search Engine Optimization) for the benefit of the GEO (Generative Engine Optimization). Experts no longer speak of “search engines” but “response engines”.
Logic changes radically. A simple request now generates 25 to 40 automatic sub-reliefs in advanced AI systems, more than 100 in “Deep Search” mode. The challenge is no longer to appear on the first page of Google, but to be the source cited by AI during its automated selection process.
This transformation imposes new rules of the game. The IA crawlers do not index your content as google bots require structured and easily accessible content. Even more critical: they now have an integrated memory and retain past interactions.
New strategic challenges
For the leaders of the sector, this revolution imposes an urgent overhaul. Three critical issues emerging:
– The loss of control over the first interaction. Users can now resolve their requests directly in the AI interface, drastically reducing direct visits to your sites. The traditional model “a click, a session” disappears in favor of user courses which never touch your site if AI completes the task.
– The requirement of readability ia. Your content must be “fueling” and explainable for multimodal AI (text, images, voice). This requires “clarity at the entity level” and structured content according to new standards. Forums and discussions become more important than blog articles for the “Grounding” of IA responses.
– The battle of quotation. The objective will no longer be to be on the first page, but to be the source cited by AI. This requires content with an EEAT (expertise, experience, authority, reliability) and impeccable technical markup.
Adapt concretely
Adaptation cannot wait. Several immediate actions are essential:
– Go from SEO to keyword to SEO by entity. You have to go beyond the keywords to focus on SEO in terms of responses, sources and entities. The emphasis must be placed on structured and annotated content, “fueled” by multimodal AI.
– Prepare MCP connectivity. AI agents will have to connect to the reservation platforms to trigger the purchase and provide real -time availability. This technical interoperability becomes a prerequisite for survival.
– Rethink the metrics. The current indicators (impressions, clicks, CTR) become insufficient. You have to develop new metrics such as “quotes” rather than impressions, integrate new surfaces (AI mode, Discover) and measure the use of the content structured by AI.
Chatgpt shopping: the taste of the future
Recent news perfectly illustrates this mutation. Chatgpt Shopping already transforms the assistant into a personal purchasing agent, with standardized carousels, images, prices and synthesized customer reviews. Currently limited to fashion, beauty, house and electronics, this feature will inevitably extend to the trip.
More significant: no advertising or sponsored placement in current results. This “pure” approach to the recommendation IA fundamentally changes the competitive situation. The real quality of the offer takes precedence over advertising capacity.
The opportunity behind disruption
This revolution, if it shakes up the established positions, opens up new opportunities. Small innovative players can compete with giants, because AI favors the relevance of size. A perfectly connected and documented rural cottage can be recommended before an international channel.
Human remains central. AI excels in logistics and personalization, but emotion, welcome and lived experience remain irreplaceable. The actors who will articulate technological efficiency and human excellence will take a decisive step ahead.
Anticipate the next turn
The MCP is only the beginning. 2025 promises to be the pivotal year when IA agents will massively impose themselves in daily uses. Tourist companies that anticipate this transfer by now adapting their tools, processes and digital strategy build their future.
Three emerging adaptation scenarios: optimize its presence on generative IA search engines, develop generative AI connections for availability and booking, and structure information for complex stays.
Artificial intelligence will not replace the journey, it will reinvent it. It is up to us, tourism professionals, to ensure that this reinvention serves our vision of hospitality and discovery. The question is no longer whether to adapt, but at what speed and with what strategy.
Technologies are evolving, the art of receiving remains. In this equation, the human factor remains our most precious differentiating.
