Discussion
  • AI Search

AI trip planners now build full itineraries. Are hotels and tour operators seeing bookings from them, or only losing research traffic?

Published 1 reply

Discussion starter: an AI search question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • admin
    Admin
    Original poster

    The evidence available in 2026 points mostly to the second: AI trip planners are absorbing research traffic faster than they are sending bookings. AI referrals remain a small share of total referral traffic for most sites, and travel is no exception. Destination guides, “things to do” lists and itinerary content are exactly the informational queries that AI answers summarise well, so publishers and tour operators who relied on that content feel the loss first. Bookings from AI sources do exist, and they tend to be high-intent when they arrive, but for most hotels and operators they are still small. The shift that matters strategically is where booking happens next: inside assistants, through OTA integrations, or through Google’s own agentic features.

    What the data and platforms show

    • AI referral volume. BrightEdge’s August 2026 data put ChatGPT at about 95% of tracked AI referral traffic, and AI still a small share of total referrals compared with search. BrightEdge does not detail its method, so treat the split as indicative.
    • OTAs inside assistants. OpenAI launched apps inside ChatGPT in October 2025 with Booking.com and Expedia among the first partners. When a user plans a trip in ChatGPT, an OTA is often already in the conversation.
    • Google’s direction. At I/O 2026 Google described agentic booking expanding to local services in the US, with Google contacting businesses on users’ behalf, alongside follow-up conversations from AI Overviews into AI Mode.
    • Measurement gaps. Clicks from AI Mode arrive in analytics as ordinary Google organic traffic, and GA4’s AI Assistant channel covers assistants such as ChatGPT and Gemini but not Google’s own AI features. Some AI-influenced bookings are invisible as such.

    Where the research traffic loss shows up

    Content type Exposure to AI answers Why
    Generic destination guides and “top things to do” High Widely covered, easy to summarise, no unique information
    Sample itineraries High AI planners build itineraries directly
    Practical logistics (transport, visas, seasons) High Short factual answers
    Property pages: rooms, facilities, policies Lower The user is evaluating a specific place and often wants to see it
    Tour detail pages with availability and pricing Lower, but OTAs compete Booking intent; assistants often route to aggregators
    First-hand local expertise: insider guides, seasonal events, niche activities Moderate Harder to replicate, and more likely to be cited than replaced

    How to tell which is happening to you

    1. Split organic traffic by content type: property and tour pages versus guides and itineraries. If the losses are concentrated in guides while booking pages hold, you are seeing research traffic absorbed rather than booking demand falling.
    2. Track GA4’s AI Assistant channel for sessions, bookings and revenue. Compare its conversion rate with organic search; small volume with high conversion is the typical early pattern.
    3. Check Search Console’s generative AI report for impressions on guide pages. Rising AI impressions alongside falling clicks confirms that the content is being used in answers.
    4. Ask guests how they planned the trip. A booking form question or post-stay survey is the only way to capture AI influence that analytics cannot see.
    5. Run the prompts your guests would use (“three days in [city] with kids, where to stay”) in the main assistants monthly, and record whether your property or tours are mentioned and which sources are cited.

    Prepare for agentic booking, carefully

    The booking layer is where AI could eventually move real revenue, and it is where hotels and operators have the least control. A few steps are sensible now without betting the business on any platform. Keep phone lines answered and staff briefed, because Google’s agentic features for local services include calling businesses on a user’s behalf, and an unanswered call is a lost booking you will never see in analytics. Make sure your connectivity partner supports Google’s free booking links for hotels, where available, so direct rates appear alongside OTA rates in Google’s hotel results. Review OTA contracts and rate parity terms with an eye on how those OTAs are being surfaced inside assistants. And keep your own booking engine fast and simple, because a user who clicks through from an AI plan has already decided and will abandon a slow, confusing checkout. None of this depends on predicting which assistant wins; it depends on being bookable wherever the decision is made.

    What hotels should do

    • Make property facts complete and consistent everywhere assistants look: your site, Google Business Profile, OTA listings and major review sites. Accessibility, family facilities, parking, pet policies and cancellation terms are the details AI planners use to match properties to requests.
    • Keep room and rate information on your own site crawlable and in text, not only inside a booking engine widget.
    • Earn reviews that mention the specific attributes guests care about; assistants summarise review themes.
    • Protect direct booking with clear benefits stated on the site (best rate, flexible cancellation, perks), because the assistant may route a user to an OTA by default.

    What tour operators should do

    • Make sure tours are listed and well reviewed on the distribution platforms assistants and OTAs draw on, even if you prefer direct bookings.
    • Publish detailed, specific tour pages: what is included, duration, meeting points, group size, accessibility, and what makes the tour different.
    • Shift content investment from generic destination guides to first-hand material only you can produce: seasonal insights, guide expertise, niche experiences.

    Hypothetical example:

    A boutique hotel’s blog of local guides loses a large share of its organic traffic over a year, while room page traffic holds steady. GA4 shows a small but growing number of AI Assistant sessions, converting at a higher rate than organic search. A guest survey finds that some guests who booked directly had first heard of the hotel in an AI trip plan. The hotel stops producing generic guides, updates its profile and OTA facts, and adds a short “what we recommend nearby” section written by staff to room pages. It treats the guide traffic as lost research traffic and focuses on being the recommended property.

    Expect research traffic to keep shrinking, measure AI-sourced bookings properly so you know they exist, and put effort into being the property or tour an assistant recommends with accurate details, rather than trying to hold onto guide traffic that an itinerary builder can reproduce in seconds.

    Need help with this on your own site?

    Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.