The hotel ranking game has a new playing field

A traveler opens ChatGPT and asks for a boutique hotel in Lisbon with a rooftop, good natural light, and the kind of neighborhood that doesn't feel like a tourism district. A shortlist forms. Three properties appear. Yours doesn't. That decision happened before any metric you track registered anything — before the session, before the click, before any ranking you've invested in doing its job. The discovery occurred. You weren't there.

This is already happening at scale. AI-native trip planning — travelers querying ChatGPT, Perplexity, and Google's AI Overviews before they open a booking interface — is leaving measurable traces for properties that look for them. Both ChatGPT and Perplexity now appear as referral sources in Google Analytics when a traveler clicks through from an AI response to a hotel website. But click-through referrals are only the visible edge. The larger share of AI-assisted discovery — where a traveler receives a shortlist and then navigates independently to book — leaves no referral trace at all. The channel is operating well beyond what any standard analytics setup captures.

Which raises the question every hotel marketer should be asking: if this layer is already shaping shortlists, what determines which properties appear in it?

The filter Google just made visible

Google's March 2026 core update — completed April 8, with volatility readings higher than any core update since August 2024 — elevated a signal called Information Gain to the dominant content quality evaluator. The mechanism is precise: Google now scores how much genuinely new knowledge a page contributes relative to everything else already ranking for the same query. Pages built on proprietary data or first-hand evidence gained significant visibility. Templated, rewritten content dropped sharply. AI-paraphrased content farms lost most of their ground.

The update is being covered as an SEO story. It is also something more significant. When Google's AI Overviews assemble an answer, when ChatGPT responds to a travel query with browsing enabled, when Perplexity builds a destination recommendation, all of them are selecting sources against the same underlying standard: does this page contribute something that can't be found expressed identically elsewhere? The ranking filter and the agentic citation filter have converged. A page that clears the Information Gain bar is a page an AI agent would cite. A page that doesn't won't appear on either surface.

For hotel marketers, this is the most significant shift in search logic in years — and it runs in their favor. Google is now explicitly rewarding content that adds something an OTA listing structurally cannot: original voice, specific character, first-hand property knowledge. For the first time, the ranking signal favors exactly the kind of content a hotel direct page is better positioned to produce than any intermediary.

Share of Model is the new competitive metric

For most of the past decade, hotel digital strategy has been organized around a single axis: rank. Position on the SERP, visibility in metasearch, placement in OTA sort order. Those positions still matter. They are not, however, the whole game anymore — and the gap between rank and actual discovery is widening.

The metric that captures the new terrain is Share of Model: how often a property appears in AI-generated responses across the queries that precede a booking decision. Research tracking AI Overview citations found that 72% of citations come from pages not in the top three organic results. A hotel can hold a strong organic position and be entirely absent from AI Overviews on the same query. Commercial teams measuring rank are not measuring the surface where an increasing share of shortlisting decisions are being made.

What determines citation eligibility is what Information Gain measures: does this page give a recommendation engine something it can actually use? Rate and availability are table stakes any OTA already provides. Character, context, and specific property knowledge are what move a property from the index onto the shortlist.

OTAs win the data layer, hotels can own the discovery layer

OTAs retain a short-term advantage where structured, machine-readable data determines visibility — rate, availability, room category. AI systems default to OTA infrastructure for transactional retrieval because it is clean and consistently formatted. But the Information Gain signal is one OTA templating cannot manufacture: a listing that describes a property as offering "comfortable rooms in a central location" gives an AI agent nothing to work with when the query is for somewhere that feels local, or quiet, or design-led. In the discovery and recommendation layer, the content gap between what an OTA listing says about a property and what the property itself can say becomes, over time, a visibility gap.

The window is open now

The discovery layer shaping shortlists upstream of your analytics is not a future condition. It is operating today, in the space between a traveler's first question and their first booking-intent action — and most hotel marketing budgets are still pointed at the layer below it.

Properties that treat their direct content as a commercial asset — specific, voice-forward, built around what makes them genuinely distinct — are already better positioned on this surface than properties that have optimized for keyword density and volume. That is a rare moment in hotel distribution: independent hotels with strong editorial identity and authentic property narrative have a structural advantage over scale operators running templated content. The playing field has tilted. The properties that look up first will find less competition there than they expect.