Only 5% of consumers act on an AI recommendation without checking first, and half of them go straight to the business's own website — a commission-free discovery channel that independent hotels must convert before AI agents start taking payments themselves.

Yext surveyed 3,848 adults in the US, UK, France, Germany and Italy this spring. After an AI recommends a local business, 53% search Google or Bing and 49% visit the business's website directly. Trust barely changes the behaviour: US respondents who rated their trust in AI five out of five checked Google at 62%, against 63% for neutral users.

Adobe's traffic data points the same way for travel. In July, visits from AI tools to US travel sites were up 119% year on year. Those visitors were 21% more engaged than other traffic and converted only 1% less, against a 47% conversion gap a year earlier.

What is the Shortlist Effect?

The pattern echoes the billboard effect documented by Chris Anderson of Cornell's Center for Hospitality Research. In 2009, Anderson listed four JHM hotels on Expedia, removed them, then relisted them. During the visible weeks, reservations through the hotels' own channels rose 7.5 to 26 percent, on top of Expedia's own sales. When Anderson and Saram Han revisited the effect in 2017, it was weaker but alive, because travelers still looked at an OTA before booking.

The visit now starts in a chatbot. A traveler asks ChatGPT for a quiet seaside hotel, gets three names, and checks one of them on the hotel's own site. That is the one page in her search where no competitor sits alongside. When it ends in a direct booking, the introduction cost the hotel nothing.

Where does the independent hold the advantage?

Brands carried the original research: a Courtyard guest knows the product before she clicks. An independent has no brand promise, so its website must verify the AI's one-line description — sea view, good breakfast, quiet at night — and justify the price.

On an OTA, the hotel competes in a price-sorted grid against twenty rivals, some carrying a Genius badge. On its own site, the guest sees only its rooms and rates. If the page confirms the AI's sentence at a sensible price, she has little reason to keep searching.

What does a lost visit cost?

If the page fails, she returns to the AI or to Booking.com and books the same room there. The commission starts at 15 percent and climbs for hotels in Booking.com's visibility programs. The AI made the introduction for free; the hotel then pays commission on the outcome.

How long does the window stay open?

Lighthouse, the hotel software vendor, put AI tools at just under 1% of visits to hotel websites in July, after a jump of more than 50% in a few weeks. The direct-booking window survives because paying inside AI remains rare and distrusted. That is changing: in late August, Google switched on hotel booking inside AI Mode in the US, with Google Pay at the checkout. Every payment step inside the answer makes the hotel website optional.

Operators can test their position now: ask ChatGPT for a hotel in their town with dates, a budget and one preference; if named, compare the first mobile screen against the AI's sentence, then set it beside the same dates on Booking.com. Whichever screen makes the case faster wins the booking.

Nobody can yet measure the effect precisely — Anderson needed Expedia's cooperation to switch listings on and off, and no one can switch off ChatGPT. The Yext and Adobe figures come from companies with commercial interests, and Booking.com commission varies by contract. Watch Lighthouse's traffic share and Google's AI Mode rollout as the indicators of when the free introduction ends.