Booking.com AI Bookings: Why Under 1% Doesn't Tell It All

Source

Booking.com AI Bookings: Why Under 1% Doesn't Tell It All ([RSU by PriceLabs](/resources/people/rsu-by-pricelabs))

What It Says

Booking Holdings reports that completed room nights traced back to AI tools like ChatGPT remain stagnant at less than 1%. CFO Ewout Steenbergen noted that this share is "still significantly below 1% of our room nights." While Booking.com appears frequently in AI-driven travel queries, these mentions are not yet converting into high volumes of direct referral traffic.

However, CEO Glenn Fogel admitted a significant data gap: the company cannot track travelers who use AI to discover a property but then book directly with the host or hotel. When asked about this leakage, Fogel stated, "I don’t know, they don’t tell me." Despite the low conversion rate, Booking.com is hedging its bets by testing paid placements within OpenAI’s interface and developing its own internal AI assistants, such as Priceline’s Penny and Agoda’s visual search.

Why It Matters

For short-term rental (STR) operators and tech vendors, the 1% figure is a lagging indicator that may mask a shift in consumer behavior. AI tools essentially automate the "savvy traveler" habit of researching on an OTA and then hunting for a direct booking site to save money or get better terms. By removing the friction of manual cross-referencing, AI makes it easier for guests to bypass OTA checkout funnels. This creates a strategic opening for property managers to capture high-intent traffic before the AI search landscape becomes dominated by paid advertisements from major platforms.

Useful Signals

• Direct Traffic Resilience: Approximately two-thirds of Booking.com’s bookings originate from unpaid channels, with over 50% coming through their mobile app.
• Paid AI Ads: Booking.com is already participating in early tests for paid placements within OpenAI and Google’s AI-powered search results.
• Checkout Ownership: About 73% of Booking’s transactions now flow through its own payment system, which provides the data necessary to train its proprietary AI models.
• The GEO Window: There is currently a "window of opportunity" for independent listings to appear alongside major brands in organic AI results before paid "pay-to-play" models become standard.

STR Tech Report Take

The 1% metric suggests that AI is not yet a primary "booking engine," but it is rapidly becoming a primary "discovery engine." The real threat to OTAs—and the opportunity for STR managers—lies in the research phase. If a guest uses ChatGPT to find the "best three-bedroom villa in Scottsdale with a heated pool," they may find an independent manager’s site just as easily as a Booking.com listing. Tech vendors should prioritize Generative Engine Optimization (GEO) and ensure direct booking sites are structured for AI crawlers to capitalize on this shift in the traveler’s path to purchase.

Original Source

Booking.com AI Bookings: Why Under 1% Doesn't Tell It All