Booking Holdings AI threat proves minimal despite stock selloff
LLM-driven bookings remain under 1% of room nights while direct traffic grows, yet shares trade at a steep discount to historical averages.

Booking Holdings has underperformed the S&P 500 by 37 percentage points over the past year as investors price in disruption from AI travel agents. The company's actual results tell a different story: artificial intelligence bookings account for less than 1% of room nights, direct traffic continues to grow, and the online travel giant just beat guidance across all key metrics.
The stock now trades at 15.3 times forward earnings, well below its 10-year average of 22.3 times, according to analysis first reported by CNBC. That valuation gap exists even as the parent company of Booking.com, Priceline, Agoda, Kayak and OpenTable continues to deliver on its long-term growth algorithm of 8% revenue growth and 15% earnings-per-share expansion.
Why it matters
The disconnect between Booking's operational performance and its stock price reveals how AI narratives can override fundamental business results. For enterprise leaders evaluating which companies will thrive or struggle in an AI-driven economy, Booking offers a case study in structural advantages that chatbots cannot easily replicate: payment infrastructure, merchant-of-record capabilities, and direct relationships with millions of independent properties that large language models cannot access on their own.
Where AI actually shows up
Traffic reaching Booking from large language models remains below 1% of total room nights, and management cited third-party data showing that the majority of LLM travel referrals that do occur route to Booking anyway. The company's direct booking channels continue to strengthen, with mobile app bookings now representing the majority of room nights.
The one area where AI has genuinely hurt the business is Kayak, Booking's metasearch brand. The company recorded a $457 million goodwill impairment in the third quarter, explicitly citing rising customer acquisition costs as AI-powered search answers absorb traffic. Kayak operates at the thin discovery layer where users click through to book elsewhere. Booking.com, by contrast, controls the fulfillment infrastructure beneath that layer.
The payments moat
Booking's defensibility stems from its role as merchant of record for nearly three-quarters of its bookings. The company processes payments across more than 100 methods, handles compliance in over 220 countries and territories, and manages customer service when flights get canceled. Both Google and OpenAI have demonstrated they do not want to carry payment risk and customer-service liability, which is why their early agentic travel products hand checkout back to online travel agencies. Booking serves as a launch partner for both companies, converting a supposed threat into a referral channel.
The company's supply advantage compounds this structural position. More than 4.7 million properties connect to Booking's platform, with the overwhelming majority coming from alternative accommodations, independent hotels and small chains rather than major brands. Much of this inventory connects directly rather than through third-party software. An AI agent can find a Marriott, but it cannot contract, price and guarantee a family-run guesthouse in southern Italy.
Results through headwinds
In the second quarter, Booking grew room nights 5%, gross bookings 9%, revenue 8% and adjusted EBITDA 9% to $2.6 billion, producing a 36% margin. The company achieved these results while absorbing the impact of Middle East conflict throughout most of its fiscal year. Customer-service costs per booking are falling at a double-digit rate as AI handles routine inquiries, and personnel expenses remained nearly flat even as revenue expanded.
The stock weakness stems from three factors beyond AI fears: ongoing geopolitical risk from Middle East conflict, and a currency headwind that created an optical deceleration in reported growth rates while constant-currency growth remained steady.
These details were first reported by CNBC.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call

