# McKinsey: only 37% of firms see AI impact on EBIT Canonical-URL: https://notitur.com/en/digest/mckinsey-ai-hotels-ebit-distribution Text-Version: https://notitur.com/en/digest/mckinsey-ai-hotels-ebit-distribution.txt Markdown-Version: https://notitur.com/en/digest/mckinsey-ai-hotels-ebit-distribution.md Language: en Published: 2026-10-31T09:46:32.595Z Modified: 2026-10-31T09:46:32.595Z Author: Joan Sanz Topics: Artificial Intelligence Summary: Hotels bought AI for inside the building, but the guest's booking decision now happens outside, in assistants like ChatGPT or Google AI Mode. ## Article McKinsey put numbers on October 5 to a question many owners have been asking quietly for months. According to its Quarterly, **80% of respondents say AI has improved their individual productivity**, but only **37% report a positive EBIT impact**. And only **6%** qualify as high performers. The gap between that 6% and the rest: nearly three quarters of the top performers substantially redesigned their workflows, versus one quarter of the others. It is the distance between using AI and reorganizing the operation around it, as [Hospitality Tech](https://hospitalitytech.com/hotels-bought-ai-building-shortlist-moved-outside-it?utm_source=rss&, utm_medium=rss&, utm_campaign=hotels-bought-ai-for-the-building-the-shortlist-moved-outside-it) analyzes. The nuance the hotel sector should read twice is another one. McKinsey notes that AI agents may do more than join a team: they can **replace the need for coordination between people or functions**. And coordinating is exactly what a traveler does when planning a trip: search engine, reviews, OTA, metasearch, advisor, hotel website, booking engine. Every step is a handoff, and that is where the big distributors built their business. That work is shifting to the guest's own assistant. On August 27, Google began rolling out hotel booking inside AI Mode in the U.S.: the traveler describes the trip, compares options and books without leaving the conversation. The hotel or platform remains merchant of record. Among the launch partners are **Booking.com, Expedia, Hotels.com, Priceline, Trip.com, Choice, Hilton, IHG, Marriott and Wyndham**. Expedia and Booking.com already run apps inside ChatGPT, and Wyndham launched its own. The shortlist no longer forms on your website. It forms in the guest's conversation, and that is where the EBIT McKinsey does not see is decided. The read for hoteliers is uncomfortable but useful. Buying a chatbot for the front desk is fine, and it probably works. But if 100% of your stack points inside the building while the booking decision moves to an external assistant, you are optimizing the part of the funnel that no longer decides. The opportunity lies in preparing inventory, rates and content so those agents choose you and cite you well. That is distribution work, not front-desk work. What to watch: how Google AI Mode expands outside the U.S., what commissions and rules apply to those conversational bookings, and whether independent hotels achieve presence in those assistants without relying only on OTAs. That is where it will show who moves from 37% to 6%. ## Sources - Hospitality Tech: https://hospitalitytech.com/hotels-bought-ai-building-shortlist-moved-outside-it?utm_source=rss&utm_medium=rss&utm_campaign=hotels-bought-ai-for-the-building-the-shortlist-moved-outside-it ## Citation McKinsey: only 37% of firms see AI impact on EBIT. Notitur. 2026-10-31. https://notitur.com/en/digest/mckinsey-ai-hotels-ebit-distribution