How AI Assistants Are Changing How Guests Find Restaurants and Hotels

A growing number of guests now ask an assistant where to eat or stay, and AI search in hospitality has become something every venue needs to understand. The surveys disagree on how many. They agree on something more useful: the assistant suggests, and the guest then checks.

That second step is where independent venues have more control than they think. Here is what the data says, why the numbers differ, and what to fix first.

How many guests actually use AI to find where to eat?

It depends on whom you ask, and the gap between surveys is itself the lesson. Two reports published within weeks of each other in autumn 2026 give very different pictures.

A Zonal and NIQ study of more than 5,000 British consumers found that 19% use AI tools to plan visits to pubs, bars and restaurants, with a further 50% considering it. A Menufy-commissioned survey of 1,000 US adults, fielded in September 2026, found that 60% already use AI tools to help choose where to eat.

Why the gap? Different countries, different sample sizes, different question wording, and different sponsors. The US survey was commissioned by a platform that sells digital growth tools to independent restaurants, which does not make it wrong, but it is worth knowing. The British report came from a hospitality technology provider working with a market research firm.

Take both as direction, not as forecast. If you want your own number, ask guests how they found you. A single line on a booking form is enough.

What do guests ask AI to do?

Guests mostly use assistants for lookup tasks that used to be several searches. In the British study, those already using AI named finding reviews and ratings (45%), average prices (43%), menus (38%) and opening times (35%) as their uses.

In the US survey, 54% used AI to find nearby restaurants, 31% to find a cuisine or dish, and 30% to find restaurants in a new city.

Notice what these tasks share. They depend on facts about your venue: hours, prices, menu, location, what other guests said. An assistant has no taste of its own. It summarises what it can find about you.

Which guests are using it, and does it favour independents?

Use is uneven, so your own guest mix matters more than any national average. In the British study, 34% of Gen Z used AI to research venues against 19% across all ages, and 39% of Londoners used it against 16% elsewhere in the UK. Higher spenders and frequent diners were also more likely to use it.

A restaurant in a student district or a business hotel in a capital may already feel this. A village inn may not, yet. Ask yourself which of your guests fit those profiles, and check the booking source data you already hold.

There is an encouraging claim for independents too. In the Menufy-commissioned US survey, 80% said they would try an independent restaurant they had never heard of if AI recommended it. Remember who paid for the survey and what it measures: stated willingness, not behaviour. The sensible reading is that a clear, well-described independent can compete for a shortlist spot without a marketing budget, as long as the facts that an assistant reads are in order.

Do guests trust the answer without checking?

No. Guests treat an AI suggestion as a shortlist, and then verify somewhere else. Coverage often skips this finding.

In the US survey, only 2% would order from an unfamiliar AI-recommended restaurant without checking another source first. Before ordering, 54% reviewed menus and prices, 48% checked ratings and reviews, and 47% visited the restaurant's website. And 59% said they trust friends or family more than AI.

The British study points the same way. Although 76% said they completely or mostly trust AI engines to find a venue, the most common influences on choice were friends and family recommendations (47%), online reviews (32%) and venue websites (25%).

The assistant draws the shortlist. Your website, your reviews and your menu decide who gets booked.

This changes the job. Being found is only the first half, and the second half is surviving a check.

What does the checking step look like for a hotel?

Hotel evidence is thinner, so be careful. The independent data behind this article is about restaurants. A sponsored piece in Hotel Dive argues that AI platforms are becoming a first point of contact for guests, and suggests hotels check how their brand appears in AI recommendations and look for visibility gaps. That is advice from a company that sells AI services, and it carries no hotel-specific figures, so treat it as a reasonable prompt rather than proof.

The logic still transfers. A guest weighing a hotel will check rates, room photos, cancellation terms and recent reviews. If an assistant shows last year's policy and your website shows this year's, the guest notices the mismatch, and trust drops before they reach you.

What should you fix first?

Fix the facts, then make them identical everywhere. Much of this is housekeeping, and it is the same work as good local search, which our guide to local SEO for restaurants and hotels covers in more depth.

Start with a one-hour audit:

1. Ask the assistants. Search for your venue by name, and for the category in your area ("quiet wine bar near the station", "family hotel with parking"). Note what is said and what is wrong.
2. Check the four basics everywhere: hours, address, phone and booking link. Compare your website, map listing and main directories.
3. Publish the menu as text, with prices, not only as a PDF or photo. Assistants and guests both read text more reliably.
4. Describe yourself plainly. One clear paragraph on your cuisine, signature dishes or room types, and who you are for.
5. Answer the questions guests actually ask: dietary options, children, dogs, delivery, parking, accessibility.
6. Keep reviews recent, and reply to them. Our piece on responding to negative reviews explains how.

The US survey's own advice to restaurants matches this list: keep hours, location, menu and ordering details accurate, describe your cuisine clearly, and maintain a steady flow of recent reviews.

What should you not do?

Do not chase tricks. Nobody outside the AI companies can tell you how a given assistant ranks venues, and the surveys above do not either. Be wary of anyone who promises a guaranteed position in AI answers.

Two other cautions:

• Do not hide facts in images or PDFs. If a person has to open a file to find your prices, assume a machine will skip it.
• Do not let your listings drift. The usual cause of a wrong answer is an old page somewhere. Update the source, not just your own site.

For the wider set of site basics, see ten restaurant website must-haves.

Key takeaways

• Surveys on AI use for choosing venues range widely, so treat them as direction and ask your own guests how they found you.
• Guests use assistants to shortlist, then check menus, reviews and websites before acting.
• Your job is to pass the check: accurate, plain and identical facts everywhere.
• Publish text menus with prices, clear descriptions and answers to common questions.
• Test assistants monthly for errors about your venue and fix the source of each one.
• Distrust anyone who guarantees a position in AI answers.