Google can book your competitor's table. Can it find yours?

Someone types "cosy Italian for an anniversary dinner, somewhere quiet" into Google. It doesn't hand back ten blue links any more. It hands back three restaurants and a booking link. No browsing. No comparing. Whichever three restaurants Google picked just won a table that used to belong to whoever ranked well on the old-fashioned search results page.

That's not a hypothetical. Google's AI Mode started booking UK restaurants directly, inside search, in April 2026. And a nationwide test of independent restaurants and bars across eight UK cities found that the overwhelming majority, including plenty with excellent ratings and hundreds of genuine reviews, simply never came up when AI was asked for a recommendation.

Good food and a full diary on a Saturday night don't protect you from this. Being invisible to the system doing the recommending does.

If this feels familiar, it should. The same shift is already reshaping how tradespeople get found. Hospitality is simply next.

Google now books the table. You don't get a say.

Here's what changed. Inside Google's AI Mode, someone can describe what they want in plain English: the occasion, the mood, the area, the budget. Google's AI reads that, picks a shortlist, and offers a direct reservation link on the spot. The booking runs through one of eight launch partners: TheFork, SevenRooms, ResDiary, Mozrest, Foodhub, Dojo, DesignMyNight and OpenTable.

Notice what's missing from that flow. There's no moment where the customer scrolls past your listing and decides to click it anyway because the photo looked nice. The AI does the shortlisting before the customer sees any options at all. If you're not on the shortlist, you were never in the running. Not ranked low. Not on page two. Simply not part of the conversation.

Great reviews. Invisible anyway.

The test that prompted this piece checked independent restaurants and bars, 1,211 of them, spread across eight UK cities. Plenty had strong ratings and genuine, substantial review histories built up over years of good cooking and good service. The overwhelming majority of them never surfaced when AI tools were asked for a recommendation in their own city.

The pattern will sound familiar if you've watched what's happened to tradespeople. Being good at the job and being findable to an AI system are two completely different achievements, and right now most hospitality venues have only managed the first one.

This isn't a fringe behaviour either. Around 47% of UK adults have now used an AI-powered search tool, directly or through Google's AI Overviews, and AI Overviews now appear on roughly a third of all Google searches, up from about one in five at the start of 2025. Hospitality recommendation questions, "best place for a birthday dinner in Edinburgh", are exactly the kind of query AI answers directly rather than pointing at a list of review sites.

Your regulars already know where you are. The AI is deciding who the next ten years of new customers get sent to instead.

The industry knows something's moving

This isn't just theory to the trade itself. Roughly 74% of UK restaurants are already experimenting with some form of AI, and marketing is the single most common use case, at around 46% adoption. Meanwhile 78% of diners now expect to book, order and pay online as standard, not as a nice-to-have. The expectation has shifted whether or not any individual venue has caught up with it.

Put those together and you get a trade that senses the ground moving, a customer base that already expects a frictionless digital front door, and a search layer that's started making the recommendation before the customer even lands on a website. The venues that get ahead of this now get years of head start on the ones that wait for it to become obvious.

"We're always full anyway"

Fair, and worth being honest about what this does and doesn't threaten. If every table is already spoken for by regulars, word of mouth, and the wedding you're catering next month, this isn't about survival. It's about the growth you're quietly not getting.

Think about who this actually costs you. The couple who've just moved to the area and don't have a "usual" yet. The visitor booking a table for a special occasion in a town they don't know. The next ten years of people who aren't your regulars yet, but would become them, if they ever found you. Being full on the customers you already have says nothing about whether you're winning the customers you don't have yet. Right now, an AI system is quietly deciding that on your behalf, and it's deciding without you in the room.

The bookable, legible venue

Getting into this conversation isn't about chasing a trend. It comes down to a short, practical list.

None of this replaces good food or a warm welcome. It just means the good food and the warm welcome have a chance of being recommended by the system increasingly doing the recommending.

What one missed recommendation is worth

Say an AI recommendation flow like this sends a venue in your town three tables a week that would otherwise have gone elsewhere. Average spend per table, £90. That's roughly £14,000 a year, quietly landing on whoever the AI actually names, for doing nothing more than being structured properly.

Now flip it round. If you're the venue that never gets named, that same £14,000 a year isn't a loss you'll ever see on a P&L. It's a number that never existed in the first place, walking straight past your door to the restaurant three streets over that turned up in the answer.

What a Zenlio Local hospitality site does about it

Zenlio Local is bringing this to hospitality next: pubs, B&Bs, restaurants and holiday lets, built the same way trades sites already are. Restaurant or LodgingBusiness schema, a robots.txt that explicitly welcomes AI crawlers, a bookable path a machine can point to, and reviews built into the page rather than scattered across four apps. Same starting principle as everything else Zenlio builds: get found, in Google and in the AI answers that increasingly decide for people before they've even opened a tab.

Being fully booked on your regulars was never the same as being found by the next ten years of new ones.

Google's already decided which restaurants get recommended in your town. Worth finding out whether you're one of them, before your competitor's table fills up with the booking that should have been yours.