Here is a job for the back end of a shift, when the last table has gone and you are waiting on the kitchen. Ten minutes, costs nothing.
Open the AI app on your phone, whichever one you use. Type the name of your place and your town. Ask when you are open, what is on and what a main costs. Then read it back the way a stranger would, somebody deciding on a Thursday whether you are worth the drive.
If you have not got one of those apps, there is no need to download anything. Search for your place on Google instead and read the summary it writes above the results. Same questions, same job. If anybody is going to meet an AI answer about your place, that is the likeliest way they meet it.
Most people who do this find something wrong. Usually something small and daft. Sometimes it is the Tuesday closing you dropped eighteen months ago, still being quoted at people.
Why it is worth ten minutes
Zonal and NIQ asked more than 5,000 British consumers about this in October 2026, reported in the Morning Advertiser on the 8th of October. 19 per cent already use AI tools to plan a visit to a pub, bar or restaurant. Another 50 per cent are considering it.
Worth knowing who is counting. Zonal sell tills and NIQ sell data, so neither is a disinterested party when the finding is that operators should worry about technology. A separate figure from Reputation and CGA puts it nearer 26 per cent. Two vendors, two numbers, both pointing the same way. Take the direction as real and the decimal point with a pinch of salt.
AI is also fourth on the list, not first. 47 per cent still go on friends and family, 32 per cent on online reviews, 25 per cent on the venue's own website. Nobody is choosing a Sunday lunch purely on what a machine tells them.
So this is not an emergency. It is a thing worth ten minutes.
What makes it worth them is who the one in five are. People eating out more than three times a week are 42 per cent more likely to use AI than people going out once or twice a week.
That is the regular. The one who already knows three places and is deciding between them.
What they are actually asking it
Here is what the people already using it want to know before a visit. Reviews and ratings, 45 per cent. Average food and drink prices, 43 per cent. Menus, 38 per cent. Opening times, 35 per cent. Deals and discounts, 28 per cent.
Reviews, prices, menus, opening times, deals.
Every one of those is yours. Every one of them also drifts out of date in a working venue without anybody deciding to let it happen. The machine is not inventing you. It is reading what you published and repeating whichever version it found.
Why it gets your venue wrong
Nothing here is a technology problem. It is a housekeeping problem in three parts.
Your opening times live in six places and you only ever update two. The Google listing, the website footer, Facebook, whatever takes your bookings, a pinned Instagram post from last summer, the sheet sellotaped inside the door. Nobody sat down and chose to be on all of those. A previous manager set one up, a rep talked somebody into another, the pubco did a third centrally.
And here is the bit nobody writes down. The reason it does not get fixed is usually a password. The assistant manager who created the Google listing left in 2023. The listing is tied to a personal email nobody has, so every time somebody says they will sort it out they get as far as the recovery screen then go back to work. If that is you, that is the job. Not the app.
Your menu is a PDF, or a photo of the board. It looks right and it prints right, yet to most of these tools it is a picture. As far as I can tell, most of them cannot reliably read the dishes or the prices off it. As far as AI is concerned you have not published a menu, so it goes looking elsewhere. What it usually finds is a third-party listing carrying your 2024 prices.
One honest warning before you go and type the menu out. A menu written out as text is a thing you have then signed up to maintain. If your specials change twice a week and the board gets rubbed off every service, a web page frozen in September is worse than no page at all, because now you are publishing wrong prices rather than none. Publish the part that holds still. The core menu, the Sunday roast, the set lunch. Leave the specials where they belong, on the board.
Your prices moved and the internet did not. New prices go on the board and in the holders the same morning, because they have to. The listing somebody set up in 2022 does not, because nobody is standing in front of it at half past twelve. That is a quiet Tuesday job. There has not been a quiet Tuesday since the spring.
None of this is slackness. It is what happens when the only person who could fix it is also the person covering a section because the chef rang in at ten.
Do the test properly
Ten minutes. Do it as a stranger would.
Ask the obvious questions. When are you open. What is on. What does a main cost. Do you do Sunday lunch. Can you take a pushchair, a dog, a party of twelve.
Ask the one that catches people out. "Is there anywhere good to eat in [your town]?" If your venue is not in the answer, that is information too. Try it a few ways, because you will not get the same answer twice.
Write down only what is wrong. Not what you would like it to say. What is factually incorrect.
Most lists come back short, with three or four things on them. At least one will be a detail you had genuinely forgotten was published anywhere.
Then fix it in this order
Not all of it. In this order, stopping when you run out of patience.
One. Opening times, everywhere. Write down every place your hours appear, then make them agree. This is the best of the five, because wrong hours do not just lose you a browse. They cost you somebody who got in the car, drove over, found the lights off and told everybody about it.
If the Google listing is the one you cannot get into, that is a solved problem and worth the twenty minutes. Go to business.google.com/add, put in your name and address, pick your venue. You will get a message saying somebody else may manage this profile. Ask for access. Whoever holds it has three days to answer. If they do not answer, you can often claim and verify it yourself. If they say no, you can still suggest an edit to the hours, which is not ownership but it does fix the thing that was wrong.
Two. Answer the awkward questions somewhere findable. Dogs. Parking. High chairs. Step-free access. Whether food runs all day Sunday or stops at four. Whether the garden is covered. These are the questions that actually come down the phone before somebody books. Most venue websites answer none of them. You already know what they are, because you answer them out loud six times a week.
Three. Get your menu into actual text. A plain web page with the dishes and the prices written out, so there is something to read rather than a picture to look at. If your site is a template you cannot edit, or the pubco runs it, or all you have is a Facebook page, then put it in the Facebook About section and the Google profile instead. Those both take text and both get read.
Four. Kill the stale listing. Whichever third-party page is carrying your old prices, claim it, correct it or get it taken down. There is nearly always one. It is nearly always the first thing anybody finds. It is also nearly always the one nobody has the login for.
Five. Reviews, which is the big one. 45 per cent is the top figure on that list. Reviews are also the one thing on it you cannot simply go and edit. What you can do is reply to them. The replies are text. They sit alongside the review and they get read the same way everything else does. If somebody complained in March about a thing you fixed in April, your reply is the only place that correction exists. Most operators reply to the bad ones and ignore the good ones. Do both. Keep them short.
That is the lot. No subscription, no new system, nothing to buy.
The honest limits of this
Two things I am not going to do.
I am not going to tell you this fills your Tuesdays. 19 per cent is real. It is still only one in five, in a trade where word of mouth is nearly half. This is housekeeping with a decent return on an afternoon. It is not a growth plan.
And I am not going to sell you a way to rank in AI search. There is a whole industry forming around that promise and most of it is noise. Nobody can tell you reliably how these tools pick what to mention. The answers change week to week. Anybody who says otherwise is guessing with more confidence than the evidence allows. What you can control is whether the information they find about you is correct. That is worth doing on its own terms, because it is the same information a human finds on your website.
The bigger point
Nothing above needs us. It is ten minutes on your phone and an afternoon of housekeeping. Most operators reading this will do the whole lot themselves. If the test comes back clean, you need nothing from anybody. That is the best outcome on offer and it happens more often than you would think.
Where it stops being a ten-minute job is when you work out why the mess built up in the first place. Your hours sit in six places because nobody ever decided which one is the real one. Same with the menu, the prices, the specials. That is not a technology problem either. It is a job that belongs to nobody in particular, which in a venue means it belongs to whoever is least busy, which means it never gets done.
That is worth knowing whether or not you ever speak to anybody about it. The fix is not a system. It is deciding who owns it.
If you want a hand with it
Do the ten-minute test first, whatever else you decide. It costs you nothing and you may well find there is nothing to fix.
If what comes back is a tangle you cannot sort out in an afternoon, or you would rather somebody looked at the whole week instead of one listing, that is the hospitality side of what we do.
The first conversation is free and there is no obligation attached to it. No hard sell, no pitch deck, just an honest look at where your hours actually go and what is worth changing. If there is nothing useful I can add, I will tell you that straight and you will have lost half an hour.
We are based in Newport and work right across South Wales. I am happy to come to you where that helps. I would rather see a kitchen in person than talk about one down the phone.