Opening hours, dogs, terrace: why contradictory data kicks your café out of AI answers

Why AI recommends your café at all, or doesn't

When someone asks today "Where nearby can I have breakfast with a dog?", they increasingly no longer type it into Google but ask ChatGPT, Google Gemini or the AI overview right at the top of the search. In doing so, the AI doesn't read your beautiful website. It pulls together facts from many sources: Google Business Profile, your website, menu portals, reviews, industry directories. From this heap of data it builds a single, confident answer.

The problem: AI models are trained to sound reliable. If your data is clean and identical everywhere, your café becomes a safe recommendation. If it contradicts itself, something unpleasant happens. The AI notices the uncertainty but can't resolve it, and rather picks the café next door whose data is unambiguous. You're not penalized for being bad. You're left out for being unclear.

For a café this is brutal, because your most important decision questions are exactly the ones that often contradict each other: when are you open? Can the dog come? Is there a terrace? Do you have vegan cake? This is exactly where it's decided whether the AI names you.

Opening hours: the classic where most cafés fail

Opening hours are the most common source of error of all. A typical café has them in four places: in the Google profile, in the website footer, on Facebook and on a delivery portal or reservation service. Then you shift the terrace season in summer, take Monday as your closing day or open Sunday only at 9 instead of 8. You change it on Google, and forget the other three places. Bang, four different truths on the net.

For an AI this is an alarm signal. If someone asks "Is the café Löffelweise open on Sundays?", the model finds "from 8", "from 9" and "closed Sundays" at the same time. It can't know which source is current. The AI's safe answer is then often: "The opening hours vary, please check directly", or it simply names another café. Both cost you the guest who's looking for a breakfast spot right now.

The solution is unspectacular but effective: keep a single master list of your opening hours, including holidays and special times. Enter it identically everywhere and work through the list completely with every change. Holiday opening hours in the Google profile are especially important, because that's exactly what people most often ask the AI about.

Dogs allowed? The question you answer nowhere

"Is the café dog-friendly?" is one of the most frequently asked AI questions about cafés, and at the same time one that's answered unambiguously on hardly any café website. You know that dogs are welcome with you. There's even a water bowl on your terrace. But is that written anywhere? Mostly not. And what isn't written doesn't exist for an AI.

It gets even worse when signals contradict each other. In one review a guest writes "great that dogs are allowed", in another it says "weren't allowed in with a dog", maybe because it's handled differently inside than outside. Your website stays silent on it. The AI sees a yes, a no and no official word from you. Result: a wishy-washy answer that puts off potential guests rather than attracting them.

Be explicit. Write a clear sentence on your website and into the Google profile: "Dogs are warmly welcome with us indoors and on the terrace, a water bowl is ready." Use the "dog-friendly" attribute in the Google profile if available. A single unambiguous sentence beats ten contradictory reviews.

Terrace, seats, indoors or outdoors: keeping spatial facts clean

The terrace is the strongest selling point for many cafés, and a data chaos. On Instagram you see it in summer, in winter it's dismantled. If someone asks in March "Does the café have a terrace?", the honest answer is "yes, from April". If that's stated nowhere, the AI guesses. It might say "yes" although it's currently closed, or "unclear" although you have one of the loveliest terraces in town.

The seating situation is often misrepresented too. A small café with twelve seats that's nonetheless touted as a venue for parties confuses both guests and AI. If someone asks "Can I sit there with eight people?", the answer has to be derivable from your data. If the website says "cozy nooks" and the portal says "ideal for groups up to 20", a contradiction arises that costs reservation inquiries.

Describe your space concretely and honestly: number of indoor and outdoor seats, terrace season, whether covered, whether heated. These details seem trivial but are exactly the fuel from which AI answers are built. Concrete numbers beat marketing adjectives.

Menu and dietary options: vegan, gluten-free, lactose-free

"Is there vegan cake there?" or "Do they have gluten-free options?" are questions that decide visit or no-visit, especially for people with intolerances who want to be sure beforehand. If your menu exists only as a pretty PDF or as a photo on Instagram, the AI can often read the content poorly. A picture of the chalkboard is nice for people, but for machines frequently invisible.

The classic contradiction: the website says "fresh vegan cakes daily", but the daily offering changes and on some days there's none. A review grumbles "no vegan cake there". Now the AI has its favorite problem again: promise and counter-proof side by side. It becomes cautious and might say "sometimes offers vegan options", noticeably weaker than a clear yes.

Make your core offerings available as real, readable text, not just as an image. Phrase it honestly and permanently valid: "We always have at least one vegan and one gluten-free cake variety on offer." If you promise that, keep it too. Consistency between statement and reality is the basis for reviews backing you up rather than contradicting you.

Reviews: your most important AI signal, and how to steer it

AI models love reviews, because they contain real guest language. If many reviews say "quiet café for working with WiFi", "best flat white in town" or "lovingly welcomed with dog", the AI learns your strengths from them. These phrasings then appear almost verbatim in recommendations. Reviews are therefore not just stars, they're training material for your AI visibility.

That's why it's strategic to ask guests for concrete reviews, not invented ones, but honest ones with details. A satisfied regular who writes "finally a café with a big terrace and vegan cake where my dog is welcome" delivers three facts in one sentence that confirm your whole positioning. That's more valuable than ten reviews that only say "nice".

Also reply to reviews, even critical ones. If someone writes "I thought dogs were allowed", and you reply "Dogs are always welcome on our terrace, indoors we kindly ask you to refrain out of consideration", you correct the signal publicly. The AI reads your reply too and gets the clear rule.

The one truth: your NAP and your data core

At its core it's about a simple principle: one truth, the same everywhere. Experts call the base data NAP, name, address, phone number. For cafés there's more to it: opening hours, terrace, dog rule, dietary options, WiFi, reservability. These core facts must be identical on every platform. Even a differing spelling of the café name or an old phone number can disrupt the AI's assignment.

Do an honest inventory once. Open your Google profile, your website, Facebook, Instagram, every delivery portal and every industry directory you appear in. Enter the core facts into a table and mark every deviation in red. Almost guaranteed you'll find several contradictions you knew nothing about, old opening hours, a wrong number, an entry from three years ago.

Then you clean up, line by line. This grunt work is unsexy, but it's the lever with the greatest effect on your AI visibility. A café with boringly consistent data is recommended more reliably by AI than the hippest café with contradictory details.

Your 7-point plan for AI-visible café data

You don't have to be a tech pro to implement this. The effort lies more in discipline than in skill. If you set up the following points cleanly once and then maintain them with every change, you clearly stand out from most cafés that leave their data to chance.

The order matters: first you clear away existing contradictions, then you build up clear, positive statements. A clean foundation brings more than any single marketing campaign. And everything you define has to correspond to the reality in the café, otherwise you produce new contradictions through your guests' reviews.

• Keep opening hours identical everywhere, including holidays and terrace season, with a master list as the source.
• Fix the dog rule in one clear sentence on the website and Google profile, naming indoors and outdoors separately.
• Describe the terrace concretely: number of outdoor seats, season, covered or not, heated or not.
• Provide the menu as readable text, not just as an image or PDF; name vegan and gluten-free options clearly.
• Actively ask guests for honest, detailed reviews and reply to every review.
• Do a NAP inventory: compare all platforms, find deviations and correct them.
• Run a short check round every three months so no new contradictions creep in.