Ask the AI about your café: how to test your visibility yourself

Why AI search suddenly matters for cafés

Two years ago the matter was clear: whoever looked for a café typed "café nearby" into Google and scrolled through the map results. Today a growing share of these searches works differently. People ask ChatGPT "Where can I get a good flat white with vegan milk in Freiburg?" or let Google's AI overview give them three suggestions directly. The answer comes as finished text, often with only two or three names mentioned.

That massively changes the rules of the game for you as a café owner. On Google there were ten results on the first page; in the AI answer there are maybe three. Whoever isn't named simply doesn't exist for that guest. There's no scrolling on, no second page. The AI makes a pre-selection, and this selection is narrower and more final than any search-engine ranking you've known so far.

The good news: you're not at the mercy of this system. You can check exactly how the AI currently talks about your café and derive from that where you need to improve. The first step is always honest measurement, not guesswork.

The right test questions for your café

Open ChatGPT, Gemini and ideally Google's AI search too, and ask the questions your guests would really ask. Start broad: "What are the best cafés in [your city]?" or "Where can I get a good breakfast in [your neighborhood]?" Note down verbatim which cafés are mentioned and in what order. Is your name there? Or do only the usual three suspects from the city center show up?

Then get more specific, because that's often where your chance lies. Ask about your unique selling points: "Where in [your city] is there a café with specialty coffee and oat milk?", "Which café in [your town] has good vegan cakes?", "Where can I work with a laptop and is there Wi-Fi?" If your café offers exactly that but isn't named, you have a visibility problem, not an offering problem.

Ask each question several times and on different days, because AI answers fluctuate. Phrase them the way a real guest types, so feel free to be colloquial: "Craving really good coffee and something sweet, where do I go in [city]?" It's exactly this natural language the AI processes, and exactly there that you realistically see whether you're in the game.

What the answers really tell you

If you don't find your café at all, that's no reason to panic yet, but a clear signal. It usually means the AI finds too little reliable information about you on the web to recommend you with a clear conscience. AI systems are cautious: they'd rather name a café with 800 reviews and a well-kept website than your insider tip with a thin digital footprint, even if your coffee is objectively better.

Also pay attention to how you're described, if you are named. Does the AI say "cozy café with homemade cakes" even though your focus is actually third-wave espresso and filter coffee? Then the AI has an outdated or skewed picture of you. That description stems from old sources, old reviews or a long-obsolete listing. Such distortions cost you exactly the guests who would fit you perfectly.

Finally, check the hard facts in the answer. Does the AI give wrong opening hours, an old address after your move, or claim you're closed Mondays when you've long been open? Such errors are dangerous because they actively keep guests away. But they're also the easiest to fix, because they almost always trace back to concrete, correctable data sources.

Where the AI gets its café information

To understand your test results, you need to know where the AI draws from. The most important source is almost always your Google Business Profile, formerly Google My Business. Opening hours, address, category, photos and above all the reviews flow heavily into what the AI knows about you. An incomplete or abandoned profile is the most common reason a good café is simply missing from AI answers.

Alongside that, mentions on third-party sites count enormously. Were you named in a local city magazine, a food blog, a "best cafés" list or on a portal like a restaurant guide? It's exactly such editorial mentions that are a strong trust signal for the AI. They show that not only you talk about yourself, but independent third parties do too. A single article in a well-read regional outlet can noticeably lift your AI visibility.

Your own website also plays a role, often an underrated one. If it clearly states in text what you offer – "specialty coffee," "vegan options," "breakfast until 2 p.m.," "outdoor seating" – the AI can pick up these terms. If all of that only lives in a pretty photo or a PDF menu, it's practically invisible to the AI. Text beats image when it comes to machine understanding.

Concrete immediate steps after your test

Start with the Google Business Profile, because this is where you have the biggest lever and full control. Check every detail: are the opening hours correct, including holidays? Is the main category "Café" and not accidentally "Restaurant"? Do you have current photos, a maintained menu and the attributes set that define you, such as "vegan-friendly," "free Wi-Fi," "wheelchair accessible"? Every correct entry raises the chance that the AI recommends you fittingly and confidently.

Then actively look after reviews, but without tricks. Kindly and personally ask satisfied regulars for an honest Google review, ideally with a sentence about what they enjoyed. Because when many guests use words like "best cappuccino" or "great vegan cakes" in their texts, the AI learns exactly that association. Also reply to reviews, including critical ones. That signals an actively maintained, living profile.

Finally, revise your website texts with the AI in mind. Write in clear sentences what defines you, who you're there for and what you're especially good at. A short page of frequently asked questions helps twice over: "Do you have oat milk?", "Can I work at your place with a laptop?", "Are there gluten-free cakes?" Such question-and-answer pairs are almost word-for-word what guests ask the AI, and thus ideal fuel for good answers.

An honest look at the competition

Your test questions incidentally give you a free competitive analysis. Look closely at which cafés the AI names again and again, and ask yourself why. Often you'll find: it's not necessarily the ones with the best coffee, but the ones with the strongest digital footprint. Many reviews, a clear website, mentions in local lists. That's not bad news, because you can build exactly these things yourself.

Feel free to ask the AI directly: "Why do you recommend Café X?" or "What makes Café X special?" The answer reveals which qualities the AI associates with a successful café and which sources it has in mind. If it says "known for its matcha" or "popular with students for working," you see precisely which profiles make an impact on the web and where you can claim your own honest niche.

Don't copy blindly, but sharpen your own profile. If three cafés are seen as "cozy and central" but you have the only in-house roastery in town, then exactly that has to come across clearly everywhere: on the website, in the Google profile, in the way you talk to guests who write a review. Your task isn't to be louder but more distinctive and unambiguously categorizable for the AI.

How often you should re-test

AI visibility isn't a state you establish once and then tick off. The models are updated, new reviews come in, competitors maintain their profiles, and your own data situation changes. That's why a fixed rhythm pays off. A realistic cadence for a café is to ask the same test questions again roughly every two to three months and compare the answers with your earlier notes.

Keep a simple log for this, with no technology at all. A table is enough: date, question asked, AI used, were you named, in what position, and how were you described. After two or three rounds you'll see a real development line. Does your position improve after you cleaned up the Google profile and collected twenty new reviews? It's exactly this feedback loop that shows you whether your measures work.

Keep proportionality in mind. You run a café, not a marketing agency, and your coffee still has to be good. Twenty focused minutes every few months is enough to stay on top of it. The effort is small, but the benefit is real: you know how the most important AI systems talk about you, and you make your decisions based on facts instead of gut feeling.

Your twenty-minute checklist

So you can get started right away, here's the condensed sequence for your first self-test. Deliberately set aside this time, ideally with a coffee, and work through the questions calmly. It's important that you write down your results, because only then can you recognize real progress in the next round instead of relying on your memory.

The test works for any café, whether a small roastery café, a breakfast spot or a bakery with a seating area. Simply adapt the questions to your city, your neighborhood and your real strengths. The more concretely your test questions reflect your actual offering, the more meaningful the result is for you.

• Ask ChatGPT and Gemini three questions each: a broad one ("best cafés in [city]"), one about your strength, one colloquial.
• Note down verbatim which cafés are named, in what order and whether you're among them.
• Check your description: does it match what you really are today?
• Verify the facts: opening hours, address, offerings – is everything correct?
• Open your Google Business Profile and correct every outdated or missing detail.
• Ask the AI why it recommends the most-mentioned competing café.
• Enter everything into a small table and set yourself a reminder for the next test in two to three months.