Does ChatGPT Mention Your Practice? How to Measure Your AI Visibility

By Fateme Taghiniai, Founder of PraxiOps · Last updated: {veroeffentlicht}

In short: There are now dedicated tools that measure how often a brand is mentioned in ChatGPT, Claude or Gemini. They are built for companies competing against many rivals in many markets. For practices there are two better options: a structured self-test (15 patient questions, once a month, about 30 minutes), or handing both measuring and improving over to PraxiOps, with no technical know-how and no time of your own. Here is how both work.

You can see how many people visit your website. You can see how many reviews you have on Google. But when a patient asks ChatGPT which dental practice in your neighbourhood is good with anxious patients, you see nothing. Not the question, not the answer. And not whether your practice was part of it.

This gap is why a new category of software is emerging: so-called AI rank trackers or "AI visibility tools". The question practices ask us: do we need one?

What an AI rank tracker actually measures

Google has rankings: position 1, position 2, position 3. ChatGPT, Claude and Gemini don't. The AI answers in full sentences, sometimes naming three practices, sometimes five, sometimes none, and the answer changes from question to question.

So these tools work differently. They automatically ask AI assistants the same questions again and again and keep count:

• Mention rate: In how many of the questions asked does the name appear?
• Share compared with other practices (in marketing terms, "share of voice"): How often are you mentioned compared with others?
• Sources: Which websites does the AI rely on? Your own website, Google, Jameda, a newspaper article?
• Description: Is what you do described correctly?

It's a sensible approach. The only question is who puts it into practice for your practice.

Do practices need a dedicated tool for this?

Most of these tools come from the US and are aimed at software companies, online shops and large brands. They track hundreds of search queries across several countries and languages. Accordingly, they often cost a three-figure amount per month, and larger plans considerably more.

For a practice, the world looks smaller:

• Your patients come from one city or one neighbourhood, not from 20 countries.

• The relevant questions can be counted on two hands.

• Many tools measure answers without a fixed location. Whether ChatGPT recommends your practice to a patient living on your street is often reflected only roughly.

• Above all: they show you that you're missing, but not what to do about it. The work on profiles, reviews and practice details stays with you.

Our honest assessment: A general brand-tracking tool rarely fits a practice. Two options make sense: measure it yourself with the self-test below, or use a solution built for practices that takes over the measuring and fixes the causes. That is exactly what PraxiOps does (more on this below).

Which AI assistants matter for practices

Not every assistant is equally important for your patients. A practical order:

1. Google with AI Overviews and Google Maps: still the most important place where patients search.
2. ChatGPT: the best-known AI assistant, including among people with no interest in technology.
3. Gemini: built into many Android phones and Google services.
4. Claude, Perplexity, Copilot: smaller user groups. Claude, for example, is often used for work. A quick look is worthwhile, more effort usually isn't.

The good news: all these assistants draw on similar sources. If your profiles on Google, Jameda and Doctolib are well maintained, you will usually be found more easily by the others too.

The 30-minute self-test: how to measure your AI visibility

1. Write down 15 questions your patients would ask

Not "dentist Berlin", but full sentences, the way people talk to an AI. Mix three types:

• General questions: "Which dermatology practice in Potsdam has good reviews?"
• Questions with specific needs: "A psychotherapist in Berlin-Neukölln who works in English and accepts statutory health insurance?"
• Questions about your name: "What do you know about the practice [name] in [city]?"

Tip: Ask your reception team. They know best what people call about.

2. Ask the questions the same way every month

• Always in a new chat, so earlier conversations don't influence the answer.
• If possible, without being logged in, or with an account that knows nothing about you.
• Name the location in the question instead of relying on your device's location.
• Same questions, same assistants, roughly the same day each month.

3. Note four things for each answer

A simple table is enough:

Your mention rate is then simple: in how many of the 15 questions were you mentioned? 3 out of 15 in the first month is a normal starting point, and a good basis for seeing change.

4. Ask new patients

The most honest measurement happens at reception: "How did you hear about us?" Include "ChatGPT / AI assistant" as an answer option, voluntary and separate from the medical history form.

What to do with the results

The result isn't a report card, it's a to-do list.

• You're not mentioned: Look at which sources the AI names for other practices. Almost always these are Google profiles, review portals and directories. That's where it pays to start: complete profiles, up-to-date services, answered reviews.
• The description is wrong: Old address, wrong specialty, outdated opening hours. The cause is usually an outdated profile somewhere online. Find it, correct it, then check again after one or two months.
• You're mentioned, but rarely: Reviews and the replies to them are among the most important sources for AI assistants. Replying regularly, personally and factually gives the AI more to rely on.

How to maintain these sources step by step is covered on our page on AI visibility for practices.

No time for this? Hand measuring and improving over to PraxiOps

The self-test is honest, but it takes time every month, and the real work only starts afterwards. Many practices we talk to have neither the hours nor anyone on the team who knows their way around directories, structured data and AI search. That's what PraxiOps is for.

What PraxiOps takes over for you:

• Measuring, automatically: PraxiOps tracks whether and how ChatGPT, Gemini, Perplexity and Claude mention your practice, using the kind of questions patients ask about your specialty in your city. No copying and pasting, no table.
• Identifying sources: You see which pages the AI relies on, and where information about your practice is missing or outdated.
• Fixing the causes: Name, address, specialty and services are kept consistent on Google, Jameda, Doctolib and other platforms, in one dashboard instead of 12+ platforms one by one.
• Answering reviews: AI-drafted replies that are checked against German healthcare advertising law (HWG) and the GDPR before they're sent. (Not legal advice.)
• Fully done for you, if you wish (Done-for-You): Our team maintains your profiles for you. You need no technical know-how and no time of your own, just a quick look at the report.

In our 5 pilot practices, this saved around 20 hours per week, about half a marketing position, and reviews were answered 40% faster. PraxiOps starts at €49 per month, with 100 free credits. For comparison: an agency often costs a small practice €700 to €1,000 per month.

What we don't promise: Nobody can guarantee that an AI will mention your practice. Anyone who promises that is selling you tricks. What we do: make sure that every source AI assistants rely on is complete, correct and up to date, and measure every month whether it's working. That's the honest way for patients to find you because they trust you. Not another Paid Placement.

What to keep in mind

• No patient data in the AI. Keep test questions general. Real names, diagnoses or patient enquiries don't belong in ChatGPT and similar tools.
• Use results internally. A sentence like "ChatGPT recommends us as the best practice in Potsdam" on your website or in ads is risky. In Germany, advertising with recommendations and comparisons is tightly restricted for healthcare professions by the Heilmittelwerbegesetz (HWG, the German healthcare advertising law) and competition law, and many other countries have similar rules. (Note: not legal advice.)
• No tricks. Bought reviews, fake profiles or texts written "just for the AI" do more harm than good, and don't fit a practice that patients are supposed to trust.

Conclusion: measuring is the first step, not the last

You don't need an expensive brand tool to know whether ChatGPT knows your practice. You can measure it yourself, with 15 good questions, a table and a fixed date each month. Or you can hand measuring and improving over to PraxiOps and simply keep an overview. The gaps this reveals are usually small and easy to close: an incomplete profile, an unanswered review, a service that isn't mentioned anywhere online. With PraxiOps, they are identified automatically and resolved for you, so your practice is presented completely and correctly wherever patients search.

Every search without a result is a person who needed your help. It's worth taking a look.

Where does your practice stand today? Our free AI visibility check shows you in 2 minutes whether ChatGPT and other assistants mention your practice, and where information is missing.

→ Check your AI visibility for free

Afterwards, you can start with 100 free credits. See how PraxiOps continuously measures and improves your AI visibility on our AI visibility page.

Frequently asked questions

What is an AI rank tracker? A tool that automatically asks AI assistants such as ChatGPT, Claude or Gemini the same questions again and again and measures how often and how a name appears in the answers.

Is there a ranking in ChatGPT like on Google? No. ChatGPT answers in sentences and names different practices depending on the question. That's why you measure the mention rate across many questions instead of a position.

How often should a practice check its AI visibility? Once a month is enough. Changes to profiles and reviews usually only show up in the answers after a few weeks.

Can PraxiOps take over the measuring for me? Yes. PraxiOps continuously tracks whether ChatGPT, Gemini, Perplexity and Claude mention your practice, shows the sources behind it and keeps your details up to date on all platforms. With the Done-for-You service, our team maintains your profiles for you. You need neither technical know-how nor time of your own.

Can PraxiOps guarantee that ChatGPT will recommend my practice? No, nobody can seriously promise that. PraxiOps makes sure the sources AI assistants rely on are complete and correct, and measures every month whether your practice is being mentioned more often.

Do I also need to check Claude and Perplexity? A quick look doesn't hurt. For most practices, though, Google, ChatGPT and Gemini matter much more.

Can I advertise that ChatGPT recommends my practice? We advise against it. Advertising for healthcare professions is tightly regulated (in Germany by the HWG and competition law), and AI answers change constantly. Use the results internally. (Not legal advice.)

Note: This article is for general information only and does not constitute legal advice.