How can you tell if you show up when your patients ask ChatGPT?

Knowing whether you show up in ChatGPT when a patient searches for a specialist is no longer just a curiosity—it’s a business metric. Prospective patients ask AI about symptoms, treatments, and increasingly, who to turn to in their city. This method shows you exactly how to measure it today, without paid tools.

# Why does what ChatGPT says about your practice matter?

It matters because usage is already widespread and the topic is healthcare. OpenAI reported more than 800 million weekly ChatGPT users in October 2025. And according to the KFF survey from March 2026, 32% of adults in the U.S. used AI chatbots for health information in the past year.

Patients asking ChatGPT don’t see ten results—they see a single answer. If your practice isn’t among the sources the model consults, you don’t exist to that patient. And if outdated information appears—an old address or an inaccurately described specialty—the first impression is shaped by an error.

There’s another point to consider: you can’t manage what you don’t measure. Most specialists have never seen what AI says about their name, specialty, and city. The audit turns that black box into a list of actionable findings.

# How do you audit your presence in ChatGPT step by step?

A basic audit takes four steps and less than an hour. First: open a session without history (or use temporary mode) so your previous data doesn’t influence the response. Second: ask questions as a patient, not a doctor. Third: document each response in a spreadsheet. Fourth: repeat the process in Gemini and Perplexity.

• Clean session: use temporary chat or a new profile; the model’s memory skews results.
• Patient language: “I have a spot on my skin; who should I see in Guadalajara?” not “board-certified Mexican council dermatologist.”
• Logging: record the date, exact prompt, whether it mentions you, who it does mention, and which sources it cites.
• Coverage: repeat the same questions in Gemini, Perplexity, and Google AI Mode.

# What prompts should you test for your specialty?

Test between 10 and 15 prompts covering the patient’s three stages: symptoms, treatment, and choosing a doctor. Most practices evaluate only the last stage and miss visibility in the first two, where patients build their shortlist.

# What should you do if you don’t show up or the information is incorrect?

If you don’t show up, the problem is almost always in your sources, not the model. ChatGPT and AI-powered search engines draw on your website, medical directories, Google Business Profile, and media mentions. Correct those sources, and the model’s response will change over time.

Prioritize three areas: a website you own with content that answers patients’ questions in clear language; consistent information (name, professional license, address, specialty) across all directories; and pages with structured data that models can read. The Google’s structured data guidelines are the technical starting point.

# How often should you repeat the audit?

At least once a quarter, and after any significant change: moving your practice, adding a subspecialty, or changing your website. Models update their sources continuously, so an annual audit is too late. Also consider that 77% of users express concern about the privacy of their medical data when it comes to AI, according to KFF: patients who do trust the answer arrive with high expectations for seriousness.

Save each snapshot in the same spreadsheet. In two or three quarters, you’ll have a time series showing which prompts you gained mentions for, which ones still mention only your competitors, and which new sources the model cites.

# Frequently asked questions

# Does ChatGPT recommend doctors by name?

Yes, when it finds consistent public sources: websites, medical directories, press coverage, and verifiable profiles. If your digital footprint is weak or contradictory, the model responds in general terms or mentions providers with stronger sources.

# Can I pay to appear in ChatGPT responses?

No. You can’t buy organic responses from models the way you buy ads. The only way is to build signals: useful content on your website, consistent information in directories, and mentions from authoritative third parties.

# Does the same audit work for Gemini and Perplexity?

Yes, the method is identical: the same prompts, a clean session, and a record of mentions and sources. Results vary by model because each weighs sources differently, which is why it’s worth measuring the three leading models.

# What should I do if AI gives incorrect information about my practice?

Correct the original source: your website, Google Business Profile, and the directories where you’re listed. Models reflect what they read; when sources agree, responses tend to be corrected in subsequent update cycles.

# Do I need paid tools to run an audit?

Not to get started. Free ChatGPT, Gemini, and Perplexity accounts are enough for a manual quarterly audit. AI mention-monitoring tools provide scale and historical time series, and make sense once you’ve decided to invest in AI search visibility.

# How long does it take for a correction to show up?

Weeks or months, depending on the model and source. Systems with live browsing, like Perplexity, reflect changes faster; models that depend on training take longer. That’s why a quarterly audit is a reasonable cadence.

# Conclusion

• 32% of U.S. adults already consult chatbots about health; the same trend is reaching Mexico.
• Audit with a clean session, patient language, and systematic tracking of mentions and sources.
• Cover all three stages: symptoms, treatment, and choosing a doctor.
• If you don’t appear, correct your sources: your website, directories, and structured data.
• Repeat this every quarter and turn the audit into a measurable series.

The audit tells you where you stand; the next step is to close the gaps. At ia¹, we do both: start with the AI visibility assessment, learn about our AI search visibility for doctors or message us on WhatsApp with the results of your first test.