AI SEO for doctors: what it is, what it is not, and where to start
AI SEO for doctors is the work of making a medical practice easy for AI assistants such as ChatGPT, Perplexity, Gemini and Google AI Overviews to find, verify and name when patients ask for a provider. It builds on classic SEO, adds off-site consistency and measurement, and cannot be bought or guaranteed.
What is AI SEO for doctors?
AI SEO for doctors means shaping the sources AI assistants read about a practice: its website, Google and Bing listings, directories, reviews and independent mentions. The goal is that ChatGPT, Perplexity, Gemini and Google's AI features describe the practice accurately and include it when patients ask who to see. Results are measured over repeated runs, never promised.
"AI SEO" is a loose label. Some vendors use it for software that writes blog posts with a language model, others for anything involving ChatGPT. In this article it means one specific thing: improving the chance that AI assistants name and describe your practice correctly when a patient asks them for help. The industry also calls this Generative Engine Optimization (GEO) or Answer Engine Optimization (AEO); the differences are mostly vocabulary, and we compare the terms in GEO vs SEO vs AEO for doctors.
What is AI SEO, in plain terms?
When a patient types "who is a good dermatologist in Denver for adult acne" into ChatGPT or Google's AI Mode, the assistant usually does three things. It runs one or more web searches in the background, it reads a handful of the pages those searches return, and it writes a short answer that names a few providers and sometimes cites where it read about them. AI SEO works on all three steps:
- Being retrieved. Your website, your Google Business Profile, your Bing listing and the directory pages that mention you need to show up for the searches the assistant runs. That is still largely classic search engine work.
- Being understood. The page the assistant reads must state plain facts it can lift: who you are, what you treat, where, for whom, with which credentials. Marketing slogans are invisible to it.
- Being trusted. Health is a high-stakes topic, so assistants lean toward providers that are described consistently across independent sources. A practice whose website, listings and directory profiles agree is a safer name to repeat.
How is AI SEO different from regular SEO?
Most of the foundations are shared. A fast, crawlable site, one strong page per service, accurate local listings and real reviews help in both. The differences are in where the weight sits and in how success is measured.
| Question | Classic medical SEO | AI SEO (GEO) |
|---|---|---|
| What you are trying to win | A position on a results page and a click | A mention or citation inside a written answer |
| Main inputs | Your pages, links to them, your Google Business Profile | The same, plus directories, reviews, third-party lists and video that the assistant reads |
| Where the work happens | Mostly on your website | Roughly half on your site, half on other sites that describe you |
| How it is measured | Rankings, clicks, calls and bookings in Search Console and analytics | How often each assistant names you across repeated runs of the same question, with an uncertainty range |
| How stable results are | Positions move slowly and are fairly repeatable | Answers change between runs; one screenshot proves very little |
That last row matters more than it seems. Ask the same assistant the same question twice and you can get a different list of names. A report that shows one screenshot of ChatGPT naming a practice is an anecdote. A useful measurement asks each question several times per engine and reports the share of answers that mention the practice, with an interval. We use eight runs per question per engine; the reasoning is in our GEO for medical practices page.
What does AI SEO not do?
Because the label is new, it attracts claims that do not survive a second look. Here is what AI SEO is not:
- It is not a paid placement. There is no ad slot inside ChatGPT's or Perplexity's organic answers that a practice can buy, and no form to submit your practice for inclusion.
- It is not a guarantee. No agency controls what an assistant writes. Anyone who promises "ChatGPT will recommend you within 30 days" is promising something they cannot deliver. The FTC expects objective marketing claims to have a reasonable basis before they are made, and it has acted against inflated AI claims.
- It is not mass-produced AI content. Publishing hundreds of machine-written pages about conditions you treat does not make an assistant trust you. Thin or inaccurate health content tends to hurt a practice in Google and gives an assistant nothing reliable to quote.
- It is not mostly technical tricks. Files such as llms.txt and extra schema markup are cheap hygiene. In our reading of the evidence they are not what moves AI answers; we explain why in llms.txt for healthcare websites.
- It does not replace your reputation. If patients leave poor reviews or your listings show the wrong address, an assistant will reflect that. AI SEO makes the true picture easier to read; it cannot paint a different one.
Which AI assistants matter for a medical practice?
Each assistant retrieves information differently, so a practice can be visible in one and absent from another. A short orientation:
| Assistant | Where its live answers come from | What that means for a practice |
|---|---|---|
| ChatGPT (with search) | OpenAI's own crawler plus third-party search providers; Bing is widely reported to be one of them | Bing Webmaster Tools, a clean Bing Places listing and fast indexing matter more than most practices assume |
| Google AI Overviews and AI Mode | Google's index, with answers written by Gemini models | Classic Google SEO plus visible question-and-answer content and a complete Google Business Profile |
| Gemini (app) | Google Search grounding | Largely the same inputs as Google's AI features |
| Perplexity | Live web retrieval with visible citations | Pages that answer one question clearly, original data, and freshness |
| Claude | Separate crawlers for training, search indexing and live fetching | Do not block Anthropic's crawlers in robots.txt if you want to be read |
Google's AI features are already common on marketing-related searches in the US. When we pulled 33 US Google searches about medical and healthcare marketing on November 19, 2026 (desktop, through DataForSEO), 32 of them showed an AI Overview, and YouTube was among the cited sources in 24 of those 32. Our full method and the raw counts sit behind the AI search visibility benchmark. That pull covers marketing searches, not patient searches, so read it as a signal about the format, not about your specialty.
What can a practice actually control?
Most of the inputs assistants rely on are within reach of a practice that is willing to keep them accurate. In rough order of impact, based on what we see in our own measurements:
- Listings and directories. Google Business Profile, Bing Places, Apple Business Connect, Healthgrades, Zocdoc, WebMD Care, Vitals, your hospital affiliation pages and your specialty board's directory. Name, address, phone, specialty and accepted insurance should match everywhere. See which directories AI assistants cite for doctors.
- Reviews. A steady, honest flow of reviews on the platforms patients use, collected without gating or incentives. Our guide to Google reviews for doctors covers the rules.
- One clear page per service. A page for each condition and procedure you want to be known for, opening with a direct answer and listing who performs it, where and for whom. This is the core of medical SEO and it doubles as AI SEO.
- A plain-facts about page. Physician names, credentials, board certifications as your state allows you to state them, locations, hours, languages and insurance. Assistants quote this kind of page far more readily than a story-led one.
- Independent mentions. Local news, hospital and association pages, podcasts and YouTube. Assistants give weight to what others say about you.
- Access for AI crawlers. Check that robots.txt and your firewall do not block OAI-SearchBot, Claude-SearchBot, PerplexityBot or Googlebot. Many practices block them by accident through a security plugin.
If you want to see where you stand before hiring anyone, our free AI visibility checker walks you through the questions to ask each assistant yourself and scores the basics.
How do you measure AI SEO honestly?
A sound measurement has four parts. First, a fixed set of questions that real patients would ask, written per specialty and city ("best knee replacement surgeon in Tampa", "who treats rosacea near Plano"). Second, repetition: each question asked several times per assistant, because answers vary between runs. Third, a result per assistant, never one pooled number, since ChatGPT and Perplexity can behave very differently for the same practice. Fourth, an uncertainty range around each rate, so that a change from two mentions to three is not reported as a breakthrough.
Alongside that, track what you can see in your own data: visits referred by chatgpt.com, perplexity.ai and gemini.google.com in GA4, AI crawler hits in your server logs, and calls or bookings from those visits. Our guide to measuring AI search traffic shows the setup.
Is AI SEO worth it for a small practice?
For most independent practices the honest answer is: do the shared foundations first, because they pay off in Google either way. Fix your listings, ask every patient for a review the same way, write one solid page per service, and make your about page factual. Those steps cost time more than money and improve both classic search and AI answers. Dedicated AI SEO work, with measurement across assistants and outreach to the directories and lists they cite, makes sense once those basics are in place and your specialty is competitive enough that being named by an assistant changes who books.
Patient behavior is moving in this direction, but published survey figures vary widely by source and method, so we do not quote a single number here. Our article on how patients use AI to find doctors goes through the published surveys and their limits.
What should you ask an agency that sells AI SEO?
- How many times do you run each question per assistant, and do you report an interval?
- Which assistants do you measure separately, and how do you handle Google's AI Overviews?
- What share of the work is on my website, and what share is on listings, directories and other sites?
- What will you not promise? (The right answer includes rankings and AI mentions.)
- Will my physicians review every page before it is published?
- Who owns the accounts, listings and content if we stop working together?
More questions, including red flags in contracts, are in how to choose a medical SEO agency.
How does Only One Medical approach AI SEO?
Only One Medical is a marketing agency for physicians based in Israel. We work remotely with practices in the United States on SEO, GEO and websites, and we do not run paid ads for US practices or handle patient health information. We measure with repeated prompts per assistant and publish our method. Our client work so far has been in Israel, so we do not show US results; what we can show is how we count, on our numbers page. If you want the practical, step-by-step version focused on ChatGPT, read how to get your practice recommended by ChatGPT.
Sources
- Google Search Central: AI features and your website. Google's guidance on how sites appear in AI Overviews and AI Mode.
- OpenAI: Overview of OpenAI crawlers. OAI-SearchBot, GPTBot and ChatGPT-User, and how robots.txt controls each.
- Aggarwal et al., GEO: Generative Engine Optimization (KDD 2024). The paper that named the field; later replications were mostly weaker.
Frequently asked questions
Is AI SEO the same as GEO?
In practice, yes. AI SEO, Generative Engine Optimization (GEO) and Answer Engine Optimization (AEO) all describe work aimed at being named and cited in AI-written answers. The labels differ more than the methods.
Can an agency guarantee that ChatGPT will recommend my practice?
No. No one controls what an AI assistant writes, and answers vary between runs. An honest agency controls the inputs assistants read and measures how often you are named across repeated runs.
Does AI SEO replace regular SEO for doctors?
No. AI assistants retrieve much of their information through search indexes, so a technically sound site, strong service pages and accurate listings remain the foundation. AI SEO adds off-site consistency and a different kind of measurement.
How long does AI SEO take to show results?
It varies by specialty, city and starting point. Live-retrieval assistants such as Perplexity can reflect changes within weeks, while Google-based features follow Google's indexing. We do not promise timelines; we report measured change over rolling 28-day windows.
Should I use AI to write my practice's website content?
AI tools can help with drafts, but every medical page should be written or edited by someone who knows the topic and reviewed by your physicians. Unreviewed machine-written health content is a liability in Google and gives assistants nothing trustworthy to quote.
Do I need llms.txt or special schema for AI SEO?
They are cheap hygiene, not the lever. Clear visible content, accurate listings, reviews and third-party mentions matter far more to what assistants say about a practice.
Want to know how AI assistants see your practice today?
Send us your website. We will check how ChatGPT, Perplexity, Gemini and Google describe your practice and tell you, in plain English, what is holding you back. Pricing depends on scope, so we quote after the first call.
Keep reading
Fargon, A. (2026). "AI SEO for doctors: what it is, what it is not, and where to start". Only One Medical. https://onlyonemedical.co.il/en/insights/ai-seo-for-doctors