Generative engine optimization (GEO) is the work of making your practice easy for AI assistants to find, trust and name when a patient asks ChatGPT, Perplexity, Gemini or Google AI Overviews for a provider. That means consistent practice facts across the web, a complete Google Business Profile, recent reviews, pages that answer patient questions in the first sentence, clean schema and letting AI crawlers read your site.
Patients already use these tools to choose care: 45% of consumers used ChatGPT or other AI tools for local business recommendations in the past year[1], and among patients who searched for a doctor, 36% cited AI tools as an influence[2]. Below is how the engines choose sources, the steps that apply to a dental, dermatology, orthopedic, optometry or med spa practice, and how to tell whether it is working. For the service side, see our overview of AI search visibility (GEO and AEO).
What is GEO, and how is it different from SEO and AEO?
GEO is optimizing your practice's information so AI tools cite or recommend it inside their written answers. SEO aims to rank a page in the list of links. AEO, answer engine optimization, aims to be the direct answer to a single question. The three overlap heavily: the same accurate, well-structured, widely confirmed information serves all of them.
SEO asks whether your page appears in the top results and gets the click. GEO asks whether, when an assistant recommends three orthodontists near a patient, your practice is one of them and is described correctly. AEO focuses on short, quotable answers to single questions, such as whether a practice offers same-day crowns or takes a particular vision plan.
GEO does not replace SEO. AI tools that search the web still lean on pages that are crawlable, fast and clearly organized. The difference is emphasis: GEO puts more weight on how consistently your practice is described across the web, on reviews and mentions you do not control, and on whether each page states its answer plainly enough to be lifted into a summary. Our practice marketing glossary defines GEO, AEO, entities and related terms, and our comparison of SEO and GEO walks through the trade-offs for one specialty.
How do ChatGPT, Perplexity, Gemini and AI Overviews choose which practices to cite?
Each tool combines what its model already learned with live retrieval from a search index, then favors sources that agree with each other. Google AI Overviews and Gemini draw on Google's index, Maps and Business Profile data. ChatGPT search and Perplexity run their own retrieval. All of them reward specific, consistent, verifiable information about who you are and what you do.
When a patient asks which dermatologist nearby does Mohs surgery, the assistant usually runs searches, reads a handful of pages and listings, and writes an answer from the sources it trusts most. Google's tools have direct access to your Business Profile, reviews and Google's index. ChatGPT search uses its own crawler and third-party search data, and Perplexity cites its sources openly, often listing directories, review sites and local news alongside practice websites.
Ranking well helps but does not decide the outcome: only 37.9% of URLs cited in Google AI Overviews also ranked in the top 10 results for the same query[3]. The engines assemble an answer from whichever pages state the relevant fact most clearly, so a short page on your site listing the insurance plans you accept can be cited even when a larger site outranks you.
What off-site signals should a practice fix first?
Start with the information you control outside your website: identical practice name, address, phone and specialty on every listing, a fully completed Google Business Profile, and a routine for earning recent reviews. Then build third-party mentions, such as local news, association directories and hospital affiliation pages, that confirm what your own site says.
Entity consistency comes first because AI tools cross-check sources. If your website lists one suite number and Healthgrades another, the assistant is less confident these are the same practice. Check Google, Apple Maps, Bing Places, Yelp, Healthgrades, Zocdoc, insurer directories and specialty society finders, and make name, address, phone, hours and specialty match exactly. Our guide to local SEO and Google Business Profile covers this listing work.
For your Business Profile, choose the most specific primary category, list each service by its real name, keep hours current including telehealth, and add current photos of the office and team.
Reviews give AI tools language about what you actually do. A review mentioning a smooth Invisalign start or a quick contact lens fitting says more than a star rating, and recency matters: 74% of consumers look for reviews written in the last three months[4]. Ask every patient, never offer incentives, and never mention a condition in a request or reply. See reviews and reputation for a compliant routine.
Third-party mentions are slower but durable: local news health segments, chamber of commerce pages, hospital or surgery center affiliation pages, professional association directories and specialty newsletters.
How do you write practice pages that AI tools can quote?
Write answer-first pages: one page per real patient question or service, with a heading phrased as that question and a first sentence that answers it plainly. Add the specifics an assistant needs to recommend you, such as location, who performs the service, insurance accepted and how to book. Short FAQ sections and clear lists make the answers easy to lift.
Your front desk is the best source of topics. A dental practice might cover whether it sees children, offers sedation and handles same-day emergencies. An orthopedic practice might explain who performs ACL reconstruction and whether patients can book a sports injury visit without a referral. An optometry practice might answer whether it takes a specific vision plan or offers myopia management for children. A med spa might state who performs injectables and their credentials.
Put the answer first: 'Yes, we see new patients for knee pain without a referral' is quotable; a paragraph of welcome text is not. Keep doctor bios factual: training, board certification, procedures performed and locations. Do not publish outcome percentages or patient counts you cannot document; medical advertising rules apply, and accurate facts are what get repeated.
- One question per heading, phrased the way a patient asks it
- A direct answer in the first one or two sentences
- Location, provider and booking details on every service page
- A short FAQ on each key page
Which technical signals matter: schema, llms.txt and AI crawler access?
Three technical items matter most. Schema markup labels your practice, physicians, services and FAQs in machine-readable form. Your robots.txt file must allow the crawlers that AI search tools use, or they cannot read your pages. An llms.txt file is optional and low cost, but no major engine has confirmed it uses one, so do it last.
For schema, use JSON-LD with the most specific type that fits, such as Dentist, Optician or MedicalClinic, plus Physician entries for each provider and FAQPage markup where FAQs are visible. Use the same name, address, phone and hours as your Business Profile, and mark up only what the page shows.
Many practice sites block AI crawlers by accident through a security plugin or hosting default. Allow OAI-SearchBot and ChatGPT-User for ChatGPT, PerplexityBot, Claude-SearchBot, Bingbot and Googlebot, which serves AI Overviews. GPTBot and Google-Extended govern model training and grounding; blocking them is a separate decision, but blocking search crawlers removes you from those answers.
llms.txt is a proposed plain-text file at your site root that summarizes the practice and links to key pages. It is quick to write and harmless, so add it once the rest is done, as a tidy summary rather than a ranking lever.
How do you measure whether AI tools are citing your practice?
Track a fixed set of patient-style prompts in ChatGPT, Perplexity, Gemini and Google AI Overviews each month and record whether your practice is named, how it is described and which sources are cited. Pair that with referral traffic from AI domains in Google Analytics, branded search trends in Search Console, and an AI option on your new-patient intake form.
Write twenty to thirty questions a real patient would ask, such as the best pediatric dentist in a neighborhood or a dermatologist who takes a named plan. Run them in a private session and note whether you appear, what is said, which competitors appear and which websites are cited; those sources show where to earn listings and mentions.
Being named has value beyond the click: brands cited in a Google AI Overview get 35% more organic clicks than when they are not cited[5]. In Google Analytics, watch referrals from chatgpt.com, perplexity.ai, gemini.google.com and copilot.microsoft.com. In Search Console, watch searches for your practice name, which tend to rise when patients hear about you from an assistant. Add 'AI assistant such as ChatGPT' to your intake form's referral question.
What should a private practice do first?
Do the work in order of effort and impact: run a baseline of AI prompts, fix listing inconsistencies and complete your Google Business Profile, check robots.txt, then rewrite your top service pages answer-first with schema. Start a steady review routine in parallel. Third-party mentions and llms.txt come after the foundation is solid.
Week one: run your prompt list and save the baseline. Week two: correct mismatched listings and fill every empty Business Profile field. Week three: confirm AI search crawlers are allowed and fix schema on the home, location and doctor bio pages. Week four: rewrite the three to five most-asked-about service pages answer-first with a short FAQ.
After that, keep the review routine running, add one answer-first page a month based on what the front desk hears, and pursue a local or professional mention each quarter. If you would rather have this done for you, DAS offers a free audit showing what AI tools currently say about your practice.
Frequently Asked Questions
How long does it take to start appearing in AI answers?
Listing and technical fixes can be picked up within weeks, because AI search tools retrieve live pages and profiles. Reviews, new pages and mentions build over months. No one can promise a given tool will name a given practice.
Can a small practice compete with hospital systems in AI search?
Often, for specific local questions. Health systems publish broad pages, while an independent practice can answer narrow questions about its own services, providers, plans and neighborhood more precisely.
Do I need a new website for GEO?
Usually not. Most of the work is rewriting key pages, adding schema, fixing robots.txt and improving mobile speed on your current site. See practice websites if you are weighing a rebuild.
Should I block GPTBot or other AI training crawlers?
That is separate from AI search visibility. GPTBot and Google-Extended affect model training and grounding; OAI-SearchBot and PerplexityBot affect whether you can be cited, so keep those allowed.
Does GEO involve patient information?
No. GEO uses only public practice information. Be careful with review requests and replies, which should never reference a patient's condition or visit.
Sources
- BrightLocal Local Consumer Review Survey 2026 (2026)
- rater8 2026 Patient Choice Report (2026)
- Ahrefs, AI Overview citations study (March 2026) (2026)
- BrightLocal Local Consumer Review Survey 2026 (2026)
- Seer Interactive, AIO impact on Google CTR (Sept 2025 update) (2025)
See if your practice is a good fit.
DAS Consultants is a New York City agency that builds AI chat, AI search (GEO/AEO), local SEO, review and patient outreach systems for independent medical practices.
Get your free audit →