How to Show Up When Patients Ask an AI Instead of Google

June 30, 2026

A growing share of patients now ask ChatGPT or a similar AI assistant a question like “who is a good dermatologist near me” instead of typing it into Google. That changes how a practice needs to show up online, because AI assistants do not rank pages the way a search engine does. They fetch, read, and cite sources in a specific, mechanical way, and understanding that process is the difference between a wasted marketing budget and a real result.

How ChatGPT actually finds and uses sources

Independent research that examined ChatGPT’s own network traffic (not just its answers) found a consistent structural pattern behind how it sources information from the web.

It fetches through scrapers, and not all scrapers are equal. Every web result ChatGPT pulls in carries a tag identifying where it came from. One tier is a licensed allowlist of major publishers — large newspapers, wire services, encyclopedic sites — and that tier is effectively closed to a local medical practice; you are not getting into it. The tier that actually matters for a practice is the open commercial scraping layer, which handles local, shopping, and general business content. That is the tier you are competing in, and it behaves more like normal web search than like a licensing deal.

Not every question triggers a web search at all. Before searching, the assistant sorts a question into a bucket — a quick lookup, a shopping question, a local question, and so on. Some buckets never touch the live web and get answered purely from what the model already learned during training, with no page fetched and nothing to win. General how-to and definition questions often land here. The local, cost, insurance, and comparison questions patients ask before choosing a provider are far more likely to actually trigger a fetch. That is where your effort belongs.

Fetched, cited, and mentioned are three different outcomes. A page can be pulled into the assistant’s context (fetched) without ever being the clickable source behind a sentence (cited), and your brand name can appear in an answer (mentioned) without either of those happening. You can win or lose each of these independently, and citations attach to one specific sentence, not to a whole page.

Local results are capped tightly. For a “best [specialty] near me” style question, the assistant typically returns only about two local businesses. Being third in a five-listing Google Map Pack might still leave you completely invisible in an AI answer. This raises the importance of your review and profile work rather than lowering it; if local visibility everywhere runs through roughly the same short list, that short list is worth fighting for.

You cannot cite yourself

This is the single most important thing to understand about AI search for a practice. Your own website can be the source of your own facts — your hours, your services, your location, your accepted insurance. But when the assistant makes an actual recommendation (“the best dermatologist in this city is…”), it typically cites a third party for that judgment, not your own site. Nobody trusts a business’s self-description as proof that the business is good.

That means the recommendation layer runs through the same places that already matter for the rest of your marketing: your Google reviews, listings on sites like Healthgrades, local press coverage, and mentions on forums like Reddit where real people are discussing their own experience. Building those third-party signals is not a separate project from your review and local SEO work. It is the same work, doing double duty.

What your site actually needs

1. Plain, readable HTML text for your core facts. Services, pricing or cost information, accepted insurance, hours, and location need to be readable directly in the page’s HTML, not hidden behind a JavaScript widget, buried in a PDF, or baked into an image. An AI assistant reads the text of the page. If your pricing only appears after a script loads a table, the assistant may never see it, and if it cannot find your facts, it will source them from somewhere else, or leave you out.
2. One clear, complete page per topic, not several thin pages saying the same thing. When results collapse, near-duplicate pages on the same domain effectively get treated as one thin page rather than several strong ones. A single well-written page beats five short ones.
3. Text-based content over video. A Reddit thread gets cited because the actual text of the discussion is sitting on the page. A YouTube video usually only returns metadata to a fetch, not the transcript, so it rarely gets cited even when it is genuinely useful. If you have valuable video content, put a full written version of it on your site as well.
4. Content written for the question a person would actually ask, including a plausible version with your city or specialty in it, since the assistant often rewrites a person’s messy phrasing into a cleaner search before it fetches anything.

What is a waste of money

Be skeptical of anyone selling a “ChatGPT ranking algorithm” service. There is no public scoring system to game, and services promising to manipulate it are selling something that does not exist in the way they describe it. Likewise, be cautious of large one-time investments in “AI-optimized content” that duplicate what your site already says in slightly different words — duplicate framing does not help if the underlying facts are already clean, readable HTML, and it can actively hurt if it produces multiple thin pages that dilute rather than strengthen your one clear page on a topic.

The actual lever is unglamorous: accurate, plainly-written facts on your own site, and genuine third-party proof (reviews, listings, press, forum mentions) that a person, not your marketing copy, thinks you are good at what you do.

What to do this week

1. Open your services and pricing pages and check whether the information is visible in the page’s raw HTML, not just rendered by a script. If you are not sure, view the page source.
2. Confirm your hours, location, and accepted insurance are stated in plain text somewhere on your site, not only inside a PDF.
3. Search your own name plus “reviews” and see which third-party sites show up. Those are your current AI-recommendation sources; make sure they are accurate and current.
4. Pick one high-intent question a patient would actually type (“[specialty] near [your city],” “does [your practice] take [insurance]”) and make sure your site answers it in a single, clear place.