A prospective patient types a description of their situation into ChatGPT, the assistant summarizes options and often names one or two specific practices, and that shortlist comes from what the underlying search layer finds well-documented, consistently described, and easy to attribute to a named provider. If your website and listings do not clearly and consistently describe your practice, you are far less likely to be one of the names mentioned.
What a patient's question to ChatGPT actually looks like
Patients rarely type a business category into ChatGPT. They describe a situation and ask for direction, phrasing things the way they would explain a problem to a friend rather than the way a directory organizes services. The assistant then interprets that description and tries to match it to providers who have described themselves clearly enough to be named with confidence.
A question might read like "I've had lower back pain for months and my doctor mentioned surgery, who should I see near me" or "second opinion before spine surgery, how do I find someone." These are longer, more conversational, and more specific than a typical search-engine query. They also often include qualifiers about location, timeline, or the fact that the patient already has an imaging report or referral in hand. The assistant has to parse all of that and still land on a name.
Why the sources behind the answer matter more than the answer itself
ChatGPT does not maintain its own independent directory of every practice in a region. When a question requires current, local, or reputational information, the assistant draws on web search results, structured business listings, review platforms, and pages that are easy to attribute to a specific named entity. It favors sources that are unambiguous about who runs the practice, where the practice is located, and what the practice covers on its website.
This means the assistant is doing a kind of synthesis: pulling fragments from several sources and stitching them into a short, readable answer. If your practice's name, location, and background appear consistently across your website, your Google Business Profile, and major directories, the assistant has an easier time treating those fragments as reliable enough to repeat. If the information is thin, outdated, or contradictory across sources, the assistant is more likely to skip your practice and name a competitor whose information is easier to confirm.
Why your website's own words decide whether you get quoted
Being listed somewhere is not the same as being quotable. An AI assistant summarizing an answer needs sentences it can lift or paraphrase with confidence, which means your website needs plain-language pages that describe your background, your approach to patient consultations, and how your practice operates, written the way a person would explain it rather than the way a brochure would. Vague marketing language gives the assistant nothing concrete to repeat.
Search engine optimization (SEO) focuses on ranking a page in a list of blue links; answer engine optimization (AEO) and generative engine optimization (GEO) focus on making a page easy for an AI system to extract a direct, accurate statement from. A page written in the passive, adjective-heavy style common to many practice websites gives the assistant little to quote. A page that explains, in plain language, how your practice works, how consultations proceed, and what patients can expect at each visit gives the assistant a specific sentence it can attribute to your name.
Practices that keep this kind of plain-language description current, and that keep their name, location, and credentials consistent across their website and directory listings, put themselves in a better position to be the name an AI assistant offers. Practices that rely only on a homepage built around broad claims and stock imagery are easier for the assistant to pass over in favor of a competitor with clearer, more citable text.
How to see exactly what ChatGPT is telling patients about you right now
You do not have to guess what an AI assistant currently says about your practice. Open ChatGPT, Gemini, or Perplexity and ask the kinds of questions a prospective patient would actually type, using your city or region and a description of a condition or situation rather than your practice's name. Note whether your practice appears at all, what is said about it, and which competitors show up instead.
Repeat the exercise with a handful of question variations, since the phrasing a patient uses can change which sources the assistant pulls from and which names surface. Pay attention to whether the description the assistant gives of your practice is accurate and current. An outdated address, a retired physician still listed as active, or a vague one-line description are all signs that the assistant is working from thin or stale material rather than a rich, current picture of your practice, and each of those gaps is an opening for a competitor to be named instead of you.
This kind of check is worth repeating regularly rather than once, because the sources an AI assistant draws from change as review platforms update, as your website is revised, and as new competitor content appears. A practice that checks its own visibility on a regular schedule can catch a gap, such as a missing service description or an inconsistent address, before it becomes the reason a patient never hears the practice's name at all.
Picture a patient sitting in a waiting room, phone in hand, typing a question about their back pain and where to get a second opinion before committing to surgery. The assistant answers in a few calm, confident sentences and names a clinic across town, complete with a short description of what makes that practice a reasonable choice. Your clinic, closer to that patient and just as capable of answering their question, is never mentioned, not because of anything about the care you provide, but because the assistant had nothing clear, current, and citable to work with when it built that answer.