Patients use ChatGPT to choose a plastic surgeon in four stages: they research the procedure, ask for a shortlist of surgeons in their metro, verify credentials and reviews for the names that appeared, and then sort out logistics such as consult cost and recovery time. Only the shortlist and verification stages name a practice, which makes those two stages the ones a surgical practice can win or lose.
AEO, GEO, AI visibility, and AI SEO are four names for the same work: getting a practice named and cited inside AI answers. This post uses AEO. It describes the patient journey stage by stage. It is a different question from the ClinicAds post on which plastic surgery queries AI answers without a click, which sorts queries by whether they send traffic, and from the post on per-engine sourcing, which covers where each assistant gets its names.
- A patient choosing a plastic surgeon in ChatGPT typically moves through four prompt stages: procedure research, surgeon shortlisting, credential and review checks, and booking logistics. Only the second and third stages name a specific practice.
- The shortlisting prompt, such as best rhinoplasty surgeon in a named metro, is where a practice either appears in the answer or does not. A practice absent from that answer is absent from the consult pool before any ad runs.
- Credential and review prompts are verification, not discovery. A patient asks ChatGPT whether a surgeon named in an earlier answer is board certified in plastic surgery, so the practice site and the certifying board listing must agree.
- Procedure research prompts rarely send a click. This post covers the patient journey by stage; the ClinicAds post on zero-click queries covers which of those queries still send traffic.
- ClinicAds plans paid media at $80 to $150 per booked consult, so each consult an AI answer produces at zero ad cost is worth that much avoided spend. These are agency averages, not guarantees.
What does a patient ask ChatGPT first?
A patient asks ChatGPT a procedure question first, such as what a rhinoplasty recovery looks like or how a breast augmentation differs from a lift. The stage builds vocabulary and rarely names a practice. ChatGPT answers it from general medical sources, so a surgical practice is almost never the named source at this stage.
The stage still matters to a practice for one reason. The words a patient learns here become the words used in the next prompt. A patient who learns the term revision rhinoplasty will type it into the shortlist prompt, and the practice pages that use that exact term are the ones a model can match.
How do patients ask ChatGPT for a surgeon shortlist?
Patients ask for a shortlist by combining a procedure, a metro, and a qualifier, for example best facial plastic surgeon for revision rhinoplasty near Phoenix. The answer names between 3 and 6 surgeons or practices. A practice that is not in that answer is not on the patient's list, and the patient will usually contact only names that appeared.
This is the discovery stage and the only one where a practice can be added to the patient's consideration set by an AI answer. It is also the stage ClinicAds measures in a baseline prompt set, as described in the post on how to measure AI visibility for a plastic surgery practice. Paid media still builds the calendar in parallel, which is why the [plastic surgery marketing program](/surgery) treats AEO as an added line, not a replacement.
What do patients check after the shortlist appears?
After a shortlist appears, patients verify the names with follow-up prompts about board certification, review quality, and complaints. This is verification, not discovery. The patient already holds the names, and the follow-up prompt tests whether the practice survives scrutiny.
A practice survives when its own site, its board listing, and its review profiles agree on the surgeon's name, credentials, and procedures. A mismatch, such as a site that lists a fellowship the directory does not, gives the model conflicting text to summarize. The review side of that check is covered in the post on whether patient reviews change AI surgeon recommendations.
- Is Dr. Name board certified by the American Board of Plastic Surgery?
- What do patients say about Dr. Name's revision work?
- How many rhinoplasties does the practice perform each year?
- Are there complaints or board actions against the surgeon?
Which stage of the journey does each prompt belong to?
Each prompt type maps to one stage, and each stage rewards a different asset. The table below lays out the four stages, the shape of the prompt, whether the answer names a practice, and the page or profile that decides the outcome. It is a planning model from ClinicAds observation, not a measured conversion funnel.
| Stage | Typical prompt shape | Names a practice? | What decides the outcome |
|---|---|---|---|
| 1. Procedure research | What is recovery from a deep plane facelift like? | Rarely | General medical sources; practice wording matters only for later matching |
| 2. Surgeon shortlisting | Best facelift surgeon in a named metro | Yes, 3 to 6 names | Directory profiles, reviews, press mentions, and procedure pages naming the surgeon |
| 3. Credential and review check | Is this surgeon board certified, and what do reviews say? | Yes, one named surgeon | Consistency of site, board listing, and review profiles |
| 4. Logistics | How much is a consult and how long is the wait? | Sometimes | A consult page with a stated fee, a wait time, and a booking route |
Where does a practice lose the patient in this journey?
A practice loses the patient at stage two when it is not named, and at stage three when its own facts conflict across sources. Losing at stage two is invisible, because no click or call is recorded. Losing at stage three shows up as a patient who asked about the practice and then booked elsewhere.
The cost is easiest to see in consults. If paid media books a consult at $80 to $150 and a practice that is absent from shortlist answers would otherwise have captured even 4 consults a month from them, the avoided-spend value is roughly $320 to $600 a month. That is illustrative arithmetic, not an observed result, and it ignores the value of the surgery itself.
- Stage 2 loss: the practice is not named, so nothing is recorded
- Stage 3 loss: the name appears but credentials or reviews conflict
- Stage 4 loss: the consult page lacks a fee, a wait time, or a booking path
What should a surgical practice do about it this month?
A practice should run its own journey as four prompts in ChatGPT, log the answers, and fix the gap at the earliest stage where it is missing. Write the prompts the way a patient would, use the practice's actual metro, and record whether the practice is named at stage two and whether stage three returns accurate facts.
The fix follows the stage. A missing stage two name is a sources and entity problem. A stage three conflict is a consistency problem and the quickest to repair, because it means editing the practice's own site and directory profiles to match the board record. Intake forms should also ask new patients whether an AI assistant was part of how they chose, since the booking form is the only place this journey becomes visible.
Do patients really choose a surgeon from a ChatGPT answer alone?
Rarely alone. Patients use the answer to build a shortlist and then verify names on the practice site, board listing, and review profiles. ChatGPT sets the consideration set; the practice's own pages and records decide whether the patient books.
Can a practice see which ChatGPT prompts led to a booked consult?
Not directly. ChatGPT does not report prompts to practices. The workable proxies are a baseline prompt log run monthly and a booking form question asking whether an AI assistant was used, which together approximate the stage two and three effect.
Does ClinicAds sell this as a separate service?
No. ClinicAds treats AEO as part of the search visibility system inside its three-system scope, alongside paid media and speed-to-lead, and does not sell GEO, AEO, or AI SEO as separate line items.