Revelus Dermatology: nothing left for a model to infer.

The challenge

Ask an AI model to describe a practice and it answers from what it can retrieve. For Revelus, that meant a picture with gaps and errors: the model associated the practice with Mohs surgery it doesn't offer, and it didn't know the practice served Kyle, Buda and Lakeway at all. A model that's unsure about a business resolves the uncertainty by leaving it out of the answer — or by answering wrong.

The opportunity

A practice knows far more about itself than its website says: every condition it treats, every service, every provider's credentials, the insurance it accepts, the policies a patient needs to know before booking. Most of that never gets published, or gets published once and goes stale. If all of it existed as a connected, current record, a model would have nothing to guess at.

The insight

Page count is not the goal. Coverage is. Each page exists because it describes something true about the practice that a patient — or a model — would ask about. A page with a defined reason to exist is a fact with a home; a site made of those pages is a record a model can trust.

What we did

Fix the wrong facts first. Before any new page was written, the errors the models were repeating were corrected at the source: the practice's own pages and structured data stopped giving a model any reason to associate it with Mohs surgery, and the service areas — Kyle, Buda, Lakeway — were stated where a model could find them. Entity clarity before volume.

Plan the record; the practice builds it. Visigent modeled what the practice knows — every condition treated, every cosmetic and medical service, every provider and credential, insurance, forms and policies — and mapped it against the site, page by page: keep, update, consolidate, build. The practice's own team wrote and published from that plan and its specs.

Track whether the answer changed. Citations of the domain and the practice's visibility across 300 category prompts were measured before and after, with the Visigent tracker and Ahrefs.

The published record, and what moved.

Revelus Dermatology, Austin. Every condition, service, provider and policy has a page of its own, connected to the rest. Nothing about the practice is left for a model to infer.

Published record, by what it describesrevelusdermatology.com

ConditionsProblems it treats

Cosmetic servicesElective offerings

Medical servicesAppointments, surgery, Rx

Resources & policiesInsurance, forms, policies

ProvidersPeople and credentials

Page counts from the site's own sitemaps, September 2026. Page count is not the goal; coverage is. Each square is a page with a defined reason to exist.

ChatGPT citations of the domain13 Jul → 6 Aug 2026, Ahrefs

Measured with the Visigent tracker and Ahrefs.

Results

ChatGPT citations of revelusdermatology.com: 2 → 14 between 13 July and 6 August 2026 (Ahrefs).

AI visibility: the practice appears in 36% of 300 category prompts (Visigent tracker run, 22 July 2026). Average position when mentioned: 2.5 across ChatGPT, Claude and Gemini. Negative mentions: 0 in the same run.

Description accuracy: both errors are gone. The models no longer associate the practice with Mohs surgery, and Kyle, Buda and Lakeway appear as service areas (same 22 July 2026 run).

Why it mattered

For a medical practice, the cost of an AI getting you wrong isn't a lost click — it's a patient told you offer something you don't, or a patient in Buda told you're not nearby. Revelus now has a record a model can read end to end. When a patient in South Austin asks what to do next, the practice is in the answer, described correctly.