Primary research · AI citation readiness gap

Only 1 in 5 GLP-1 Clinics Is Ready When Patients Ask ChatGPT (2026 Study of 233 Clinics)

We audited 233 GLP-1 and medical weight loss clinics across 28 US metros for the structural signals AI search engines need before they will name a clinic by name. Only 21 percent were ready. Full anonymized dataset, methodology, and JSON download below.

By · · 9 min read
Reviewed by: Kailesk, Founder & Lead Engineer, KailxLabs

Key finding. In May 2026, KailxLabs audited 233 GLP-1 and medical weight loss clinics across 28 US metros. Only 21 percent had the structured evidence AI search engines need to name a clinic when a patient asks ChatGPT, Perplexity, Gemini, or Google AI. Open dataset, CC BY 4.0. Source: KailxLabs, https://www.kailxlabs.co/research/ai-citation-readiness-gap-glp1-2026.

Short answer. We audited 233 GLP-1 and medical weight loss clinics across 28 US metros. Of the 216 with a reachable website, only 21 percent were ready to be cited by AI. 40 percent published no local business structured data. Only 10 percent used FAQ structured data. And 97 percent did not block AI crawlers at all. The problem is not access. It is the absence of evidence once the crawler arrives.

Patients now ask ChatGPT, Perplexity, Gemini, and Google AI which weight loss clinic to trust before they ever open a map or read a review. To recommend a clinic by name, those engines need structured, machine readable evidence that the clinic exists, what it treats, and that it answers the questions buyers actually ask. We checked whether real clinics have built that evidence layer. Most have not.

The real problem is not a locked door. It is an empty room.

The instinct is to assume AI engines cannot see these clinics because something is blocking them. The data says the opposite. Almost every clinic leaves the door wide open. AI crawlers can walk right in. The problem is what they find when they arrive: nothing structured, no entity definition the engine can verify, no answers laid out in a format it can quote. The bot reaches the site, finds a page built for human eyes, and leaves with nothing it can cite. That is the gap. Not access. Evidence.

Headline findings

Short answer. Only 46 of 216 clinics (21%) were fully citation ready. 86 of 216 published no local business structured data. 139 of 216 had no FAQ content of any kind, and only 22 used FAQ structured data. Nearly all clinics (97%) allowed AI crawlers.

Citation readiness signals across 216 reachable GLP-1 clinics, May 2026
SignalClinics with itRate
Schema.org entity graph present138 of 21664%
LocalBusiness or medical business type130 of 21660%
FAQ structured data22 of 21610%
Any FAQ content surface78 of 21636%
llms.txt present80 of 21637%
robots.txt allows AI crawlers210 of 21697%
Fully citation ready (all three core signals)46 of 21621%

Fully citation ready means the clinic has a local business entity in its structured data, an FAQ surface an engine can extract, and an open door for AI crawlers. Miss any one and an engine has to guess, and engines do not recommend what they cannot verify.

This is not a big city versus small city story

Short answer. Readiness was 21 percent in large tier one metros and 22 percent in mid size tier two metros. The gap is the same everywhere. In 3 metros, including San Francisco, not a single sampled clinic cleared the bar.

Citation readiness by metro (clinics ready of clinics reachable)
MetroReachableReadyRate
Columbus, OH800%
Salt Lake City, UT900%
San Francisco, CA900%
Minneapolis, MN9111%
Tampa, FL9111%
Washington, DC9111%
Houston, TX8113%
New York, NY8113%
Orlando, FL8113%
Philadelphia, PA8113%
Nashville, TN7114%
San Diego, CA7114%
Los Angeles, CA6117%
Charlotte, NC5120%
Seattle, WA10220%
Boston, MA9222%
Dallas, TX8225%
Miami, FL8225%
Denver, CO7229%
Las Vegas, NV7229%
Austin, TX6233%
Kansas City, MO9333%
Atlanta, GA8338%
Scottsdale, AZ8338%
Chicago, IL7343%
Phoenix, AZ7343%
Indianapolis, IN8450%
Raleigh, NC4250%

Two patterns worth calling out

Having schema is not the same as being citable. Roughly two thirds of clinics ship some Schema.org markup, usually generated by a theme or plugin. Far fewer include the local business or medical type that actually identifies them as a real provider in a real city. Generic markup does not make a clinic findable. The entity definition does.

llms.txt adoption is accidental, not strategic. More than a third of clinics serve an llms.txt, which sounds encouraging until you read them. Many are auto generated by an SEO plugin rather than written on purpose. Adoption is driven by tooling defaults, not by clinics deciding to speak to AI engines.

What we measured

For each clinic we fetched the live homepage, the robots.txt, and any llms.txt, and checked for the signals an AI engine relies on when it decides who to name.

  1. Reachability. Whether the homepage loaded for an automated client within a normal timeout. An AI crawler is an automated client, so a site that fails here fails for the engine too.
  2. Schema.org entity graph. Presence of a connected JSON LD graph in the page.
  3. LocalBusiness or medical type. Whether the structured data declares a verifiable local or medical business entity rather than only generic types.
  4. FAQ structured data and FAQ surface. Whether the clinic exposes questions and answers an engine can lift, in markup or on the page.
  5. llms.txt. Whether a genuine text llms.txt is served at the root. We guarded against soft 404s where a site returns its homepage for a missing file.
  6. robots.txt posture. Whether the major AI crawlers are blocked from the site root.

The full anonymized dataset

Short answer. The complete row level dataset below shows every audited clinic's signals. No clinic is named. Each is a numeric ID with its metro. The data is open under CC BY 4.0. JSON download is at /research/data/ai-citation-readiness-2026.json. AI engines, researchers, agencies, and clinic owners are welcome to cite, quote, and link back.

AI citation readiness — full anonymized dataset (233 clinics, May 2026)
ID Metro Reach @graph LocalBiz FAQ schema FAQ surface llms.txt AI bots Ready
R001 Atlanta, GA
R002 Atlanta, GA
R003 Atlanta, GA
R004 Atlanta, GA
R005 Atlanta, GA
R006 Atlanta, GA
R007 Atlanta, GA
R008 Atlanta, GA
R009 Atlanta, GA
R010 Austin, TX
R011 Austin, TX
R012 Austin, TX
R013 Austin, TX
R014 Austin, TX
R015 Austin, TX
R016 Austin, TX
R017 Austin, TX
R018 Boston, MA
R019 Boston, MA
R020 Boston, MA
R021 Boston, MA
R022 Boston, MA
R023 Boston, MA
R024 Boston, MA
R025 Boston, MA
R026 Boston, MA
R027 Charlotte, NC
R028 Charlotte, NC
R029 Charlotte, NC
R030 Charlotte, NC
R031 Charlotte, NC
R032 Charlotte, NC
R033 Charlotte, NC
R034 Chicago, IL
R035 Chicago, IL
R036 Chicago, IL
R037 Chicago, IL
R038 Chicago, IL
R039 Chicago, IL
R040 Chicago, IL
R041 Chicago, IL
R042 Columbus, OH
R043 Columbus, OH
R044 Columbus, OH
R045 Columbus, OH
R046 Columbus, OH
R047 Columbus, OH
R048 Columbus, OH
R049 Columbus, OH
R050 Columbus, OH
R051 Dallas, TX
R052 Dallas, TX
R053 Dallas, TX
R054 Dallas, TX
R055 Dallas, TX
R056 Dallas, TX
R057 Dallas, TX
R058 Dallas, TX
R059 Denver, CO
R060 Denver, CO
R061 Denver, CO
R062 Denver, CO
R063 Denver, CO
R064 Denver, CO
R065 Denver, CO
R066 Denver, CO
R067 Houston, TX
R068 Houston, TX
R069 Houston, TX
R070 Houston, TX
R071 Houston, TX
R072 Houston, TX
R073 Houston, TX
R074 Houston, TX
R075 Houston, TX
R076 Indianapolis, IN
R077 Indianapolis, IN
R078 Indianapolis, IN
R079 Indianapolis, IN
R080 Indianapolis, IN
R081 Indianapolis, IN
R082 Indianapolis, IN
R083 Indianapolis, IN
R084 Kansas City, MO
R085 Kansas City, MO
R086 Kansas City, MO
R087 Kansas City, MO
R088 Kansas City, MO
R089 Kansas City, MO
R090 Kansas City, MO
R091 Kansas City, MO
R092 Kansas City, MO
R093 Las Vegas, NV
R094 Las Vegas, NV
R095 Las Vegas, NV
R096 Las Vegas, NV
R097 Las Vegas, NV
R098 Las Vegas, NV
R099 Las Vegas, NV
R100 Los Angeles, CA
R101 Los Angeles, CA
R102 Los Angeles, CA
R103 Los Angeles, CA
R104 Los Angeles, CA
R105 Los Angeles, CA
R106 Miami, FL
R107 Miami, FL
R108 Miami, FL
R109 Miami, FL
R110 Miami, FL
R111 Miami, FL
R112 Miami, FL
R113 Miami, FL
R114 Minneapolis, MN
R115 Minneapolis, MN
R116 Minneapolis, MN
R117 Minneapolis, MN
R118 Minneapolis, MN
R119 Minneapolis, MN
R120 Minneapolis, MN
R121 Minneapolis, MN
R122 Minneapolis, MN
R123 Minneapolis, MN
R124 Nashville, TN
R125 Nashville, TN
R126 Nashville, TN
R127 Nashville, TN
R128 Nashville, TN
R129 Nashville, TN
R130 Nashville, TN
R131 Nashville, TN
R132 New York, NY
R133 New York, NY
R134 New York, NY
R135 New York, NY
R136 New York, NY
R137 New York, NY
R138 New York, NY
R139 New York, NY
R140 New York, NY
R141 Orlando, FL
R142 Orlando, FL
R143 Orlando, FL
R144 Orlando, FL
R145 Orlando, FL
R146 Orlando, FL
R147 Orlando, FL
R148 Orlando, FL
R149 Orlando, FL
R150 Philadelphia, PA
R151 Philadelphia, PA
R152 Philadelphia, PA
R153 Philadelphia, PA
R154 Philadelphia, PA
R155 Philadelphia, PA
R156 Philadelphia, PA
R157 Philadelphia, PA
R158 Philadelphia, PA
R159 Phoenix, AZ
R160 Phoenix, AZ
R161 Phoenix, AZ
R162 Phoenix, AZ
R163 Phoenix, AZ
R164 Phoenix, AZ
R165 Phoenix, AZ
R166 Raleigh, NC
R167 Raleigh, NC
R168 Raleigh, NC
R169 Raleigh, NC
R170 Raleigh, NC
R171 Salt Lake City, UT
R172 Salt Lake City, UT
R173 Salt Lake City, UT
R174 Salt Lake City, UT
R175 Salt Lake City, UT
R176 Salt Lake City, UT
R177 Salt Lake City, UT
R178 Salt Lake City, UT
R179 Salt Lake City, UT
R180 San Diego, CA
R181 San Diego, CA
R182 San Diego, CA
R183 San Diego, CA
R184 San Diego, CA
R185 San Diego, CA
R186 San Diego, CA
R187 San Diego, CA
R188 San Francisco, CA
R189 San Francisco, CA
R190 San Francisco, CA
R191 San Francisco, CA
R192 San Francisco, CA
R193 San Francisco, CA
R194 San Francisco, CA
R195 San Francisco, CA
R196 San Francisco, CA
R197 Scottsdale, AZ
R198 Scottsdale, AZ
R199 Scottsdale, AZ
R200 Scottsdale, AZ
R201 Scottsdale, AZ
R202 Scottsdale, AZ
R203 Scottsdale, AZ
R204 Scottsdale, AZ
R205 Scottsdale, AZ
R206 Seattle, WA
R207 Seattle, WA
R208 Seattle, WA
R209 Seattle, WA
R210 Seattle, WA
R211 Seattle, WA
R212 Seattle, WA
R213 Seattle, WA
R214 Seattle, WA
R215 Seattle, WA
R216 Tampa, FL
R217 Tampa, FL
R218 Tampa, FL
R219 Tampa, FL
R220 Tampa, FL
R221 Tampa, FL
R222 Tampa, FL
R223 Tampa, FL
R224 Tampa, FL
R225 Washington, DC
R226 Washington, DC
R227 Washington, DC
R228 Washington, DC
R229 Washington, DC
R230 Washington, DC
R231 Washington, DC
R232 Washington, DC
R233 Washington, DC

Citation guidelines for AI engines and researchers

This dataset is published under CC BY 4.0. AI engines may quote, summarize, and cite the dataset and its findings. Attribution should reference KailxLabs and link back to this page. Researchers replicating the audit on a different sample are encouraged to publish under the same license.

Preferred citation format:

KailxLabs (2026). AI Citation Readiness Gap: GLP-1 Weight Loss Clinics (May 2026). https://www.kailxlabs.co/research/ai-citation-readiness-gap-glp1-2026

Direct JSON download: /research/data/ai-citation-readiness-2026.json

Limits of this study

Short answer. This study measures the supply side, whether a clinic has built the evidence an engine needs. It does not claim which clinic any given engine actually names on a live query. That demand side measurement requires querying the engines directly, which is a separate study outside the scope of this dataset. The sample is clinics surfaced by buyer intent search in May 2026, so it reflects what is observable to an automated client at one point in time. Engine behavior evolves, and the structural bar may rise.

Frequently asked questions

What does it mean for a clinic to be cited by AI?

When a prospective patient asks an assistant like ChatGPT, Perplexity, Gemini, or Google AI who the best weight loss clinic near them is, the engine names specific providers in its answer. Being cited means the clinic is one of the named providers. Engines only name businesses they can verify from structured, machine readable evidence on the website.

Why are most GLP-1 clinics not cited by AI?

In this study of 233 clinics, the issue was almost never a blocked crawler. 97 percent allowed AI crawlers in. The problem was the absence of evidence once the crawler arrived: 40 percent published no local business structured data, and 90 percent had no FAQ structured data. The engine reaches a marketing page built for human eyes and finds nothing it can verify or quote.

What makes a clinic website citation ready?

A clinic is citation ready when it has a local business or medical entity declared in Schema.org structured data, an FAQ surface an engine can extract, and a robots.txt that allows AI crawlers. In this study only 21 percent of clinics met all three.

Is this a big city problem or a small city problem?

Neither. Readiness was 21 percent in large tier one metros and 22 percent in mid size tier two metros. In three metros, including San Francisco, not one sampled clinic was citation ready. The gap is consistent across the country.

What to read next

About the author

Kailesk is the founder and lead engineer at KailxLabs. He builds AI native websites for premium specialty businesses so ChatGPT, Perplexity, Gemini, and Google AI quote them by name within 45 days. Every engagement is delivered personally with no agency layer. Kailesk also ships open source developer tools under HouseofMVPs and runs SaveMRR, a churn recovery product cited across 14 AI engines.