GEO Readiness Benchmark 2026: We Audited 50 US Practices for the Signals AI Engines Use to Cite Them

Only 2% of the 50 US dental, med spa, and law firm websites we audited had FAQPage schema — the single structural signal AI engines most rely on to extract and cite answers. 80% had no entity (sameAs) links, and the average GEO Health Score was just 41 out of 100. Most local practices are structurally unprepared to be recommended by ChatGPT, Perplexity, and Google AI Overviews, regardless of how they rank on Google.

Why we ran this study

AI search — ChatGPT, Perplexity, Google AI Overviews, Claude, and Gemini — now answers a large and growing share of the "find me a local provider" queries that used to go to Google's ten blue links. When these systems recommend a dentist, med spa, or law firm, they draw on structured data, entity signals, and third-party citations to decide who to name.

Yet almost all of the marketing spend in these verticals still targets Google keyword rankings. We wanted to quantify a specific, measurable question: how structurally ready are typical US practice websites for AI citation? Not "how good is their marketing," but "do they have the machine-readable signals AI engines look for before they cite a local business?"

What we measured (and what we didn't)

We built an automated crawler that fetches each practice's homepage and robots.txt and checks seven on-page GEO-readiness signals. Every number in this report comes directly from that crawl.

We measured structural readiness, not live citation rates. This study does not claim to measure whether a given practice currently appears in ChatGPT or Perplexity — those results vary by query phrasing, location, and model version and are not stable enough to benchmark at scale. Instead we measure the prerequisites AI engines use: the structured data, entity links, and directory presence that make a business citable in the first place.

Methodology

We audited 50 US practices across three verticals in April 2026:

• 23 dental practices — general, cosmetic, and multi-specialty
• 14 med spas — single- and multi-location aesthetic practices
• 13 law firms — personal injury, immigration, and family law

Practices were sampled from public business listings in New York City and Phoenix. None were AltorLab clients and none paid to be included. Each site was scored on a 0–100 GEO Health Score built from seven weighted signals: FAQPage schema (25 pts), LocalBusiness schema (15), practitioner credentials (15), AI-crawler access in robots.txt (15), third-party directory presence (15), insurance/service information (10), and sameAs entity links (5).

Headline results across all 50 practices

The pattern is consistent: practices generally have the "human trust" signals (credentials, open crawler access) but almost none of the "machine-readable" signals (structured Q&A data, entity links) that AI engines need to extract and attribute an answer. GEO Health Scores ranged from 15 to 80 out of 100, with a 41 average.

Readiness by vertical

Med spas were the least prepared vertical — not one had FAQPage schema, and the average score was 10 points below dental. Dental practices scored highest, helped by more consistent LocalBusiness schema and directory presence. Not a single law firm or med spa in the sample had FAQPage schema on its homepage.

The three gaps that matter most

Ranked by how often they appeared across the 50 practices, these are the structural gaps most likely to keep a local business out of AI answers:

1. No FAQPage schema (98% of sites). This is the format AI engines most readily parse into a citable answer. It can be added to existing pages in a day, and almost no one has done it — which makes it the single highest-leverage fix available right now.
2. No entity (sameAs) links (80% of sites). Without sameAs links tying the website to its Google Business Profile, Healthgrades, or other profiles, AI engines struggle to confirm that the business on the site is the same entity they see elsewhere — so they hesitate to cite it.
3. Weak or no third-party directory presence (58% of sites). AI engines build their initial knowledge of local providers from aggregator sources. A practice absent from Healthgrades, Zocdoc, Yelp, and Google Business Profile has few external footholds for an engine to reference.

What this means for practice marketing

There is a structural mismatch between where practices invest (Google rankings, design, ad spend) and what makes a business citable by AI (structured data, entity links, directory presence). A practice can rank well on Google and still be structurally invisible to AI engines, because the two systems reward different things.

The encouraging part: the biggest gaps are technical and cheap to close. FAQPage schema, sameAs entity links, and directory consistency are one-time engineering fixes — far less effort than moving from page two to page one on Google. In a market where 98% of competitors haven't done them, the practices that act now become the AI-recommended default in their city.

Frequently asked questions

What percentage of US practices have FAQPage schema?

Only 2% — one of the 50 practices we audited. FAQPage schema is the structured-data format AI engines most readily extract answers from, which makes its near-total absence the largest AI-readiness gap we found.

Which vertical is least ready for AI search?

Med spas, with an average GEO Health Score of 31/100 and zero practices carrying FAQPage schema. Dental practices scored highest at 47/100, and law firms averaged 40/100.

Is blocking AI crawlers the main problem?

No. Only 8% of audited sites blocked any AI crawler in robots.txt. The dominant problem is the opposite: crawlers can reach these sites freely, but the sites lack the structured data and entity signals AI engines need to cite them.

Does this study measure whether a practice appears in ChatGPT?

No — and we're explicit about that. This benchmark measures structural AI-search readiness: the on-page signals that are prerequisites for citation. Live citation results on a specific platform depend on query phrasing, location, and model version and aren't stable enough to benchmark at scale. Readiness is the controllable input; citation is the downstream outcome.

How can I see my own practice's score?

Use the free AI visibility checker for an instant GEO Health Score, or book a free audit for a detailed breakdown with a prioritized fix list.

City-level data

We break the national sample down by market. Each city page reports the real audited signals for that metro:

• Med spas in New York City — 14 audited, avg GEO score 31/100
• Dental practices in New York City — 13 audited
• Dental practices in Phoenix — 10 audited, avg 52/100
• Law firms in New York City — 13 audited