The 50-point AEO audit framework we run on every dental practice

When a dental practice asks us “where do I stand on AEO right now?”, we do not eyeball it. We do not estimate based on a 15-minute look at the site. We run a 50-point audit, scored across seven weighted pillars, and the practice gets a composite score out of 100. The whole thing takes our analyst about three hours per practice. We have run it on roughly 200 practices to date, and the average score for an unoptimized practice is 31 out of 100.

We are publishing the framework openly because we are tired of dental agencies hiding their methodology. If you want to run this on your own practice, you can. If you want us to run it for you, the audit is free and you keep the deliverable whether or not you ever work with us. Below is the structure, the scoring, and what each pillar actually tests.

The seven pillars and their weights

• Pillar 1 — AI Visibility: 20% weight, 10 points
• Pillar 2 — Schema & Structured Data: 15% weight, 10 points
• Pillar 3 — Content & E-E-A-T: 15% weight, 10 points
• Pillar 4 — Brand Mentions & Citations: 15% weight, 10 points
• Pillar 5 — Technical SEO Foundation: 10% weight, 10 points
• Pillar 6 — Local Pack & Google Business Profile: 15% weight, 10 points
• Pillar 7 — AI Bot Accessibility: 10% weight, 10 points

Each pillar is scored 0–10. Pillar score times weight, summed across all seven, gives a composite 0–100. Below 40 is critical. 40–59 is the average dental practice. 60–79 is meaningfully optimized. 80+ is rare — we see it on roughly 1 in 50 practices, usually ones that have hired an actually-good agency.

Pillar 1 — AI Visibility (20% weight)

We run 15 patient-style queries live in ChatGPT, Perplexity, Google AI Overviews, Gemini, and Copilot. Five local (“best [specialty] in [city]”), five procedure-specific (“[procedure] cost in [city]”), five comparison/cost (“[procedure A] vs [procedure B]”). We screenshot every result and tag whether the practice was named, whether it was the primary recommendation, and which competitors were named instead.

The five free-tier scored checks:

1. Practice named in ChatGPT for “best [specialty] in [city]”
2. Practice named in Perplexity for “[city] [primary procedure] cost”
3. Practice cited in a Google AI Overview for any procedure-explainer query
4. Practice named in Copilot for “emergency dentist [city]”
5. Practice named in Gemini for “[city] [specialty] reviews”

Why this pillar is weighted heaviest: it is the only one that measures the actual outcome. Everything else is a leading indicator. If you are scoring 8/10 on schema but 1/10 here, the schema work has not paid off yet.

Pillar 2 — Schema & Structured Data (15% weight)

The five visible checks:

1. Dentist schema on the homepage with full NAP, hours, geo, payment methods
2. Physician schema for the lead provider with medicalSpecialty, alumniOf, memberOf
3. MedicalProcedure schema on at least three top-revenue service pages
4. FAQPage schema on at least one service page
5. AggregateRating + Review schema with real review data

The other five (paid tier): Organization with sameAs linking to social profiles, BreadcrumbList, MedicalScholarlyArticle for any reviewed content, MedicalCondition on condition explainer pages, and ImageObject with creator attribution on original photography.

Schema is the single most fixable gap on most dental sites. We have not audited a practice yet where schema work alone did not yield at least 2.5 incremental pillar points within 30 days of deploy.

Pillar 3 — Content & E-E-A-T (15% weight)

Healthcare is a Your Money or Your Life category, so AI engines apply Experience-Expertise-Authoritativeness-Trustworthiness scoring aggressively. The five visible checks:

1. Named, credentialed author byline on the homepage and at least one service page
2. “Medically reviewed by [DDS] on [date]” block on at least one clinical page
3. Top three service pages each over 1,500 words
4. 2026 cost ranges visible on at least one service page
5. Original (non-stock) photography on the homepage and at least one service page

The hidden five: provider credentials linked out to state board, hospital, or ADA Find-a-Dentist; last-reviewed date visible on every clinical page; topical coverage breadth (how many of the 20 standard dental topics are covered); comparison tables on at least two service pages; insurance acceptance list with payer logos.

What we usually find: practices have nailed one or two of these and missed the rest. The 1,500-word threshold is the single most-failed check — most service pages we audit are 400–700 words.

Pillar 4 — Brand Mentions & Citations (15% weight)

This is the off-page pillar. The four visible checks:

1. Foursquare listing exists, claimed, and complete
2. Mentioned in at least one “best dentists in [city]” listicle in the past 12 months
3. Quoted or featured in any high-authority publication (Healthline, Verywell, Forbes Health, WebMD, local news) in the past 12 months
4. Reddit or forum thread mentions in the past 12 months

The hidden ones: podcast guest history, ADA Find-a-Dentist completeness, Wikipedia mention, scholarly citation, and a full sweep across 30 directories.

The Foursquare check alone disqualifies 80 percent of practices. The Healthline check disqualifies 95+ percent. These are not unreasonable bars — they are exactly the signals AI engines weigh, and most practices have simply never been told to chase them.

Pillar 5 — Technical SEO Foundation (10% weight)

The four visible checks:

1. Core Web Vitals passing — LCP under 2.5s, INP under 200ms, CLS under 0.1
2. Mobile-friendly: no horizontal scroll, tap targets at least 48px, base font at least 16px
3. All key service pages indexed in Google (verified via the site: operator)
4. Homepage internal links to all primary service pages within one click

This is the most commoditized pillar. Most practices score 5–7 here automatically because the website builder did most of the work. The most common failure is mobile LCP, which we routinely see at 4+ seconds because of huge hero images and unoptimized hero videos.

Pillar 6 — Local Pack & Google Business Profile (15% weight)

The four visible checks:

1. GBP primary category correct + 2–4 secondary categories used
2. 30+ photos uploaded in the past 90 days, with new photos within the past 30
3. 4+ GBP Posts published in the past 30 days
4. Review velocity of 10+ over the past 90 days, with reply rate at 80%+

This is the pillar where elbow grease beats strategy. There is no schema trick that fixes a stale GBP. Practices that score 9–10 here have an internal team or an external partner who actually does the weekly photo upload and review reply. Practices that score 2–3 have an agency who promised “GBP management” and then stopped after month 2.

Pillar 7 — AI Bot Accessibility (10% weight)

The four visible checks:

1. robots.txt explicitly allows GPTBot, Google-Extended, PerplexityBot, ClaudeBot, CCBot, anthropic-ai
2. XML sitemap submitted to both Google Search Console and Bing Webmaster Tools
3. Pages render in plain HTML (not JavaScript-only) — verified via View Source and curl
4. No Cloudflare or CDN bot challenge blocking AI crawlers (verified via user-agent test)

The Cloudflare check trips up a shocking number of practices. Cloudflare’s 2024+ default WAF posture blocks PerplexityBot and ClaudeBot. If your site sits behind Cloudflare and nobody has touched the WAF rules, you are invisible to two of the highest-converting AI engines and nobody told you.

Beyond the seven pillars — the prioritized fix list

The point of the audit is not the score. The point is the prioritized fix list at the end. Every audit closes with a list of the top seven fixes, ranked by expected impact on the composite score, with a 90-day deploy plan. Most practices, executing on those seven fixes alone, can move 25 composite points in 60–90 days.

The most common top-seven we see:

1. Claim and complete the Foursquare listing
2. Deploy Dentist, Physician, and MedicalProcedure schema across homepage and top three service pages
3. Rewrite top three service pages to 1,500+ words with author bylines, “medically reviewed by” stamps, and 2026 cost ranges
4. Allow AI bots in robots.txt and fix Cloudflare WAF rules if applicable
5. Restart weekly GBP photo uploads and weekly GBP Posts
6. Add FAQPage schema to at least three service pages with patient-shaped Q&A
7. Pitch a single Healthline or Forbes Health editor for inclusion in a procedure-explainer article

None of those are expensive. Most of them, a competent in-house team can do in 30 days. None of them require a 12-month contract or a $5,000-a-month retainer. That is the version of dental marketing we believe should exist — and the version we built Thorli to deliver.

Get the audit run on your practice

We run this 50-point audit free for any US dental practice. You get a 28–34 page PDF with screenshots of every AI engine test, your scores by pillar, the prioritized seven-fix list, and a 90-day roadmap. No call required to get the report. We turn them around in 48 hours.

You can request one at thorli.com/free-audit. The audit is the audit — it is not a 15-minute lead-magnet that opens a sales conversation. It is the actual deliverable. If after reading it you want to talk, the next step is on the back page of the PDF. If you do not, you keep the audit and you go fix the things yourself. That is the deal.