Key findings
29.8%
of orthodontic practice websites serve an llms.txt file with real content. Adoption tracks the platform: 92% of Wix sites have one, compared with 2% of Squarespace sites.
2.2%
of orthodontic websites block a major AI crawler such as GPTBot or ClaudeBot. Most blocks come from one vendor's defaults: 45 of 46 ProSites-built sites block OpenAI's GPTBot.
22.5%
of orthodontic websites still run Universal Analytics code, more than three years after Google's Universal Analytics stopped collecting data on July 1, 2023.
14.9%
of orthodontic websites have no Google Analytics at all.
24.5%
of orthodontic websites with an appointment or consultation page carry Meta (Facebook) Pixel code on it. In a browser test of 38 such pages from the study sample, the pixel fired on 30.
59.6%
of orthodontic websites run on WordPress, and 17.3% are built on a dental- or orthodontic-specific platform such as Sesame, Rooster Grin or ProSites.
31.0%
of orthodontic websites load an accessibility overlay widget, mostly UserWay.
40.1%
of orthodontic websites carry Google Ads conversion tracking, and 16.2% use call-tracking software.
Figures are shares of orthodontic practice websites unless marked otherwise. Each figure's base excludes sites it could not measure.
AI readiness
Platforms decide whether a site talks to AI
Website builders and SEO plugins now create llms.txt files and robots.txt rules by default. Whether an orthodontic website has them depends far more on its platform than on the practice.
llms.txt is a proposed standard: a plain-text summary of a website written for large language models. 29.8% of orthodontic websites serve one with real content, and 1.3% also serve the longer llms-full.txt. No major AI provider has committed to reading llms.txt, so adoption shows readiness, not proven benefit.
Sites with an llms.txt, by platform or SEO plugin
- Duda69 sites97.1%
- Wix88 sites92.0%
- WordPress + All in One SEO84 sites82.1%
- WordPress + Rank Math328 sites44.2%
- Rooster Grin56 sites37.5%
- WordPress, none of the three429 sites26.3%
- WordPress + Yoast SEO886 sites25.1%
- Sesame194 sites2.6%
- Squarespace84 sites2.4%
320 of 507
orthodontic llms.txt files name the tool that generated them: Yoast SEO, Rank Math, All in One SEO or Wix. The rest carry no generator name.
45 of 47
PBHS sites serve an llms.txt that is an empty placeholder. We count those as not having one.
AI crawlers in robots.txt
Only 2.2% of orthodontic websites block a major AI crawler from the whole site. Again, platform defaults drive the number.
Orthodontic sites blocking each crawler
- ClaudeBot1.9%
- GPTBot1.8%
- PerplexityBot1.4%
- Google-Extended0.8%
- CCBot1.9%
- Bytespider2.5%
ProSites sites that block OpenAI's GPTBot
- 45 block GPTBot, ClaudeBot and PerplexityBot
- 1 do not
ProSites ships a robots.txt that allows named search engines, such as Google and Bing, and blocks every other crawler. Of the 69 sites in the sample that block GPTBot, 45 are ProSites sites. Officite blocks only ByteDance's crawler, on 56 of 57 sites.
Separately, 6 sites in the sample block Googlebot itself, which keeps them out of Google search.
Structured data
54.8% of orthodontic websites include schema.org markup that describes a local business, and 44.9% use a dental or medical type such as Dentist. Search engines and AI assistants use this markup to identify a practice, its location and its services.
Privacy
Tracking on appointment pages
About one in four orthodontic websites with an appointment or consultation page carry Meta Pixel code on it. In a browser test, the pixel fired on 30 of 38 pages.
We found an appointment or consultation page on 1,355 orthodontic sites, and 24.5% of them carry the pixel there, either in the page itself or in a Google Tag Manager container the page loads.
Browser test: 38 appointment pages with pixel code, from the study sample
- 30 pixel fired
- 2 pixel loaded, no event
- 3 cookie-consent banner showing
- 3 did not load
What the test does and does not show
- The pixel sent a standard page-view event. We did not fill in or submit any forms, so what it sends after a patient submits a request was not tested.
- Few practices use privacy tooling: 9.2% have a cookie-consent tool and 1.9% use analytics marketed as HIPAA-compliant.
- This study reports what websites load. It does not assess whether any practice complies with HIPAA or any other law.
The U.S. Department of Health and Human Services has warned healthcare providers about tracking technologies on their websites. On June 20, 2024, a federal court in Texas vacated part of that guidance in American Hospital Association v. Becerra, and HHS withdrew its appeal on August 29, 2024. The risk for a practice depends on what a page collects and what the pixel sends.
Analytics and advertising
Three in four run GA4. One in five still run code Google retired in 2023
22.5% of orthodontic websites still run Universal Analytics code, which has collected nothing since July 1, 2023. 14.9% have no Google Analytics at all.
Orthodontic sites using each tool
- Google Analytics 476.6%
- Google Tag Manager48.8%
- Google Ads conversion tracking40.1%
- Meta (Facebook) Pixel22.9%
- Universal Analytics still running22.5%
- Call tracking16.2%
- No Google Analytics14.9%
- TikTok Pixel1.3%
Hidden inside Tag Manager
Many tools load only from inside a Google Tag Manager container, so a scan of page code alone misses them. Across all sites in the sample:
- CallRail80 of 309
- PatientPrism68 of 87
- Microsoft Clarity83 of 200
Gold: installs found only inside Tag Manager.
"Still running" means a hardcoded Universal Analytics snippet, or a live Universal Analytics tag in Tag Manager. Counting any leftover Universal Analytics ID raises the figure to 24.0%. The code does no harm, but it points to sites whose tracking nobody has reviewed since 2023.
Website platforms
WordPress runs most orthodontic websites
WordPress runs 59.6% of orthodontic websites. Dental- and orthodontic-specific platforms hold 17.3%, led by Sesame.
- WordPress 59.6%
- Dental and orthodontic builders 17.3%
- Wix, Squarespace and Duda 7.7%
- Other or not identified 15.4%
Platform share, orthodontic sites
- Sesame176 sites10.5%
- Rooster Grin55 sites3.3%
- Wix51 sites3.0%
- Squarespace42 sites2.5%
- Duda37 sites2.2%
- Adobe Experience Manager17 sites1.0%
- TNT Dental16 sites1.0%
- ProSites16 sites1.0%
- PBHS10 sites0.6%
- Officite8 sites0.5%
- Custom JavaScript builds24 sites1.4%
- Other platforms44 sites2.6%
- Not identified184 sites10.9%
Agencies credited in the footer
580 of the 1,661 orthodontic websites we could read credit a web designer or marketing agency in the footer. Many agencies add no credit, so these counts are a minimum.
- Sesame Communications172
- Neon Canvas97
- HIP Agency58
- Doctor Multimedia18
- Houmanity16
- Wonderist14
Orthodontic and dental tools
A fragmented market for booking and patient tools
27.3% of orthodontic websites use a third-party booking or lead-capture tool, and no single vendor has more than 7% of sites.
A site counts as using a tool when it loads the tool's script or embed, or links to its booking or patient portal.
Patient communication and portals
any: 41.7%
- Sesame14.0%
- Ortho2 (Edge booking and portal)9.4%
- Practice Beacon5.5%
- Weave3.6%
- Rhinogram2.1%
Booking and lead capture
any: 27.3%
- Jotform6.5%
- SmileSnap4.7%
- NexHealth4.2%
- OrthoFi3.7%
- Simplifeye3.7%
- Greyfinch3.7%
- Thread Communication2.8%
- LocalMed2.6%
- Modento2.1%
Reviews
any: 18.4%
- Trustindex8.1%
- Podium3.0%
- Birdeye2.4%
Patient financing
any: 7.0%
- Cherry3.0%
- CareCredit2.5%
- Sunbit2.2%
Call tracking
any: 16.2%
- CallRail10.1%
- calls.net2.2%
- LeadSigma1.7%
- PatientPrism1.3%
- CallTrackingMetrics1.1%
Marketing automation and chat
any: 12.9%
- GoHighLevel8.1%
Accessibility overlays
any: 31.0%
- UserWay27.5%
- accessiBe2.7%
AI receptionist
any: 2.0%
- Truelark2.0%
All tool bars share one scale, 0–30%.
By region and state
Regions differ less than platforms do
The four U.S. regions are close on most measures. The Northeast has the lowest WordPress use, and the South the highest Meta Pixel use.
| Region | Ortho sites | WordPress | llms.txt | Meta Pixel | UA still running | No Google Analytics |
|---|---|---|---|---|---|---|
| Northeast | 238 | 49% | 32% | 23% | 23% | 19% |
| Midwest | 291 | 59% | 30% | 17% | 21% | 16% |
| South | 666 | 65% | 30% | 25% | 21% | 14% |
| West | 486 | 60% | 29% | 24% | 25% | 14% |
The state table covers the 22 states with at least 25 orthodontic websites in the sample. The other 28 states and DC count in the regional figures only. With 25 to 40 sites, a state's percentage can move by about 15 points from sampling alone, so compare states with care.
| State | Ortho sites | WordPress | llms.txt | Meta Pixel | UA still running | No Google Analytics |
|---|---|---|---|---|---|---|
| Alabama | 43 | 70% | 28% | 16% | 19% | 12% |
| Arizona | 65 | 59% | 22% | 27% | 30% | 8% |
| California | 188 | 54% | 28% | 21% | 21% | 20% |
| Colorado | 37 | 67% | 28% | 31% | 14% | 8% |
| Florida | 107 | 63% | 34% | 36% | 17% | 20% |
| Georgia | 27 | 70% | 22% | 22% | 30% | 15% |
| Massachusetts | 33 | 25% | 25% | 19% | 34% | 22% |
| Michigan | 31 | 53% | 50% | 20% | 17% | 10% |
| Missouri | 32 | 56% | 28% | 25% | 19% | 9% |
| Nevada | 39 | 77% | 46% | 23% | 23% | 13% |
| New Jersey | 48 | 46% | 33% | 17% | 24% | 26% |
| New York | 63 | 60% | 30% | 29% | 18% | 11% |
| North Carolina | 74 | 70% | 30% | 18% | 23% | 10% |
| Ohio | 72 | 52% | 27% | 16% | 37% | 18% |
| Oklahoma | 25 | 68% | 28% | 20% | 32% | 4% |
| Oregon | 32 | 72% | 34% | 16% | 38% | 16% |
| Pennsylvania | 40 | 55% | 43% | 25% | 30% | 18% |
| South Carolina | 35 | 63% | 26% | 26% | 20% | 20% |
| Tennessee | 51 | 62% | 38% | 26% | 28% | 12% |
| Texas | 165 | 68% | 29% | 28% | 16% | 14% |
| Virginia | 36 | 60% | 29% | 37% | 23% | 11% |
| Washington | 43 | 56% | 24% | 30% | 30% | 12% |
Methodology
How we measured
We crawled public pages on September 24, 2026 with a crawler that identified itself as National Orthodontic Registry research, gave a contact address, and honored robots.txt.
01
Sample
3,962 practices in the registry directory, deduplicated to 3,330 websites. A group with 14 locations on one website counts once.
02
Crawl
For each site: robots.txt first, then the homepage, llms.txt, the sitemap and up to six appointment, contact and treatment pages. 3,067 sites returned a usable page.
03
Tag Manager
We downloaded 1,611 Google Tag Manager containers and matched the tags inside. Paused tags are published without their settings and do not count.
04
Detection
Platforms from the open-source webappanalyzer fingerprints plus our own rules for dental builders. Tools from 91 vendor signatures built from what the crawl found. Commented-out code and DNS-prefetch hints do not count.
05
Orthodontic sites
A site counts as orthodontic when the directory classes the practice as orthodontic and its own homepage confirms it. In a hand check of 30, 26 were orthodontic practices, 3 dental practices with orthodontics as a main service, and 1 was not.
06
Validation
We loaded 30 random sites in a real browser. Agreement with the crawl: 28 of 30 for Tag Manager, 28 for Google Ads, 27 for the Meta Pixel, 25 for GA4, and 82% for vendor tools.
Rules worth knowing
- An llms.txt counts only if it returned plain text, not HTML, differed from the response to a random non-existent file name, and had at least 200 characters.
- robots.txt was read as crawlers read it (RFC 9309): a crawler's own rules override the catch-all rules, and a crawler counts as blocked only when it is disallowed from the whole site.
- Our first crawl pass fetched some pages that a site's robots.txt disallows. We deleted those responses, excluded them from every figure, and fixed the crawler. 57 sites (22 of them orthodontic) count only in figures read from robots.txt.
Limitations
- Most figures are minimums. Tags that load only after cookie consent or on certain actions may be missed, and server-side tracking cannot be seen from outside.
- The sites come from the registry directory, which covers every state but is not a complete list of U.S. orthodontists.
- 178 sites showed a bot challenge instead of the page and 50 blocked our crawler. We did not try to get around either, and they are excluded.
- We read up to seven pages per site, not whole websites, on one day. Websites change often.
About the National Orthodontic Registry
The registry is a directory of U.S. orthodontic practices, funded in part by paid placement for practices. Building and maintaining the directory means collecting data on thousands of practices, including their websites, which is what makes a study at this scale possible. Paid placement played no part in this research. Every practice website in the directory was eligible, whether or not the practice pays, and all were measured the same way. No company named in this report paid for this research.
Data and citation
The aggregate data behind this report is free to reuse under CC BY 4.0. It contains no figures for individual practices.
Download data.csvHow to cite
National Orthodontic Registry. The State of Orthodontic Practice Websites 2026. September 2026. https://nationalorthodonticregistry.com/research/orthodontic-website-technology-2026
We plan to repeat the crawl each year with the same method. Researchers and journalists with questions about the method can reach us through the contact page.
Sources
- llms.txt proposal (llmstxt.org)
- Google Analytics Help: Universal Analytics has been replaced by Google Analytics 4
- AHA News: HHS will not appeal AHA court victory in online tracking case (Aug. 29, 2024)
- Holland & Knight: American Hospital Assn. v. Becerra (June 2024)
- webappanalyzer technology fingerprints (open source)
- RFC 9309: Robots Exclusion Protocol