Dental Website Statistics 2026: What 832 US Practice Sites Actually Do

In September 2026 we measured 832 dental practice websites across 47 US states — a random sample drawn from public OpenStreetMap listings, each site read once from its homepage. The short version: 41.6% show no online booking we could detect, 74.2% never mention price, and 90.5% carry no review markup, so their star ratings cannot show in search. 34.3% fail all three at once. The full aggregates are published below and free to reuse.

What we measured, and what we did not

The sampling frame was every US dentist listing in OpenStreetMap that publishes a website — 3,687 unique domains. We attempted 1400 of them at random on 24 September 2026, one request to the homepage each, honouring robots.txt; 832 sites answered. From the returned HTML we detected scheduling platforms and booking wording, JSON-LD types, price and financing wording, click-to-call links, chat widgets, Spanish versions, the Meta Pixel, where forms post, and how long the server took to respond. No practice is named anywhere: every figure below is an aggregate. We did not judge design, copy or clinical quality, and we did not open inner pages — a practice may well sell beautifully two clicks deeper.

The numbers

Nine in ten give up their stars before anyone clicks

90.5% of the sites have no review structured data at all, so Google has nothing to turn into stars under the link — the cheapest visual advantage in a local result. 55.4% skip local-business markup as well, 29.0% publish no structured data whatsoever, and only 7.3% have both pieces in place. 35.8% do not even link their Google Business Profile from the site, which is the one signal that ties the page and the map listing together. None of this is a redesign: it is a day of work on markup and a link, and it is exactly what local SEO for dentists is built on.

Four in ten cannot take a booking, and a third only look like they can

A booking platform is on 27.4% of sites. Another 31.0% show a “book now” link or a contact form — which is a request, not a booking, and it waits for someone to open it. On top of that, 21.2% have no click-to-call phone link, so a patient on a phone has to copy the number by hand, and only 9.9% run any chat or messaging widget. That gap is where new patients quietly go to a competitor: our own take on it is an AI front desk that answers calls and messages, and the site-side fixes are in medical website design that converts.

Three in four never mention money

74.2% of homepages contain no price, no starting-from figure and no financing wording. Patients treat that as a signal: the ones who need to know the number leave to find it elsewhere, and the ones who stay arrive at the phone unqualified. Nobody has to publish a full fee schedule — a starting figure, a financing line and an honest “it depends on the case” beat silence. We publish our own prices for the same reason.

Speed: a third are fast, a sixth are slow enough to lose people

One site in five is shut to crawlers it does not recognise

21.1% of the sites we attempted refused us: either robots.txt came back as 401 or 403, which under the robots standard means “stay out entirely”, or it disallowed our crawler by name. Bot protection like this usually still lets verified Googlebot through — but it also blocks newer, less-known crawlers, and that is the class AI assistants use to read the web. Of the 621 sites that did serve a robots.txt, 13 (2.1%) block AI crawlers on purpose — GPTBot, CCBot, Google-Extended and the like. If a practice wants to be the answer inside ChatGPT or an AI Overview, step one is letting those crawlers in; the rest is in AI search optimisation for clinics.

Tracking is on 15% of sites, and a small slice sits next to health questions

The Meta Pixel is present on 15.1% of sites. On 1.7% it shares a homepage with a field that asks something about the visitor’s health — a reason for visit, a dental concern, insurance or a date of birth. That combination is a risk indicator, not proof of a violation: what matters is whether any of it reaches the ad platform. 2.2% post their forms to a third-party domain, and the word HIPAA appears on 10.5% of homepages. If you run ads at all, the platform rules on health data are worth an hour: we keep them sourced and dated in the Google and Meta healthcare ad rules.

What we would fix first, in this order

Ordered by return per hour of work, not by how it looks in a proposal:

1. Review and local-business markup, plus a link to the Google Business Profile. A day of work, and it is the difference between a plain link and a starred one.
2. Real online booking — a platform, not a form that waits for a human.
3. Click-to-call on every phone number, because most of this traffic is on a phone.
4. Speed: get the first byte under half a second and the homepage HTML under 300 KB.
5. A price or financing line, even a range.
6. Crawler access: let the AI crawlers in, or accept being absent from AI answers.

Method, limits and the raw data

Recorded 24 September 2026. Frame: 3,687 US dentist listings with a website in OpenStreetMap; 1400 sampled at random, 832 answered, 47 states represented. One homepage request per site, no mobile emulation, response time measured once from a single European location on an uncached request. Limits worth knowing: we read homepages only, so a booking widget or price page deeper in the site is invisible to us and the “no booking” share is an upper bound; detection of booking relies on known platforms and common wording; the OpenStreetMap frame skews toward practices someone bothered to map; and sites behind bot protection are missing from the answered sample, which likely skews it toward more open, more technically maintained sites. Download the aggregates as CSV or JSON. They are published under CC BY 4.0 — use any number with a link to this page, and cite it as “Tepexa, Dental Website Statistics 2026”. Ask us anything about the method through the audit form; we answer method questions from journalists and agencies too.