Third-party trackers on 362 healthcare practice sites

We ran our scanner over 362 dental and physical therapy practice websites in September and October 2026 — 5,382 pages, public HTML only. If you run a small practice site, the row to look at is the third one: it is the one that separates ordinary analytics from a page where somebody books an appointment. No practice is named anywhere on this page.

How many practice sites load a tracker?

Median trackers per site: 2. 97 sites load nothing third-party at all.

Why a tracker on a booking page is different

Analytics on a home page is ordinary. The same script on the page where somebody books an appointment, asks about a symptom or fills in a new-patient form is a different thing: the address of that page says what the visit was about, and reporting the address is what these scripts are built to do.

52 of the 362 sites do that — about one in seven. That share is a floor rather than an estimate. We counted it only when the page's own address said booking, appointment, scheduling, contact, consultation, intake, portal, new-patient forms, insurance verification or referral. A form sitting on a page with a neutral address is not in the 14%.

Which finding surprises people most?

29 sites run a session recorder — Microsoft Clarity, Hotjar or FullStory. These do not count visits; they replay them, including mouse movement and, depending on how they are configured, what was typed into a field. They are usually switched on by whoever built the site, which is why they are worth checking for even on a site you are happy with.

Which trackers appear most often?

What the grades mean

97 sites load no tracker but only 72 score A: the other 25 embed a third-party widget — a booking iframe, a chat box, a map — which sends a visitor's address to another company even though it is not a tracker.

How this was measured

• Up to 25 public pages per site, read as HTML. No login, no form was ever submitted, at most one request a second per site, robots.txt respected on every host (scan bot policy).
• A script counts once per site however many pages it appears on, and a site counts once however many times we have scanned it, so neither a page-heavy practice nor a repeat customer can move the numbers.
• Tags that only fire after a click are invisible to a read of the HTML. They need a headless browser, which belongs to a paid assessment rather than this pass, so the figures are conservative in that direction too.
• The sample is not random. 360 of the 362 are practices with a website listed in OpenStreetMap around Seattle, Portland, Denver, Boise and Spokane; the rest asked us for a scan. These are the scans we completed — sites that blocked the crawler or did not answer are not in the sample, and we correct for none of that. Read it as a picture of small practice websites in the western United States, not a national statistic.

By metro and by kind of practice

• Seattle practice websites
• Portland practice websites
• Denver practice websites
• Phoenix practice websites
• Austin practice websites
• Dallas practice websites
• San Diego practice websites
• Minneapolis practice websites
• dental practice websites
• physical therapy practice websites

What we are not saying

We are not telling you what any of this means for a particular practice's obligations. TagQuiet performs technical assessments and is not a law firm. What the law asks of you is a question for your counsel, and counsel usually starts from an inventory of what actually loads.

What if your own site has one?

The same scan runs on request, free, and takes about ten minutes:check your site. You get the grade and every finding with the page it loads on, and no call from us unless you ask for one.

If it does find something, most of it is a settings change your web person can make in an afternoon, and the free result is written so you can forward it to them. If you would rather hand the whole thing over, we remove the trackers ourselves, re-scan and send you a log of what changed — what that costs is $499, or $699 with the written assessment. Nothing on your site changes without your approval in writing.

Scanned 2026-10-08. Every figure here comes from one computation over our scan table (app/benchmark.py) and moves as we scan more sites; the date says which run this page shows. Cite it as: TagQuiet, Third-party trackers on 362 healthcare practice sites, 2026-10-08, tagquiet.com/benchmark.