Every med spa we measured has a Book button.
Nine aesthetic practices, measured the same way we measured roofers. The fundamentals were flawless — and the pages were carrying roughly four times the image weight and half again as many scripts.
Same measurement, different trade.
We ran the same measurement we used on roofing sites against aesthetic practices around Houston, Kingwood and Humble on 8 September 2026: one fetch per homepage, recording document weight, server response, referenced image sizes, script counts and the technical basics. We added one check specific to this trade — whether the page connects to a recognised online booking platform.
We did not measure Core Web Vitals. Those need a real browser and field data from actual visitors. Claiming them from a single fetch would be exactly the sort of unearned number this series exists to avoid.
We started with twelve sites and report nine. Three were excluded because we could not fetch them cleanly — one returned 403 to an automated request, which usually means bot protection rather than a fault; one did not connect at all; one returned 404 at its own root. None of those are safe to characterise from the outside, so they are out of the sample rather than written up as failures.
A button is not a booking.
- 01
Every site has a booking call-to-action. Four of nine have a booking system.
All nine surface a book, schedule or reserve action. But we could only positively identify an actual booking platform on four of them — every one of those was Zenoti. The remaining five either use a system our detection does not recognise, or the button leads to a contact form and a callback. For a client deciding at eleven at night, those are very different experiences.
- 02
The pages carry roughly four times the image weight of roofing sites
Median payload across the first twelve images was 489 KB, against 131 KB for the roofing sample. The heaviest single image we found was 3.1 MB. This is not surprising for a trade that sells on visual results — but it is a cost being paid on every visit, including the ones on mobile data.
- 03
And half again as many scripts
Median 60 script tags, against 41 for roofers. One site loaded 121. Booking widgets, chat, review embeds, tracking and financing tools accumulate, and each is code that must be fetched and executed before the page settles.
- 04
The technical fundamentals were flawless
All nine had HTTPS, a viewport tag, a meta description, structured data, a robots file and a sitemap. Nine out of nine on every one. Whoever builds sites for this industry knows the checklist — this is not where the problem is.
Med spas against roofers.
Two trades, same method, same week. The contrast is the useful part — it separates what is an industry habit from what is just how websites are.
| Measure | Med spas (9) | Roofers (12) |
|---|---|---|
| Median script tags | 60 | 41 |
| Heaviest site, script tags | 121 | 65 |
| Median image payload (first 12) | 489 KB | 131 KB |
| Largest single image found | 3.1 MB | 2.2 MB |
| Median HTML document | 68 KB | 47 KB |
| Median image elements on page | 39 | — |
Aesthetic practices sell on images, so heavier pages are not a surprise and not automatically wrong. What is worth questioning is whether the weight is chosen. A 3 MB photograph is rarely a decision anybody made; it is usually a file that went up at whatever size the camera produced.
What this does not establish.
Nine is a small sample, and it skews toward practices already doing enough marketing to surface in ordinary Houston searches. Each site was fetched once, from one location, on one day.
The booking figure is the one to read most carefully. Our detection recognises a fixed list of platforms. A practice using something outside that list, or handing off to a booking system on another domain, would show as no platform found even though clients can book perfectly well. The honest claim is narrow: on four of nine we could positively confirm a booking system, and on the other five we could not — not that the other five lack one.
And a homepage is not a website. Treatment pages, where most decisions actually happen, were not measured.
The two things worth checking.
Count the taps from a treatment page to a confirmed appointment. If the answer involves a form and a callback tomorrow, that is the gap — and it is not a design problem, it is a plumbing one.
Then open your own gallery on mobile data, not office wifi. If it is slow for you it was slower for the client who left. More on what aesthetic practices need is on the med spa websites page, and the free audit runs this properly against your site rather than from the outside.
Find out which of the nine you resemble.
Send your address and we will run the full twelve-check audit, including the image weight and booking path this sample says are doing the damage.