21 of 23 South Coast Dental Websites Say "Book Online." The Median One Takes 10.3 Seconds to Load.

If you run a dental practice on the South Coast, you have heard this pitch. Probably last month. Probably from someone who found your number on Google.

Patients want to book online now. Nobody calls anymore. Add online booking and you'll stop losing new patients to the practice down the road.

It is a good pitch because it sounds obviously true. So we tested it.

What we actually did#

We took 23 dental practice websites across New Bedford, Dartmouth, Fairhaven, Somerset, Swansea and Westport. For each one we ran two independent passes.

Pass one: Google PageSpeed Insights on mobile, which is Google's own Lighthouse engine, measuring Largest Contentful Paint, meaning the moment the main content actually appears on a phone screen. Google's threshold for "good" is 2.5 seconds. Over 4.0 seconds is rated "poor."

Pass two: we fetched the homepage plus up to eight likely appointment pages on each site and read the source. We looked for two different things: booking language ("request an appointment," "book online," "schedule an appointment") and actual booking software, meaning a third-party scheduling or patient-engagement vendor loaded on the page.

You could repeat this yourself in an afternoon. Everything we used is free.

Two of the 23 returned no Lighthouse result at all, so their speed is not measured and they are excluded from every speed number below. One site's server refused our structure checker outright, which means it exists and is working. It just would not talk to us. We report it as blocked, not broken.

Finding one: the button is already everywhere#

21 of 23 sites use booking or appointment-request language somewhere on the site.

Whatever you were told you are missing, your competitors are not quietly winning with it. They all have it. It is not a differentiator. It is wallpaper.

Finding two: almost none of it is real booking#

Here is where it gets interesting. Of those 21 sites talking about booking, only 6 of 23 load any third-party scheduling or patient-engagement software at all. The vendors we found, across all six, were Simplifeye, Modento, Flex, Swell and Kleer. Kleer is a membership-plan product, not a scheduler.

The other fifteen say "Request an Appointment," and what happens when you tap it is that you fill in a form, or you get a phone number. Seven of the 23 sites have no contact form on them anywhere, so on those, "request an appointment" resolves to "call us during business hours."

That is not online booking. That is a message.

The distinction matters more than the industry admits, and the best statement of why comes from the research itself:

When designing online appointment systems, it is imperative to create a true online booking application, not just an online system that operates as a messaging app, which then will be followed up by email or phone. Consumers desire more convenience, not additional steps.

A form that generates a callback tomorrow morning has not removed a step from the patient's life. It has added one. They still have to take your call.

Finding three: this is the actual problem#

Across the 21 practices where we could measure it, the median mobile Largest Contentful Paint was 10.3 seconds.

Not one site cleared Google's "good" bar. Not one landed even in the "needs improvement" band. Every single measured practice in six towns is in the poor category, and the spread runs from 4.1 seconds at the very best to 21.2 seconds at the worst.

Read the bottom two rows twice. The fastest, highest-scoring site in the entire study is the one practice that does not advertise online booking at all. It just loads.

Google ran the definitive study on this with Deloitte, across 37 brands and 30 million sessions. For lead-generation sites, the kind whose entire job is getting someone to complete a form, which is exactly what a dental site is, a 0.1 second speed improvement produced "an uplift of 21.6% from the first step of the form to the form submission page." One tenth of one second. (Think with Google, Milliseconds Make Millions, p.24.)

Now apply that to a page that takes 10.3 seconds.

The fix you can do today, free#

Do not buy scheduling software this week. Do this first, and it costs nothing:

1. Go to PageSpeed Insights (pagespeed.web.dev). Enter your homepage. Choose Mobile.
2. Scroll to Largest Contentful Paint element. It will name exactly one thing on your page. Nine times out of ten it is your hero photo, the smiling-patient shot at the top.
3. Find that image file. Open squoosh.app. Drag it in. Resize to 1600px wide, export as WebP, quality 75.
4. Send the new file to whoever maintains your site and ask them to swap it.

Twenty minutes of your time. On a site serving a 4MB camera original as its hero, and several in this study are, that one swap is routinely worth two to four seconds. It will do more for your new-patient numbers than any booking widget you could buy this month.

What we could not measure#

Being straight about the limits, because a study that hides them is marketing:

• Two of the 23 sites returned no Lighthouse result. Their speed is not measured, not "slow."
• Our software detection reads the page source. A scheduler injected later by a tag manager could be missed, so 6 of 23 is a floor, not a ceiling.
• We measured whether a booking system is present, not whether it works. A vendor script on the page is not proof a patient can grab a Thursday 4pm slot.
• We did not measure phone answer rates, and for most South Coast practices the phone is still doing the heavy lifting.

What a South Coast dental practice should do Monday#

Cheapest first, with honest time estimates.

This morning, free. Run PageSpeed Insights on your homepage and fix the hero image. Twenty minutes.

This week, free. Long-press every phone number on your site on an actual phone. Confirm each one dials your actual front desk. Ten minutes.

This month, free. Nine of the 23 sites have no LocalBusiness schema, the structured data that tells Google your hours, address and phone. Ask whoever runs your site to add it. It is an afternoon of someone's time and it is often the difference between appearing in a map pack and not.

Before you buy a scheduler. Ask the vendor one question: does a patient pick a real, open slot in my actual calendar, or does this send my front desk a message? If it is the second one, you are buying a contact form with a subscription attached. You already have a contact form.

Then, and only then, buy the scheduler. A real one, on a page that loads in under two and a half seconds, taken at 11pm by somebody with a toothache who was never going to call you in the morning. That is the version that works. The order matters.

---

Avalon Partner is in Fall River. If you want to know what your own site measures, the real numbers rather than a sales sheet, I will run it and send you the results, whether or not you ever hire me.

774.559.8992 · Joshua.Amado@AvalonPartner.com