What should a dental website have? I read 55 West Michigan dentists’ sites: 40 ask for the appointment, 13 say they take new patients

The measurement: 61 practices, 7 towns, one homepage each

The source is Brave Search’s local results for “dentists <town> MI”, read 23 September 2026 — not Google’s map pack. Between 12:54 and 12:56 UTC that day I asked the Brave Search API for those results in the eight towns I track — Grand Rapids, Wyoming, Kentwood, Grandville, Walker, Holland, Muskegon and Kalamazoo — and took the first ten listings in each: the practices a search engine shows for the town, not a ranking. Walker returned no local listings at all, so it is not in the numbers. The other seven gave 70 listings. Five repeated a website already listed in the same town, three of them in Grand Rapids; three practices appear in two towns each, and two Grandville listings with different web addresses land on the same practice’s site; counting each practice once leaves 61. One listing names no website, which is not the same as the practice having none. Of the 60 websites, 55 could be read as the practice’s own homepage. Of the other 4, one answered “page not found” (HTTP 404) at the exact address the listing gives, a link carrying a tracking parameter, on the sweep and again on the retry, although the plain homepage loads; one domain no longer resolves; one refused a single fetch with “too many requests” (HTTP 429), which is a refusal, not a dead site; and one is drawn entirely by script: the HTML a crawler receives carries only its title, so a single fetch cannot read it. One more listing’s web address now redirects to a dental practice in Florida, so it is set aside as not the practice’s own site.

The 55 are what the search shows, not a hand-picked set of independent practices. Among them are a dental chain’s location page, the homepage of a health centre that offers dentistry alongside medicine, vision and counselling, a page on a national emergency-dentist referral site, and one homepage that was a “site under maintenance” notice on the day.

I fetched each homepage with curl, the way a search crawler or an AI engine reads it, and tested the HTML for the signals in the table. Two are dental: whether the page mentions new patients, and whether it mentions insurance. The rest are the ones I test on every trade’s sites, each described in the table exactly as the test reads it. A test counts words, not meaning, so I fetched every page a second time the same day and read what each test had matched. The second read matched every count, and the splits below come from that hand reading. The raw files are in the receipts.

What the homepage has (55 dentist sites, read 23 Sep 2026)	Sites	Of 55
HTTPS	55	100%
A phone viewport (built for mobile)	55	100%
A tap-to-call link	47	85%
Booking wording (“book online”, “schedule an appointment”, “request an appointment” and the like) or a scheduling service’s code	40	73%
The words “new patient” or “new patients”	39	71%
“Emergency”, “24/7” or “24 hours” mentioned	27	49%
Insurance mentioned (the word “insurance”, “PPO” or one insurer’s name)	26	47%
LocalBusiness-type schema markup (JSON-LD)	22	40%
A contact form in the page’s HTML (a form with an email or phone field, or a message box)	14	25%
Reviews shown on the page (widget, star rating or count)	14	25%
Financing or monthly payments mentioned	13	24%
A dollar figure in the text	6	11%
A warranty or guarantee mentioned	4	7%
“Free” before an estimate, inspection, quote or consultation	3	5%
An FAQ (FAQPage markup or an FAQ heading)	2	4%
“Licensed and insured”, or a number after “licence #”	0	0%

Town (23 Sep 2026)	Sites read	Asks for appointment	Tap-to-call	New patients	Insurance	Schema	Reviews	FAQ
Grand Rapids	7	5	6	5	3	3	2	0
Wyoming	7	6	6	5	6	2	2	1
Kentwood	9	6	6	7	3	2	0	0
Grandville	7	6	7	5	4	3	3	1
Holland	9	7	9	6	3	6	2	0
Muskegon	9	5	8	5	6	1	5	0
Kalamazoo	9	7	6	8	3	5	1	0
Walker	no local listings returned — not measured

Every town but Walker gave ten listings; Grand Rapids’ ten were seven practices and Wyoming’s and Muskegon’s were nine each, because some listings repeated a practice already on the list. The town rows count a practice once per town, so one listed in both Wyoming and Kalamazoo appears in both rows; the first table counts it once. Read across the towns and the pattern holds everywhere: the appointment ask and the phone link are in the majority in every town, and five of the seven towns have no FAQ at all. The towns differ most on schema, reviews and insurance: schema runs from 6 of 9 in Holland to 1 of 9 in Muskegon, reviews from 5 of 9 in Muskegon to none of 9 in Kentwood, and insurance from 6 of 7 in Wyoming to 3 of 9 in Kentwood, Holland and Kalamazoo.

The appointment ask: 40 of 55, and what is behind it

This is where dental sites part company with the trades. 40 of the 55 carry booking wording — book online, schedule an appointment, request an appointment — or a scheduling service’s code. The same test, the same day, found one on 12 of 46 West Michigan roofing sites, and the dental share was the highest of the eight trades I read that way; only plumbers and HVAC companies came near, at a little over half.

The test counts words, though, so I read all 40 by hand. 10 connect to an outside scheduling service: the button opens a hosted booking page, or the homepage loads the service’s widget. 21 keep the visitor on the practice’s own site: 14 of those buttons are labelled as a request — “Request an appointment”, “Appointment request” — 4 go to the contact page, and 3 go to a booking or appointment page I did not follow, because this read stops at the homepage. In the other 9 the wording leads nowhere on the site: five only ask in a sentence, such as a line inviting you to call the office to schedule; one button opens an email; one button has no link behind it; one asks only in the description a search engine shows; and one matched on the site builder’s own code.

A request is not a booking. A patient who books picks a time and has it; a patient who requests leaves their details and waits for the office to confirm. Both can work, but they are different promises, and the button should say which one it makes. Three of the 21 label the button “Book” and send the visitor to the contact page or to a page called an appointment request — that is the version to avoid. Which to offer is the practice’s call, not the website’s: in a housing campaign I ran, asking people to pick a time brought leads in at under a quarter of the category’s median cost, but that was a different market and not a controlled test, so for a dental practice it is worth testing rather than assuming.

What a new patient cannot find: a welcome, an insurance answer, a price

A welcome. The two dental signals show the gap most clearly. The words “new patient” appear on 39 of the 55 homepages, which sounds like plenty until you read them. Only 13 say in words that the practice is taking new patients or welcomes them, and 3 put a price on a new patient’s first visit. On the other 23 the phrase is a menu link or a form — “New Patients”, “New Patient Forms” — a patient’s review, a passing mention or, once, a free offer; one asks the question in its FAQ and answers that you should call to check. 16 never use the phrase at all. Whether a practice is taking new patients is something a person choosing a dentist cannot work out from anything else on the page, and it costs one line to say.

Only 13 of 55 dental homepages say, in words, that the practice is taking new patients.

An insurance answer. Insurance is the other dental question, and the homepages split down the middle. 26 of 55 mention it, and all 26 use the word “insurance” itself. Read by hand, 11 of those say in a sentence where the practice stands — that it takes most major plans, or that it does or does not take Medicaid — and 3 of them list the insurers they are in network with. On the other 15 the word is a menu link, a patient’s review, a “No insurance?” line or a passing mention. Of the 29 homepages without the word, three come close: one says “insurances” in its menu, which the test’s exact-word match misses, and two link to a membership plan. The other 26 say nothing about insurance, or about paying without it.

The fix is one line near the top of the homepage: the plans you are in network with, whether you take Medicaid, and what someone without insurance does. A “no” is an answer too. It saves the patient a call, and the front desk the same call.

A price. Six homepages print a dollar figure, and I read each one by hand (read 23 September 2026). Three are new-patient specials: a flat $80, $99 or $149 for the first exam and X-rays — one adds a cleaning — for patients without insurance. Two price an in-house membership plan, the practice’s own answer to having no insurance: one prints a table of plans starting at $38 a month, and one says only “starting at less than $1/day”. The sixth is a clear-aligner offer, $175 a month with $0 down, which is financing rather than a price. So 5 of 55 tell a patient without insurance what anything costs. I would not price treatment on a homepage, but the first visit and a membership plan are the two prices a patient can act on, and the case for publishing a starting price, made there for contractors, holds for both.

The gaps every trade shares: an FAQ, reviews, schema, the phone link

An FAQ. 2 of 55 have an FAQ by the test — an FAQ heading or FAQPage markup — the fewest of the eight trades I read that day. By hand it is thinner still. One of the two is a real FAQ: the first visit’s cost and insurance, children, new patients, what to bring. The other is FAQPage markup that no visitor sees, still carrying a template’s questions, such as “What do you know about my business?” Google’s general structured data guidelines say not to mark up content that is not visible to readers of the page (read 23 September 2026), and the rich result that markup was written for is gone anyway: Google’s documentation changelog records the FAQ rich result leaving Google Search on 7 May 2026 (read 23 September 2026). A chain’s location page, meanwhile, answers seven questions under question headings without calling them an FAQ; counting it, 14 homepages without an FAQ link to one elsewhere on the site. The reason to have one is the patient, not a search feature: it is where insurance, new patients, the first visit and emergencies get answered before the phone rings (why and how).

Reviews. 14 of 55 show reviews on the page — a review widget, or a printed star rating or review count — so 41 give a visitor comparing practices no rating to see without leaving. How to show them honestly is on the reviews page.

Schema. 22 carry LocalBusiness-family schema, the markup that tells a search engine what the business is and where. schema.org has a type made for exactly this — Dentist, which sits under LocalBusiness, MedicalBusiness and MedicalOrganization, and lists no more specific types (schema.org/Dentist, read 23 September 2026) — and 12 of the 22 use it; 8 use plain LocalBusiness and 2 MedicalBusiness. Of the rest, the health centre declares MedicalClinic, a medical type the test does not count, and one practice declares “Family Dentist”, which is not a schema.org type at all; the other 31 declare no business type, or only a generic organisation. It is a small job, and the schema page says how to check yours.

The phone link. 8 of 55 have no tap-to-call link a crawler can see. One of those is the maintenance notice, so seven working practice homepages leave a visitor on a phone with no number to tap (the case for the link). The typical homepage also names just one of the eight towns in this sweep; if a practice draws patients from the next town over, a line saying so is the honest way to show up for it.

What the trade words mean on a dental site

Some of the words I count on every trade’s sites mean something else here, so they stay out of the verdict. “Emergency”, “24/7” or “24 hours” is on 27 of 55 homepages, mostly as a service in the menu — emergency dentistry, sometimes same-day. Only two homepages use “24/7” or “24 hours” at all, and neither is a practice promising round-the-clock hours: one is the referral site, the other an accessibility toolbar’s “24 hours” setting. The trades’ free estimate turns up as a free consultation on 3, one of them to explain insurance benefits. The 4 matches for a warranty or guarantee are two disclaimers that make no guarantee and two slogans. Financing is mentioned on 13, mostly as a menu link. And none of the 55 says “licensed and insured” or prints a number after “licence #”: that is how contractors show credentials, and none of these practices does it that way.

Verdict: what a dental website should have

Element	West Michigan dentists (of 55)	Verdict
HTTPS and a phone-ready layout	55 and 55	Table stakes. Missing either puts you behind every practice nearby.
A tap-to-call link	47	Yes. Seven working homepages have none a crawler can see.
An appointment button	40 ask; 31 lead somewhere to book or request; 10 of those to a scheduling service	Table stakes. Say whether it books or requests, and never label a request “Book”.
“We are taking new patients”, in words	13	Yes, near the top. It costs one line.
Your insurance position	26 mention insurance; 11 say where they stand	Yes: the plans you are in network with, Medicaid yes or no, and what to do without insurance.
A price for the first visit or a membership plan	5	Yes, if you offer one. They are the prices a patient without insurance can act on.
Reviews on the page	14	Yes. A widget, or a rating and count linked to your Google profile.
An FAQ a patient can read	2	The rarest item on the list. Answer insurance, new patients, the first visit and emergencies.
Dentist schema	12 (22 with any business type the test counts)	Yes. Cheap, and schema.org has the type.
Emergency care, if you offer it	27 mention the word	Say what you actually offer: same-day slots, hours, who answers after hours.
The towns your patients come from	median 1 of 8 named	Name the ones that are true; a page per town only with something true to say.

What this reading does not prove

• Brave’s listings, not Google’s map pack. These are the dentists one search engine lists for each town on one day. They are not a ranking and not every practice in the region, and they include a chain’s location page, a health centre’s homepage, a referral site’s page and a maintenance notice.
• Homepages only, HTML only. An insurance page, an FAQ or a booking page one click away, or anything drawn by script after the page loads, is invisible here — one practice’s homepage is drawn entirely that way and could not be read at all. Every count is a floor, not a ceiling.
• Words, not substance. “Mentions insurance” means the word is on the page, not that the practice takes your plan. I did not follow the 21 on-site appointment pages to see which of them book and which only request.
• Correlation, not cause. Nothing here says which element wins patients. It says what a person comparing these sites will and will not find.
• Websites, not dentistry. This reads what the homepages say. It says nothing about the care any practice gives, and nothing on this page is advice about dental care.

How we build it

A dental site we build starts from this list rather than a template: the baseline, an appointment button that says whether it books or requests (linked to your scheduling service if you use one), a line on new patients, your insurance position, the price of a first visit or a membership plan if you offer one, reviews from your Google profile, an FAQ that answers what the front desk hears most, Dentist schema, and the towns your patients come from. The Starter Site is from $999 for up to three pages; the Business Site is $2,499 starting for five to eight pages, which leaves room for separate new-patient, insurance and service pages (how many pages a site needs). They are the same packages as our contractor websites, with the dental questions answered instead of the trade ones. Everything on this page applies whether or not we build yours.

---

Common questions