Our own check · October 2026
What we found on 100 dental practice websites
On 18 October 2026 we ran our free check on 100 independent, mostly private dental practices in England, chosen at random from the CQC register. We looked at where each came on Google, whether ChatGPT named it, and what was wrong with its website. No practice is named here.
6 in 10
weren’t in Google’s top 10 for “dentist” in their own town
57 of 94 practices. The map box isn’t counted.
3 in 10
weren’t named in any of 5 answers when we asked ChatGPT about their town
16 of the 50 practices asked about their own town.
7 in 10
had a website problem our check calls urgent or important
72 of 99 websites, such as broken links between their own pages.
With numbers this size the shares are rough, so we give them to the nearest tenth: about ±10 points on a share of 94, about ±14 on a share of 50.
6 in 10 weren’t in Google’s top 10 for “dentist” in their own town
For each of the 94 general practices, one search on a phone, from the practice’s own town or village (London practices from London, and five small places from the nearest town). The 6 orthodontists are left out: “dentist” isn’t their search.
Where each practice came
- 57 practices (about 6 in 10) weren’t in the top 10
- 15 practices (about 2 in 10) came 4th to 10th
- 22 practices (about 2 in 10) came in the top 3
What came first in each search
- Another practice54 practices
- A directory or listing (CQC, the NHS website, Facebook, a listing site)22 practices
- A dental chain or group11 practices
- The practice itself7 practices
Google showed its AI summary above the results in the searches for 21 of the 94 practices.
ChatGPT
3 in 10 weren’t named in any of 5 answers
For each practice’s town, we asked ChatGPT directly, with web search on, the question a patient might type, five times:
Which dentist would you recommend in [the practice’s town]?
How often it named the 50 practices asked about their own town
- 16 practices (about 3 in 10) named in none of the 5 answers
- 9 practices (about 2 in 10) named in 1 or 2
- 25 practices (about 5 in 10) named in 3 to 5
Where its answers came from
- dentistlocator.co.uk41 of 50
- nhs.uk34 of 50
- alldentists.co.uk34 of 50
- dentalchoices.org20 of 50
- cqc.org.uk10 of 50
Directories and the NHS website came up in its answers far more often than practices’ own websites.
Who isn’t in this figure, and why
| Practices | Why |
|---|---|
| 20 | Our check asked about a nearby town or city, not the village or town the practice is in. |
| 11 | London practices our check asked about “London” as a whole, not their own area. |
| 4 | Orthodontists, asked about dentists. |
| 14 | Our check wasn’t sure of the town or the trade, so it didn’t ask. |
| 1 | The website didn’t answer. |
A practice can’t fairly be measured against a town it isn’t in, so these are left out.
Both together
ChatGPT named the practices Google shows far more often
Of the 19 practices in Google’s top 10 for their town, ChatGPT named 17 in most of its answers. Of the 31 outside it, it named 8 in most answers, and 15 not at all. It shows the two go together in this group, not that one causes the other.
In Google’s top 10 for their town17 of 19 named in most answers
Not in Google’s top 108 of 31 named in most answers
| On Google | Practices | None | 1–2 | 3–5 |
|---|---|---|---|---|
| Top 3 | 9 | 1 | 0 | 8 |
| 4th to 10th | 10 | 0 | 1 | 9 |
| Not in the top 10 | 31 | 15 | 8 | 8 |
Their websites
7 in 10 had a problem our check calls urgent or important
Our free check read up to 100 pages of each of the 99 websites that answered (the whole site for 66), the way Google reads them.
Problems worth knowing about
- At least one urgent or important problem72 of 99
- Broken links between its own pages39 of 99
- Pages that share the same title32 of 99
- Old or placeholder pages still live30 of 99
- No Google Analytics found25 of 99
- A sitemap that lists broken or moved pages22 of 99
- Pages marked “don’t show in Google”9 of 99
Common, and less serious
Images without descriptions (93), pages without one clear main heading (80), pages without a description (74), main pages with very little text (67) and titles too long or too short (62).
Speed on a phone, with care
Google’s lab test on a simulated slow phone called 83 of 88 homepages slow, but that test is strict. Only 11 practices had enough real visitors for Google to measure, and 7 of those were slow.
Method
How we checked
- Who
- 100 independent, private-focused dental practices in England whose websites run on WordPress, drawn at random from the CQC care directory of 7 October 2026. 867 practices qualified, out of 12,300 dental locations. At most 2 a town, London by borough: 100 websites in 90 towns. Independent means an owner of at most 5 practices; private-focused, a site that sells private treatment and doesn’t mainly offer NHS care.
- When
- 18 October 2026.
- One search a practice, “dentist” and its own town or village, on a phone, from that place, through DataForSEO, a search data service. London practices were searched from London, and five small places from the nearest town its location list knows. Ordinary results only.
- ChatGPT
- The free check’s own question for the practice’s town, five times, asked of ChatGPT directly with web search on. The free check’s matching decides whether an answer names the practice.
- Websites
- Our free check, reading up to 100 pages of each site the way Google reads them.
- Checked
- For ten practices drawn at random, we opened every broken link reported, a site that answered on two addresses, and every ChatGPT answer against its verdict. Then every figure was checked again against the raw data, twice. That moved the Google figure from 68 of 100 to 57 of 94: some searches hadn’t run from the practice’s own place, and the 6 orthodontists came out.
What it doesn’t tell you
- These are independent, mostly private practices on WordPress, not all dental practices.
- Google’s results and ChatGPT’s answers vary by place, by person and by day. This is one day.
- The map box isn’t counted, so a practice outside the top 10 may still show on the map.
- The numbers are small, so the shares are rough: “6 in 10”, not “61%”.
- Google and ChatGPT going together doesn’t show that one causes the other.
- Only 50 of the 94 general practices were asked about their own town. The table under ChatGPT says why.
How to quote this
Quoting these figures
Please quote them with the group, the date, and a link to this page.
Source
Findwick, “What we found on 100 dental practice websites”, October 2026. findwick.com/research/dental-websites-2026/
- “About 6 in 10 independent dental practices Findwick checked weren’t in Google’s top 10 for ‘dentist’ in their own town (the map box not counted).”
- “Of the practices in Google’s top 10 for their town, ChatGPT named 17 in 19 in most of its answers. Of those outside it, 8 in 31.”
- “About 4 in 10 had broken links between their own pages.”
Each line is about independent, mostly private dental practices in England with websites on WordPress, checked on 18 October 2026. Questions about the data, or how we checked? Ask us through the contact form.
Is your practice one of them?
The free check shows where you come on Google for your town, whether ChatGPT names you, and what’s holding your site back.
Free, with no sign-up. It takes a minute or two, and we only read pages anyone can see. More for dental practices.