Almost one in four of the 200 dental websites we checked did not open

How we did it

200 sites out of the 278 drawn from across the country

The query to the open OpenStreetMap database covers all of Ukraine, not selected cities. It lists 2,964 dental practices, but only 293 give a website address — roughly one in ten. A Ukrainian directory added 39 more records, making 332 candidates. After removing social-media pages, directories and domains outside Ukraine, 278 were left, and we checked 200 of them. So 278 is almost every dental practice whose website the open map knows. That is not the same as every dental practice in the country with a website: the map fills that field far from everywhere.

• 2,964Dental practices on the open OpenStreetMap database, all of Ukraine
• 293Of those, give a website address — roughly one in ten
• 332Plus 39 records from a Ukrainian directory
• 278Left after removing social pages, directories and non-Ukrainian domains
• 200Checked
• 121Opened and allowed the full check

What was intended and what came out

The proportion could not be held: only 39 eligible sites were found in smaller towns against the 60 required, and the shortfall was filled from the large cities. The sample therefore tilts towards large cities, which should be kept in mind when reading any figure below.

Every percentage on this page is counted over all 200 checked sites unless stated otherwise beside it. Where a figure can only come from the sites that answered, we say so — "95 of 116", for instance. A site we could not reach is counted neither as having the problem nor as free of it: nothing is known about it.

Speed

It is not the hosting

We measured the time until the main block of the page appears — the large heading or the photo at the top. That is the LCP metric, and the 2.5-second threshold is Google’s. The measurement itself is made on our request by Google PageSpeed Insights: mobile profile, five runs, median taken. These are lab figures, not the real-visitor data Google collects separately in its own reports — the two are usually close but not identical.

• 8,816milliseconds to the main contentthat is 8.8 seconds — three and a half times over the limit
• 287milliseconds for the server to answerthat is fast — the server is fine
• 59.5points out of 100 for speedGoogle's score for mobile

The same picture can be shown another way. Counted not over all 200 but only over the 116 sites whose speed we could measure at all, the slow ones number 95 out of 116 — that is 82%. We kept the first way on the page because everything else is counted that way too, but there is no sense hiding the second: both are correct, they simply answer different questions.

The server answers in a quarter of a second, yet the main content takes eight. So the time goes not into the server's reply but into what happens after it — heavy images, fonts and scripts. What exactly is to blame on any one site only a separate diagnosis can show.

Technical risk

Almost a quarter of the sites simply do not open

• Served no working home page

  In 35 cases the domain has no DNS record, in 2 the host is unreachable, and 10 servers returned a 404. None are blocks on our auditor — those are counted separately

  4723.5%

• Server files left open

  Files that are normally closed off are reachable. In 3 cases they are of the kind that holds database passwords and keys; in the other 13 they reveal internal structure but not credentials

  168%

• Domain expires within two months

  Unless it is renewed, the site disappears and joins the very first row of this list

  115.5%

• Running without a secure connection

  The browser marks the address as not secure, and anything typed into a form travels in the clear

  52.5%

• Certificate expiring with no auto-renewal

  Fewer than 60 days left and no automatic renewal configured

  31.5%

• Signs of a break-in

  Injected text, redirects to other sites, spam

  0of 200

Visibility in search

A third of the clinics never tell search engines they are clinics

Besides the text for people, a page can carry a short description for machines: what kind of business this is, where it is, when it is open, what number to call. Google and AI assistants can read it, and it helps them understand the page correctly. It is not a condition of being in search — a site is indexed without it — but with it there is less room for the business to be described wrongly.

• No business description for search

  Address, opening hours and phone are not given in machine-readable form, so search has to pick them out of the text

  6834%

• No machine-readable description at all

  To a search engine the page is plain text with no hints about what is on it

  5527.5%

• No page description

  Google writes the caption in results itself, taking a passage from the page

  3417%

• No list of pages for search engines

  Search finds pages through links as well; a list only helps it happen sooner

  2110.5%

• Closed off from search

  An explicit rule forbids showing it in results

  31.5%

• No page title

  The line shown in search results and on the browser tab

  10.5%

Getting in touch

Someone reaches the site — and has no way to book

A number that will not dial from a phone, and a button that leads nowhere, are the cheapest problems on this whole page. An hour's work. And this is exactly where the call that advertising already paid for goes missing.

Checking ourselves

We cut everything in half and counted twice

The auditor ran over this industry twice: on 29 September with one draw and on 1 October with another, each time 200 sites from the same list of 278. The list is small, so the draws overlap — 150 sites are common to both, making 250 distinct sites in all. The headline figures above come from the 1 October run. Here we take all 250 and split them into two halves of 125 that share no site at all.

The widest gap between the halves is seven points, and on exposed server files they matched exactly.

The two halves came out close on every figure. That points to internal consistency — our auditor gives a steady result across different sets of sites. It does not show that the same picture holds across the industry: the limits are unchanged — the open map records a website for roughly one dental practice in ten, and the sample tilts towards large cities.

What we do not know

What these numbers do not tell you

1. Only those the open map knows are visible. Roughly one dental practice in ten has a website address recorded in OpenStreetMap. A practice absent from it, or with the field empty, never entered the check — so this is not a slice of the whole industry but of the part that marked itself in open data.
2. The sample tilts towards large cities. The intended proportions could not be held — the table is above, in the section on how sites were drawn. Kyiv alone accounts for 53 of the 200.
3. Dental practices only. These percentages cannot be carried over to other sectors or countries — we did not check them.
4. We looked at the home page only. Service and practitioner pages were not checked; they may be better or worse.
5. This is one particular day — 1 October 2026, with the list of clinics fixed two days earlier. Some of those 47 closed sites may have returned the next day, and some live ones may have gone down.
6. The source is OpenStreetMap, which is not an official business register. A practice absent from it, or listing no website, never entered the check.
7. Measured from a single location. In your city and on your connection the numbers will differ. The overall picture holds; the exact seconds do not.
8. An automated check is not a full audit. The auditor sees what is visible from a browser. It does not replace a manual review, and it does not prove the absence of other problems.
9. We did not count anyone's money. We have no data on these clinics' calls or revenue, so we do not claim a slow site cost anyone a particular sum.

• The data behind this study

  Every aggregate figure on this page in a single file, so it can be checked or recounted your own way.

  Download the data file

• Found a mistake?

  If you think one of these numbers is wrong, write to us. We will check it, and if we got it wrong we will correct it on this page and note the date.

  Write to us

1 October 2026 · AI X-Ray · research-xray-lab 1.2 · run Ds9oV6lZFXUw · seed 2027. These figures may be reproduced with a link to this page.