Finnish municipal front pages, September 2026
We scanned the front page of every one of Finland's 308 municipalities with an automated accessibility check. This is what scan results look like, still.
The numbers
- 299 / 308 municipalities scanned
- 9 municipalities whose site showed the scanner a bot challenge instead of the page
- 79 % (237) of the scanned municipalities had at least one automated WCAG failure on the front page
What was found, by criterion
Share of scanned front pages where the automated check found at least one failure against the criterion. One page can appear on several rows.
| WCAG 2.1 criterion | What it means | Who it affects | Share of front pages | Elements in all |
|---|---|---|---|---|
| 4.1.2 Name, role, value | A button, link or field has no name a screen reader can read, or its role is marked wrong. | Screen reader and voice control users | 58 % (172) | 1458 |
| 2.4.4 Link purpose | The link text does not say where it leads ("read more", an empty image link). | Screen reader users who browse by link list | 37 % (110) | 940 |
| 1.4.3 Contrast | The contrast between text and background is below the required 4.5:1. | Practically everyone; essential for people with low vision | 34 % (102) | 2070 |
| 1.3.1 Info and relationships | Headings, lists or tables are made by looks rather than structure, so assistive technology does not recognise them. | Screen reader users | 20 % (61) | 677 |
| 1.1.1 Text alternatives | An image has no text alternative, or the alternative is a file name. | Blind and low-vision people | 15 % (45) | 326 |
| 1.4.4 Resize text | Text does not grow, or is cut off, at 200 % zoom. | People with low vision | 4 % (12) | 24 |
| 1.4.1 Use of colour | Information is carried by colour alone, for example a link differs from text only by colour. | Colour-blind people | 4 % (11) | 40 |
| 1.3.5 Identify input purpose | A form field does not tell the browser what belongs in it (name, email), so autofill does not work. | People for whom typing is hard | 0 % (1) | 4 |
| 2.1.1 Keyboard | A function cannot be operated with the keyboard. | People who cannot use a mouse | 0 % (1) | 2 |
| 2.4.2 Page titled | The page has no title that says what page it is. | Screen reader users, and anyone with many tabs open | 0 % (1) | 2 |
The rulebook: what each criterion requires →
Site-level checks
These checks are made by trying things in a browser (the Tab key, a 320 px width, zoom). The verdict is "likely fails", never a certain failure: a person has to look.
| Check | What it means | Who it affects | Likely fails | Check by hand | Looks OK |
|---|---|---|---|---|---|
| Page at 320 px | The page needs sideways scrolling on a narrow screen or at 400 % zoom. | People with low vision who zoom, and users of small phones | 26 % (79) | 0 | 220 |
| Focus visible | When moving with the keyboard, it is not visible where the focus is. | Keyboard users | 18 % (54) | 96 | 148 |
| Everything reachable with the keyboard | The page has controls the Tab key never reaches. | People who cannot use a mouse, and screen reader users | 10 % (29) | 86 | 184 |
| Label in name | A button's visible text is not part of its technical name, so voice control cannot find it. | Voice control users | 34 % (43) | 0 | 85 |
| No keyboard trap | Focus got stuck in one element, for example a chat window. | Keyboard and screen reader users | 1 % (4) | 26 | 268 |
| Pause, stop, hide | Automatically moving content (a carousel, a video) cannot be paused. Movement interrupts a screen reader and makes the page harder to make sense of. | Screen reader users, blind and low-vision people, and people who find it hard to concentrate next to motion | 3 % (10) | 0 | 289 |
By region
Municipalities in the region, those scanned, and of those the ones with at least one automated failure on the front page. In small regions a few municipalities move the percentage a lot.
| Region | Municipalities | Scanned | At least one failure |
|---|---|---|---|
| Ahvenanmaa | 16 | 16 | 12 (75 %) |
| Etelä-Karjala | 9 | 9 | 7 (78 %) |
| Etelä-Pohjanmaa | 18 | 18 | 14 (78 %) |
| Etelä-Savo | 11 | 11 | 11 (100 %) |
| Kainuu | 8 | 8 | 7 (88 %) |
| Kanta-Häme | 11 | 11 | 9 (82 %) |
| Keski-Pohjanmaa | 8 | 8 | 7 (88 %) |
| Keski-Suomi | 22 | 22 | 15 (68 %) |
| Kymenlaakso | 6 | 6 | 5 (83 %) |
| Lappi | 21 | 21 | 18 (86 %) |
| Pirkanmaa | 23 | 23 | 18 (78 %) |
| Pohjanmaa | 14 | 14 | 10 (71 %) |
| Pohjois-Karjala | 13 | 7 | 5 (71 %) |
| Pohjois-Pohjanmaa | 30 | 30 | 27 (90 %) |
| Pohjois-Savo | 19 | 19 | 14 (74 %) |
| Päijät-Häme | 10 | 10 | 9 (90 %) |
| Satakunta | 16 | 14 | 11 (79 %) |
| Uusimaa | 26 | 26 | 17 (65 %) |
| Varsinais-Suomi | 27 | 26 | 21 (81 %) |
How this was done
Sample: the front page of every one of the 308 municipalities' websites, addresses taken from the municipalities' own pages in September 2026. One page per municipality in two views, desktop (1366 px) and phone (390 px).
Method: the same automated check as in Sigila's reports. WCAG 2.1 A/AA rules are checked by axe-core, the engine behind the free browser tools; the site-level checks (keyboard, layout, zoom) are made by Sigila's own scanner by trying. Scans ran at night, respected robots.txt and identified themselves as SigilaScanner.
What this is not: automated checks cover roughly a third of WCAG 2.1 AA. The rest needs a person. One front page does not describe a whole site, and a failure does not mean a municipality breaks the law; compliance is supervised by Traficom. This is a description of what scan results look like.
No municipality is named. Every municipality gets its own front page's result on request, free of charge, and a report of the whole site if it wants one.
Your municipality's result
A municipality's web owner gets their own front page's result free of charge. Ask through the support form, naming the municipality and your work email.
Request your municipality's result →
Open data
Criterion, check and region level figures as a CSV file (CC BY 4.0, source: Sigila). Per-municipality results are not published.
For journalists and organisations
Questions are answered in writing through the support form; the contact person is Pauliina Härkönen. The figures may be used with attribution.
Computed 22 September 2026 · Sigila