An accessibility statement that claims conformance it does not have is worse than none — it is a false representation to the people it is supposed to help. What follows is the measured state of the software today.
We want Okount to be usable by every shopkeeper, including those who use a keyboard rather than a mouse, a screen reader, larger text, or who do not distinguish certain colours.
Standard we are working to
Web Content Accessibility Guidelines (WCAG) 2.2, Level AA.
How well we currently meet it
Partially conformant. Parts of the software do not yet meet the standard. They are listed below rather than summarised, because a list you cannot check is not useful.
What does not work yet
Keyboard
- There is no visible focus indicator on most controls. Buttons, chips, list rows, tabs and icon buttons give no sign of where keyboard focus is. Text fields do. This fails WCAG 2.4.7.
- Dropdowns cannot be operated by keyboard. They open on tap or click only, with no type-ahead, arrow-key or Enter support. Because the app uses dropdowns in many places by design, this currently makes parts of it slower by keyboard than by mouse — the opposite of what was intended.
- Dialogs and sheets do not manage focus. Focus does not move into them when they open, is not kept inside them, and does not return to what opened them.
Screen readers
- Most icon-only buttons are unlabelled. At the last count, 38 of 53 across 27 screens. A screen reader announces these as nothing useful.
Colour and contrast
- Three colours fall below the 4.5:1 minimum for text: secondary grey text (3.80:1), the warning colour on white (4.42:1), and the teal accent (2.98:1).
- Some status labels are distinguished by colour alone, which does not work for everyone.
Text size
- Some text is set at a fixed size and does not grow with the device or browser setting. Some layouts cut text off instead of wrapping when it is enlarged.
What already works
- Every screen has its own address, so browser back, refresh and bookmarks behave normally.
- The layout adapts to phone, tablet and desktop widths, and to a browser window being resized.
- Light and dark are both supported; light is the default because receipts and QR codes are designed for it.
- An unfinished bill survives a page refresh rather than being lost.
What we are doing about it
The items above are being worked on in this order: visible focus indicators, then focus handling in dialogs, then making dropdowns fully keyboard-operable, then labelling every icon button, then the contrast and text-size fixes.
Target date for meeting WCAG 2.2 Level AA in full: 30 June 2027. The focus-indicator, dialog-focus and icon-labelling work above is scheduled to land before 31 March 2027, with the remaining contrast and text-size fixes after it.
The legal deadline has already passed. The European Accessibility Act has applied to services of this kind since 28 June 2025. This software does not meet it today, and the date above is a genuine engineering estimate rather than a compliant one. Anyone relying on this statement should read it as notice of a current failure, not as a plan that is on time. Offering the app in the EU before the work is finished carries real enforcement risk, and that is a decision to take deliberately.
Telling us about a problem
If something here stops you using the software, we want to hear about it — including anything not on the list above.
Write to accessibility@okount.com. Tell us what you were trying to do, what got in the way, and what you use to read the screen if you would like to say. You will get a reply within 5 working days, and an answer about a fix within 20 working days.
If the reply does not satisfy you, you may take the matter to the enforcement body for accessibility in your own country; in the EU each member state names one.
This mailbox is not live yet. The address is published here so it does not have to change later, but it is not yet receiving mail. Under the European Accessibility Act a working contact route is mandatory, so this must be a real, monitored inbox before the app is offered in the EU — a published address that bounces is worse than none.
How this was assessed
By reviewing the software's own code and measuring colour contrast directly. It has not yet been tested with a screen reader by someone who uses one daily, and it has not been audited by anyone outside the team. Both should happen before any claim here is strengthened.