Lumessa

Accessibility

Last updated 20 August 2026. This statement covers every page of getlumessa.com. It does not cover the Lumessa app, which is not yet released; a statement for the app will be published before it is.

The standard we aim at

Web Content Accessibility Guidelines 2.1, Level AA.

How it was measured

By measurement rather than by eye. Every text and interface colour on this site is computed against the ground it actually sits on: the colour's transparency is composited over that ground, any transparency inherited from a parent element is folded in, and the resulting contrast ratio is calculated to the WCAG 2.1 formula. Where an element can appear on more than one ground, the worse of the two governs.

Alongside that: the site was operated with a keyboard alone; checked at 320 pixels wide and at 200% zoom; and checked with the operating system's reduced-motion setting turned on.

Re-measured on 12 August 2026, when the hero recording replaced a still image; on 13 August 2026, when the privacy notice was rewritten; on 14 August 2026, when the recording was found not to start on its own in several mobile browsers; and on 15 August 2026, when a second moving thing was added to the home page and the recording moved further down it. Re-measured again on 19 August 2026, when the home page was relaid to a new design, the lock-screen drawing was rebuilt, and the recording moved beside the account of the method. Re-measured on 20 August 2026, when the two Pause controls were withdrawn from view and a menu button was added to the narrow-screen header; the keyboard, 320-pixel and reduced-motion checks were re-run that day, and the two sentences above describing the controls were rewritten after the re-run rather than before it. The 11-pixel finding was re-tested on 15 August, in the same batch, and an exception is what the re-test produced: type set in a picture of a phone was smaller than type set as content, and the sentence had been written before there was any such picture on the site. The finding was qualified rather than withdrawn then, and the qualification was retired on 19 August when the rebuilt drawing measured clean. The statement above previously said the fade was the only animation on the site. That stopped being true the moment the recording shipped, so it was measured again rather than reworded — a statement about what was found is only worth reading if it is re-found when the thing it describes changes. New paragraphs are new text, and untested text is not covered by a finding made before it existed.

What has been found

  • No text on the site is set below a contrast ratio of 4.5:1, at any size.
  • No text on the site is set below 11 pixels. From 15 to 19 August this sentence carried an exception: the drawing of a phone lock screen on the home page contained three lines of type that rendered between 7.5 and 8.9 pixels, because type inside it scaled with the picture. The drawing was rebuilt on 19 August with every size inside it set directly, none below 11 pixels — measured in the browser before this sentence was restored, not assumed from the stylesheet.
  • Every interactive element can be reached and operated by keyboard, in a sensible order, with a focus outline that is visible against its background.
  • Colour is never the only thing carrying a meaning. The question rows, for example, show their open or closed state by the direction of a marker as well as by anything else.
  • There is no horizontal scrolling at 320 pixels wide, and nothing is cut off or overlapping at 200% zoom.
  • There are three moving things on the site. A slow fade as a section comes into view is switched off entirely when reduced motion is requested, and no content or information is lost when it is. The second is a recording of the app, which sits beside the four steps of the method further down the home page and is the only recording on the site. The third is a drawing of a phone lock screen at the top of the home page, which changes the word it is showing once a second; it is a drawing rather than a recording, and it is built from the page itself rather than from a file.
  • That recording repeats, and where it starts on its own depends on the browser. Some mobile browsers will not start any recording until you have touched the page; on those, it begins at your first tap anywhere and not before. It carries a Pause control in every case, positioned directly beneath it. The control is not shown until it is reached: it appears when it receives keyboard focus, as the skip link at the top of each page does, and it is announced to a screen reader whether or not it is visible. Once pressed, the recording stays stopped.
  • The drawing carries its own Pause control, positioned directly beneath it and reached in the same way, and it is separate from the recording’s. Stopping one does not stop the other. The drawing starts on its own in every browser, because nothing in a browser restricts it the way playing a recording is restricted.
  • If you use a mouse or touch alone and want nothing on the page to move, turning on your operating system’s reduced-motion setting stops both the recording and the drawing before they start. We know this is a weaker position than a control every visitor can see, and it is a deliberate one.
  • When reduced motion is requested, neither the recording nor the drawing starts on its own. A still frame stands in place of the recording and a single word stands in place of the drawing, and either can be started from its control if you want to watch it.
  • The recording has no sound, and everything it shows is also written in words on the same page — so nothing in it is available only to someone who can watch it. The same is true of the drawing: it is announced to a screen reader as a single description and its interior is not read out, because the words in it change five times a minute and say nothing the page does not already say in writing.

What has not been assessed yet

We would rather say this than imply a completeness we have not tested for.

  • Screen-reader passes with VoiceOver and TalkBack have not yet been run against this site. That includes the recording and the lock-screen drawing, and both of their Pause controls, which have been checked by keyboard but not with a screen reader running.
  • The lock-screen drawing has not been checked on a browser that does not support container queries, which is the mechanism it uses to hold its proportions at any size. Where that support is absent the drawing is expected to render at the wrong text sizes rather than to disappear. It has not been seen in that state, and saying so is more useful than predicting it.
  • The launch-notification form has been measured for contrast, keyboard operation and labelling. What has not been tested is what happens after you submit it, because that page is served by MailerLite rather than by us.

The European Accessibility Act

The Act's requirements bite on the Lumessa app and on anything sold through it rather than on this informational website. We are aiming this site at the same standard regardless, on the view that a site that fails the test is not made acceptable by being out of scope.

If something here does not work for you

Write to hello@getlumessa.com and tell us what happened and what you were using. Every message is read by a person, and answered.