Conformance report
Most statements like this one say a company is committed to accessibility and leave it there. A commitment is not a measurement. So instead of a promise, here is the score: every criterion Vitalis is held to, what we found when we checked, and the barriers still standing on this site today — named, owned and dated.
Verdict
Partially conformant with WCAG 2.2 Level AA
"Partially conformant" is a defined term, and it is the honest one. It means most of the standard is met, and some of it is not. The parts that are not are listed on this page rather than left for you to discover.
49/56
Of the 56 WCAG 2.2 Level AA success criteria in scope, 49 are met, 5 are partially met and 2 are not met.
01 — Scope
This statement applies to every page published on this domain: the homepage, the six health desks — Fitness, Nutrition, Mental Health, Healthy Recipes, Weight Loss and Sleep — every article within them, the five interactive calculators on the tools bench, and the legal pages including this one.
It was written against the templates those pages are built from, which is how a site of this shape has to be assessed. If a fault exists in the article template, it exists in every article; if the calculator shell announces its results poorly, all five do. Where a finding applies to one section only, the ledger says so.
Two things sit outside it. The first is content embedded from other services, which we do not control and cannot repair — covered in section 07. The second is anything you reach by following an external link. Once you leave this domain you are subject to somebody else's standards, and we make no claim about them.
02 — Conformance ledger
WCAG 2.2 Level AA is 56 success criteria: 32 at Level A and 24 more at Level AA. Below are the ones that carry the daily use of this site, plus every single one that does not fully pass. Criteria that pass without anything worth saying are counted in the map above and left there — printing forty untroubled rows would bury the seven that matter.
Whether the content reaches you at all — as text a screen reader can speak, at a size you can enlarge, in a contrast your eyes can separate.
Every photograph, diagram and icon carries either a written alternative or an explicit marker telling assistive software to skip it. Decorative shapes are hidden rather than described, because a screen reader announcing "abstract green shape" is noise, not access.
Headings, lists and tables are built from real HTML elements. A recipe method is an ordered list, not a stack of paragraphs beginning with numbers, so software can tell you there are nine steps before you commit to reading them.
Body text, headings and controls clear the 4.5 to 1 threshold throughout. A handful of small captions sitting on tinted panels inside the calculators measure closer to 4.3 to 1 — legible, but under the line. Logged as DEF-06.
Text is sized in relative units, so raising your browser or operating-system font size enlarges the page instead of being overridden by it.
At 400% zoom, or on a 320-pixel-wide screen, the page reflows into a single column. Nothing requires you to scroll sideways to finish a sentence.
Buttons, inputs and focus rings pass. Some of the finer rules and axis ticks inside the hydration and sleep instruments fall below 3 to 1 against their background. They are drawn as texture rather than as information, but the boundary is closer than we would like.
Override our line height, letter spacing or paragraph spacing with your own stylesheet and nothing clips or overlaps. We tested this with the standard spacing bookmarklet on the article and calculator templates.
Whether you can actually drive the thing — with a keyboard, a switch, a voice, or a finger that does not land precisely.
Every link, button, field, slider and tab on the site can be reached and operated without a mouse. The five calculators were the hard case, and they are the ones we test first.
Nothing on the site captures focus and refuses to give it back. The cookie banner can be answered and dismissed entirely from the keyboard.
A skip link is the first thing in the tab order on every page. Press Tab once on arrival and it appears; press Enter and you land past the masthead and navigation, at the start of the content.
Focus moves in the order the page is written, top to bottom, left to right. We do not reorder the tab sequence with positive tabindex values, which is the usual cause of focus jumping somewhere unexpected.
Focus rings are never suppressed. Every interactive element on the site draws a visible outline with an offset when it takes keyboard focus, including on dark panels where a default ring would disappear.
The header sticks to the top of the window, and on a short viewport it can sit over a link that has just received focus after a jump. The element is still active and still announced, but you may not see it. Logged as DEF-05.
Buttons, nav items and calculator controls meet the 24-pixel minimum comfortably. Some inline citation links inside article prose do not, because they sit in a line of running text. Logged as DEF-04.
Nothing on this site flashes. There is no auto-playing video, no carousel that advances on its own, and no animation that repeats indefinitely.
Whether the language, the layout and the error messages behave the way you expect them to, on every page.
The page declares English in its markup, so a screen reader reaches for an English voice and pronounces "read" and "lead" the way you meant them.
The header, footer and search sit in the same place on every page of the site. Learning the Fitness desk teaches you the Sleep desk.
When the newsletter form rejects an address, the message is shown in text and colour beside the field — but it is not yet wired to the input in a way assistive software can follow. A screen reader user is told the submission failed without being told which field caused it. Logged as DEF-03.
Every input has a permanent visible label rather than a placeholder that vanishes the moment you type. Where a field needs units or a range, that guidance sits next to it instead of in a tooltip.
Whether the page still works when it is read by software rather than looked at by a person.
Standard controls report themselves correctly. The range sliders in the calculators are the exception: they announce a bare number where they should announce "82 kilograms". You can still set them accurately, but you have to already know what the number means. Logged as DEF-02.
This is the most consequential thing on the sheet. When you change an input, the calculators recompute instantly and the result changes on screen — silently. A screen reader user gets no announcement and has to navigate back to the result to discover it moved. Logged as DEF-01.
03 — Known defects
These are the barriers we already know about. Publishing them is not an apology — it is the only way you can tell whether this statement was written by someone who actually opened the site with a screen reader. Each has an owner inside Vitalis and a date. If a date passes without the fix landing, the entry stays here and the date moves, visibly.
DEF-01
Result panels will be given a polite live region so the new figure is spoken as soon as it settles, without stealing focus from the control you are still adjusting.
DEF-02
Each slider will carry a human-readable value text, so it reads as "82 kilograms" or "7 hours 30 minutes" rather than "82" or "450".
DEF-03
The error text will be programmatically associated with the input and the field marked invalid, so the message is read out as part of the field rather than stranded beside it.
DEF-04
Reference links inside running prose will get additional padding and spacing so they can be hit reliably on a touchscreen without disturbing the line height of the paragraph.
DEF-05
Scroll margin will be applied to focusable targets so anything receiving focus is scrolled clear of the fixed masthead rather than under it.
DEF-06
The muted foreground token will be darkened for use on tinted surfaces, and the finer instrument rules lifted to clear 3 to 1.
DEF-07
Everything on this sheet is our own assessment, which means our own blind spots are in it. We intend to commission a third-party review and publish the report alongside this statement.
04 — Keyboard
Nothing on this site needs a pointing device. If you navigate by keyboard, by switch, or by speaking commands to your computer, this is the whole control scheme — there are no custom shortcuts to memorise, because inventing new ones tends to collide with the ones your screen reader already owns.
| Key | What it does | Notes |
|---|---|---|
| Tab | Move forward one controlPress it once on arrival to reveal the skip link before anything else. | Press it once on arrival to reveal the skip link before anything else. |
| Shift + Tab | Move back one controlRetraces the same order in reverse. Nothing is skipped either way. | Retraces the same order in reverse. Nothing is skipped either way. |
| Enter | Follow a link, press a buttonWorks on the skip link, the nav, and every card on a desk page. | Works on the skip link, the nav, and every card on a desk page. |
| Space | Press a button, or scroll a screenfulOn a checkbox it toggles. On open page background it scrolls down. | On a checkbox it toggles. On open page background it scrolls down. |
| Escape | Close the cookie banner or an open menuFocus returns to whatever you were on before it opened. | Focus returns to whatever you were on before it opened. |
| Arrow keys | Nudge a calculator slider one stepLeft and down decrease, right and up increase. Faster and more precise than dragging. | Left and down decrease, right and up increase. Faster and more precise than dragging. |
| Home / End | Send a slider to its minimum or maximumUseful for finding the range of a calculator before you set it. | Useful for finding the range of a calculator before you set it. |
One caveat worth knowing before you start: the calculators will respond to your keys correctly, but they will not yet tell a screen reader that the answer has changed. That is DEF-01 above, and it is the first thing on the repair list.
05 — By default
The cheapest accessibility work is the kind you never have to do twice. These are structural choices baked into the templates and the stylesheet, which is why they apply to an article published next year as surely as to this page.
Real text everywhere. Nothing on this site is a picture of words, so every sentence can be enlarged, recoloured, selected, translated or read aloud.
One first-level heading per page, and a heading structure that never skips a level, so you can navigate by headings and get an accurate table of contents.
Body copy set at sixteen pixels or larger with generous line height, because the single biggest readability win is not a feature — it is type that was set properly to begin with.
Motion honours your system setting. If you have asked your device to reduce motion, every animation on every desk stops. This is written into the stylesheet for each section, not bolted on.
Colour is never the only signal. Status on this very page is carried by a written word and a distinct shape as well as a hue.
Landmarks are real. Header, navigation, main and footer are marked up as such, so assistive software can jump straight to the part you want.
06 — Method
Automated tooling was the starting point, not the finding. Scanners catch missing alternative text and thin contrast reliably, and they are useless at the questions that actually matter — whether an alternative is a good description, whether the focus order makes sense, whether a result you cannot see was announced to you.
So each template was also walked by hand: through the whole page on the keyboard alone, at 400% browser zoom, at a 320-pixel viewport width, with a custom text-spacing stylesheet applied, and then listened to with NVDA on Firefox under Windows and with VoiceOver on Safari under macOS and iOS. The calculators were tested last and hardest, because they are the only part of the site that computes rather than reads, and they are where every remaining high-severity defect was found.
The limitation is straightforward and worth stating plainly: this is a self-assessment. We have not been audited by an independent specialist, and we have not published a VPAT. Testers who rely on assistive technology every day find things that sighted developers running a screen reader for an afternoon simply do not. That gap is logged as DEF-07, and it is the reason this statement carries a review date rather than a completion date.
07 — Beyond our control
Where Vitalis embeds something from another service, that component arrives with whatever accessibility its own makers gave it, and no amount of care on our side can rewrite it. We keep the number of such embeds deliberately low for exactly this reason — the site loads no advertising, no social widgets and no third-party trackers, which removes the most common source of inaccessible foreign markup before it arrives.
When we do find that an embedded component blocks people, our order of preference is: report it to the supplier, provide the same information in accessible text on the page beside it, and if neither closes the gap, remove the component. A chart nobody can read is not worth the pixels it occupies.
08 — Feedback
If something on this site blocked you, we want to hear about it, including — especially — if it is not on the list above. Reports from people using the site in ways we cannot simulate are worth more than another internal pass.
Email ahmadahmii0.0.9@gmail.com with Accessibility in the subject line.
Helpful if you have it, but never required: the address of the page, what you were trying to do, and what your device or assistive software did instead. A single sentence is a valid report. Do not spend your afternoon writing us a bug ticket.
We aim to reply within five working days. If the fix is quick we will make it and tell you when it is live; if it is not, we will tell you honestly what it involves, add it to the register above with a date, and — where you need the content now — send you the information in a format that works for you.
09 — Status of this document
This statement was prepared on and reflects the site as it stood on that date. It is reviewed every six months and rewritten whenever a defect closes, a new one is found, or a significant new template ships. The next scheduled review is 12 February 2027.
Numbers on this page are not decoration. When a fix lands, the count in the map changes and the entry leaves the register. If you return in six months and nothing has moved, that tells you something true about us — which is rather the point of publishing it this way.