Web Accessibility Audit Report

Executive summary

shop.demostore-example.com does not currently meet Level AA. We found 37 distinct issues across the six templates in scope. Six of them are critical, which in this report means a customer using a keyboard or a screen reader cannot get past them without help.

Three of those six sit directly on the path to a sale. The cookie banner cannot be dismissed with a keyboard, so a keyboard-only visitor never reaches the shop at all. The cart and quantity controls are announced only as “button”, so a screen-reader user cannot tell “remove item” from “add one more”. And when checkout rejects an address, nothing is announced — the form simply appears not to respond.

None of this needs a redesign. The whole fix list is roughly 11–15 hours of developer time, and most of it is markup: labels, accessible names, focus handling and four colour values. The visual design does not change.

Fixing these 37 issues brings the audited scope to Level AA on the retest date. It does not make the site permanently conformant: accessibility breaks again with the next theme update or the next batch of product photos. That is what the monthly re-check is for.

Result summary

The score is EAA Comply’s own weighted composite of severity, frequency and where in the buying journey each issue sits. It exists to track progress between audits. It is not a standard measure — WCAG has no score, only pass or fail per criterion.

Issues by severity

Conformance by success criterion (extract)

The delivered report covers all 56 Level A and AA criteria of WCAG 2.1 and 2.2, each with the evidence behind the verdict and criteria marked not applicable where the site has no such content. This is a representative extract.

The fix plan, in the order we would do it

Ordered by customer impact divided by effort, not by severity alone. F-02 is one hour of work that unblocks twelve controls, so it goes ahead of slower items of the same severity.

Findings in detail

Six of the 37 findings are reproduced below in the format the full report uses. Every issue is documented this way, with the corrected code written for your stack on the Guided fixes tier.

<div class="cc-bar" tabindex="-1">
  <span class="cc-ok" onclick="acceptAll()">Accept</span>
</div>

<div class="cc-bar" role="dialog" aria-modal="true"
     aria-labelledby="cc-title">
  <h2 id="cc-title">Cookies</h2>
  <button type="button" class="cc-ok">Accept</button>
  <button type="button" class="cc-no">Reject non-essential</button>
</div>

<button class="qty-up">
  <svg viewBox="0 0 16 16">...</svg>
</button>

<button type="button" class="qty-up" aria-label="Increase quantity">
  <svg viewBox="0 0 16 16" aria-hidden="true" focusable="false">...</svg>
</button>

<input type="text" name="postcode" placeholder="Postcode">

<label for="ship-post">Postcode</label>
<input type="text" id="ship-post" name="postcode"
       autocomplete="postal-code">

<div class="err"></div>
<input type="email" id="mail" class="is-error">

<input type="email" id="mail" aria-invalid="true"
       aria-describedby="mail-err">
<div id="mail-err" class="err" role="alert">
  Email address: enter an address in the form name@example.com.
</div>

.price--sale { color: #e05c5c; }   /* 3.59:1 on white */
.site-foot a { color: #8b95a5; }   /* 2.86:1 on #f6f9fc */

.price--sale { color: #c0392b; }   /* 5.44:1 on white */
.site-foot a { color: #5a6678; }   /* 5.51:1 on #f6f9fc */

<img src="/img/product-img-04.jpg" alt="product-img-04.jpg">

<img src="/img/product-img-04.jpg"
     alt="Navy wool scarf, folded, showing the ribbed weave">

What automated testing found, and what it did not

axe-core found 14 of the 37 issues. The other 23 came from testing by hand with a screen reader and a keyboard. That ratio is normal, and it is the reason a scan is not an audit: automated rules are good at what a machine can measure — a missing attribute, a contrast ratio, a duplicate id — and blind to whether a name is meaningful, whether focus can escape a dialog, or whether an error message actually helps. F-01 and F-06 are the clearest examples here: one is invisible to the scanner, the other is recorded by it as a pass.

Scope, limits and what we did not test

• •The payment step is a hosted iframe supplied by the shop’s payment provider. We verified that focus enters and leaves it correctly; its internal markup belongs to the provider and cannot be changed by the shop.
• •PDF invoices and the two marketing videos are outside the scope of this report. Both fall under the EAA and can be quoted separately.
• •The findings describe the site as it stood between 24 and 27 June 2026. Theme updates, new plugins and new product content can reintroduce issues after that date.
• •This is a technical assessment, not legal advice. EAA Comply is an independent audit service, not a public authority and not a certification body.

What happens next

• →Your developer works through the fix plan. All of it is markup, CSS and focus handling — nothing here needs a redesign.
• →Tell us when it is done and we retest the same scope. The retest is included.
• →Once the scope passes, the paid tiers issue the technical conformance certificate. For this fictional shop that is sample certificate EAAC-2026-0428-17A, which is the other half of this example.
• →Your accessibility statement is written from these findings and updated at every retest. An extract of the draft is below.

See the matching sample certificate →

Your accessibility statement (draft extract)

The delivered statement is complete, dated and ready to publish; this extract shows the parts that come out of the audit. It says “partially conformant” because on the day of the audit that was true. It is rewritten at the retest.

© 2026 EAA Comply — ← Back to eaacompliance.com · hello@eaacompliance.com