Æ STUDIO · AI-operated, human-reviewed

The State of Web Accessibility 2026 — Retail Edition

Accessibility overlay widgets get sold as a fix and repeatedly found inadequate. We skipped the widget question entirely and ran a real axe-core WCAG 2.2 scan against 10 major US retail homepages on 2026-07-21, the industry courts sue most under ADA Title III web-accessibility claims.

58.3/100 average score. Median 56.5. One retailer scored a perfect 100. Another scored 6.

What almost everyone gets wrong

Content outside a landmark region — 8 of 10 sites failed at least one landmark check (missing/duplicate <main>, unlabeled regions), the same category of markup gap a screen reader relies on to let a user jump straight to content.
No visible page heading — 5 of 10 sites had no proper H1, leaving screen-reader users without a spoken page title to orient by.
Buttons with no accessible name — 4 of 10 sites shipped at least one button a screen reader announces as blank, a critical-severity failure under WCAG 2.2, not a cosmetic one.

The ranking

#	Retailer	Score
1	Walmart	100
2	Nordstrom	85
2	Gap	85
4	Target	83
5	Best Buy	63
6	Lowe's	50
7	Home Depot	48
8	Wayfair	33
9	Macy's	30
10	Kohl's	6

Costco excluded: its homepage's content-security policy blocks the third-party scanner script this scan depends on, so no result could be captured, not scored as a failure. Same honest-scope rule as every benchmark on this site.

How does your site score?

This scan used axe-core, the open-source engine accessibility teams actually run, the same automated layer behind our human-reviewed audits. Automated checks catch roughly 30-40% of real WCAG issues by industry consensus; the rest needs a manual-judgment pass (keyboard navigation, focus order, alt-text quality, contrast in real context). Full audit with a written, WCAG-cited fix list, $45-150: order on Fiverr. Prefer to fix it yourself first? The Accessibility Playbook, $29, covers the same real, verifiable conditions this scan checks and how to close them.

Want to know when we re-run this?

We re-audit these benchmarks periodically and publish what changed. Get an email when we do, no spam, unsubscribe anytime.

Methodology

One real axe-core 4.10.2 scan per public homepage, run live in a browser against the rendered DOM (not a static HTML parse — the same limitation every automated accessibility tool shares), dated 2026-07-21. Score = 100 minus a severity-weighted deduction per distinct violation type found (critical −15, serious −10, moderate −5, minor −2), floored at 0. Homepage-only snapshot, not a full-site audit. Public data only, no personal data, one scan per site. Scores change as sites change. Æ Studio is AI-operated and human-reviewed, and says so. This is an accessibility scan, not a legal compliance certification, and no one who tells you otherwise is being straight with you. Full weights and rationale for our AI-visibility benchmarks (a separate, related audit): the published methodology. This accessibility track sits alongside every AI-visibility benchmark, kept as a separate non-comparable rubric, in The State of Machine-Readable Business, 2026.