Accessibility overlays: what they do and what to use instead
What overlays and widgets change at runtime, the limits accessibility organizations describe, the 2025 FTC order against accessiBe, and the evidence-first alternative.
Accessible scanned 12,682 US small-business websites across seven industries. 90.1% had at least one automatically detectable WCAG failure. A separate cohort of 256 national brand sites had a detectable failure on 88.7% of sites. These are failures a free automated checker can flag. Check your own site. A page with no detected failures is not thereby accessible.
. Main Street Accessibility Report. 13,735 sites attempted, 53,464 pages tested, scanned November 9–10, 2026.
Every number on this page comes from those scans. Small business methodology · National brand methodology
90.1%
of sites had at least one automatically detectable WCAG A/AA failure
6.2
detectable failures per page on the median site
71.5%
of sites had text with too little color contrast
95.5% vs 89.4%
had a detectable failure with an overlay, against sites without one. Medians were 8.8 failures per page with an overlay and 6.0 without.
These are failures a free automated checker can flag. Check your own site.
The figures above, and the tables below, are traced in the methodology for each cohort. Small business methodology · National brand methodology
In every industry measured, at least four in five sites had an automatically detectable WCAG failure. Dental and medical practices were highest: 95.6% of 2,102 sites, with a median of 8.6 failures per page. Med spas were next, at 94.9% of 356 sites.
Restaurants were lowest, at 80.8% of 3,665 sites, with a median of 3.5 failures per page.
The HHS Section 504 rule (45 CFR 84.84 (opens in a new tab)) requires WCAG 2.1 AA for web content of recipients of HHS federal financial assistance. Compliance dates were extended by an interim final rule on May 7, 2026 (opens in a new tab). They are May 11, 2027 for recipients with 15 or more employees and May 10, 2028 for recipients with fewer than 15. The requirement is not in force yet.
| Industry | Sites | % with a failure | Median failures per page |
|---|---|---|---|
| Dental and medical | 2,102 | 95.6% | 8.6 |
| Med spas | 356 | 94.9% | 7.8 |
| Home services | 2,543 | 94.5% | 8.6 |
| Apparel and retail | 1,284 | 93.8% | 7.8 |
| Beauty and wellness | 1,573 | 92.5% | 6.4 |
| Hotels and hospitality | 1,196 | 90.9% | 5.0 |
| Restaurants | 3,665 | 80.8% | 3.5 |
| All industries | 12,682 | 90.1% | 6.2 |
1,610 sites (12.7%) load an accessibility overlay, the toolbar widget sold as an automatic fix. 95.5% of those sites had a detectable failure, against 89.4% of sites without one. The median was 8.8 failures per page with an overlay and 6.0 without.
Overlay sites had a higher median number of failures per page in all 7 industries. They had a higher failure rate in 6 of 7. Beauty and wellness is the exception: 91.0% with an overlay against 92.6% without.
Overlays are not associated with fewer failures. What an overlay changes. The figures are a correlation. They do not show that an overlay causes failures.
| Industry | Without an overlay | With an overlay | ||||
|---|---|---|---|---|---|---|
| Sites | % with a failure | Median per page | Sites | % with a failure | Median per page | |
| Home services | 2,367 | 94.1% | 8.0 | 176 | 98.9% | 16.0 |
| Apparel and retail | 1,160 | 93.5% | 7.6 | 124 | 96.8% | 10.8 |
| Med spas | 291 | 94.2% | 7.0 | 65 | 98.5% | 10.2 |
| Dental and medical | 1,512 | 94.3% | 8.0 | 590 | 99.0% | 9.8 |
| Beauty and wellness | 1,495 | 92.6% | 6.4 | 78 | 91.0% | 8.9 |
| Hotels and hospitality | 988 | 90.5% | 4.7 | 208 | 92.8% | 5.8 |
| Restaurants | 3,292 | 79.8% | 3.4 | 373 | 89.8% | 4.4 |
| All industries | 11,072 | 89.4% | 6.0 | 1,610 | 95.5% | 8.8 |
Text with too little color contrast was the most common failure in every industry. It appeared on 71.5% of all 12,682 sites. Links with no accessible name appeared on 48.3% of sites: a screen reader can only announce that link as “link”. Touch targets too small for WCAG 2.2 appeared on 22.1%.
The top three are cheap to fix and are usually fixed once in the theme or template, which corrects every page. The same check is available in the browser extension.
| Barrier | Rule | % of sites | How to fix |
|---|---|---|---|
| Text with too little color contrast | color-contrast | 71.5% | 1.4.3 guide |
| Links with no accessible name | link-name | 48.3% | 2.4.4 guide |
| Touch targets too small (WCAG 2.2) | target-size | 22.1% | 2.5.8 guide |
| Embedded frames without a title | frame-title | 20.0% | 4.1.2 guide |
| Images without alternative text | image-alt | 18.4% | 1.1.1 guide |
| Links told apart by color alone | link-in-text-block | 12.9% | 1.4.1 guide |
| Zoom disabled on mobile | meta-viewport | 10.9% | No guide page. This rule relates to WCAG 1.4.4. |
| Buttons with no accessible name | button-name | 8.9% | 4.1.2 guide |
| Lists built incorrectly | list | 6.8% | 1.3.1 guide |
| Controls nested inside controls | nested-interactive | 6.7% | 4.1.2 guide |
4 in 10 small-business sites (40.5%) had at least one page missing a main content region or a top-level heading, twice the national-brand rate (19.5%).
86.5% of 12,682 small-business sites had at least one page with a best-practice structure issue that affects screen reader navigation. National brands were close on that combined measure, at 87.5%. These figures come from automated checks of the page structure screen readers rely on. They are not WCAG failures, and they are not counted in the failure figures above. Dental and medical sites were highest, at 93.1%, and home services were at 91.1%.
Screen reader users rarely listen to a page from top to bottom. They jump between headings and page regions. See how to test with a screen reader and keyboard-only navigation. What a clear structure is for is covered in WCAG 1.3.1. The checks in this finding are separate from that WCAG failure.
National brands do the basics better and still have the same structure issues. 19.5% had at least one page missing a main region or a top-level heading, about half the small-business rate. Content outside any page region was on at least one page of 65.2% of brand sites and 64.1% of small-business sites. Headings that skip levels were 57.0% and 60.8%.
Links with no accessible name are WCAG failures, and 48.3% of small-business sites had at least one. A screen reader can only announce that link as “link” (WCAG 2.4.4). Buttons with no accessible name, also WCAG failures, were on 8.9% of sites and are announced as just “button”. Together with the structure issues, a screen reader user often can neither find nor identify where a link goes.
Structure checks show where navigation will break. Journey testing shows what a screen reader user actually experiences completing a task.
The two combined rows in the table were computed from the same scans and are not in the download files. The “any navigation issue” row includes region, heading-order, landmark-one-main, page-has-heading-one, empty-heading, and skip-link. Small business methodology.
| What a screen reader user runs into | Rule | Small businesses | National brands |
|---|---|---|---|
| Content outside any page region, so it cannot be reached by jumping between regions | region | 64.1% | 65.2% |
| Headings that skip levels, so the page outline makes no sense | heading-order | 60.8% | 57.0% |
| Page regions that share the same name | landmark-unique | 39.0% | 38.7% |
| No main content region, or no top-level (H1) heading | combined | 40.5% | 19.5% |
| No main content region to jump to | landmark-one-main | 27.9% | 10.5% |
| No top-level heading (H1) | page-has-heading-one | 22.4% | 11.3% |
| Empty headings, announced as just “heading” | empty-heading | 13.5% | 15.6% |
| Any of the navigation issues above: region, heading-order, landmark-one-main, page-has-heading-one, empty-heading, or skip-link | combined | 86.5% | 87.5% |
| Industry | Any navigation issue | No main region or no top-level heading |
|---|---|---|
| Dental and medical | 93.1% | 46.2% |
| Home services | 91.1% | 46.2% |
| Hotels and hospitality | 91.0% | 38.7% |
| Med spas | 91.0% | 39.3% |
| Beauty and wellness | 86.8% | 42.8% |
| Apparel and retail | 86.1% | 30.8% |
| Restaurants | 77.8% | 36.4% |
The brand cohort is a separate run, scanned with the same engine, the same page sample and the same seed as the small-business cohort. Brand and small-business figures sit side by side. They are never pooled into one number.
316 well-known US brand sites were attempted, from a hand-curated list. 256 were counted. 60 were excluded, mostly because bot protection blocked the scanner.
88.7% of brand sites had a detectable failure, against 90.1% of small businesses. Brands usually have fewer failures per page: a median of 5.2, against 6.2.
Two exceptions run the other way. National home-services franchise sites have more failures per page than local contractors (median 12.0 against 8.6). Apparel brands also have more (9.0 against 7.8).
The 34 brand sites with an overlay had a 97.1% failure rate and a median of 7.8 failures per page. The 222 without an overlay had 87.4% and a median of 5.0. This is the same pattern as Finding 2: overlays are not associated with fewer failures.
The most common failures on brand sites were low-contrast text (55.5% of sites), links with no accessible name (34.8%) and small touch targets (30.9%). Touch targets are more common on brand sites than on small-business sites (22.1%).
Methodology for the national-brand cohort
| Industry | National brands | Small businesses | |||
|---|---|---|---|---|---|
| Sites | % with a failure | Median per page | % with a failure | Median per page | |
| Home services | 49 | 93.9% | 12.0 | 94.5% | 8.6 |
| Apparel and retail | 43 | 93.0% | 9.0 | 93.8% | 7.8 |
| Dental and medical | 42 | 92.9% | 4.3 | 95.6% | 8.6 |
| Beauty and wellness | 40 | 92.5% | 5.6 | 92.5% | 6.4 |
| Hotels and hospitality | 39 | 82.1% | 3.0 | 90.9% | 5.0 |
| Restaurants | 43 | 76.7% | 1.0 | 80.8% | 3.5 |
| All industries | 256 | 88.7% | 5.2 | 90.1% | 6.2 |
Automated testing finds only some barriers. The three layers of testing cover what a scan does not. A page with no detected failures is not thereby accessible. These figures count automatically detectable WCAG failures. They are not a legal determination.
The small-business sample is local businesses found through public business listings in US metro areas, not a random sample of all US websites. Content inside iframes was not tested. Best-practice structure issues are reported separately and are not counted as failures. Results marked for manual review are not counted.
The measurement note for each cohort is the methodology file. Small business methodology · National brand methodology
Agencies can keep one record per client site.
Two cohorts, four files each: cells.csv, rules.csv, best_practice.csv and methodology.md. The files are aggregates. They name no business.
Data released under CC BY 4.0 (opens in a new tab); please credit Accessible and link to this report.
12,682 sites counted, scanned 9–10 November 2026.
256 sites counted. A separate run, never pooled with the small-business figures.
Accessible scanned 12,682 US small-business websites on November 9–10, 2026.
Accessible, Main Street Accessibility Report, November 2026, https://accessible.so/research/main-street-2026
Accessible scanned 12,682 US small-business websites in November 2026. 90.1% had at least one automatically detectable WCAG A/AA failure, and the median site had 6.2 per page. A page with no detected failures is not thereby accessible, so this is not a count of accessible websites.
In this scan, 95.5% of sites with an accessibility overlay had an automatically detectable WCAG failure, against 89.4% of sites without one. Medians were 8.8 failures per page with an overlay and 6.0 without. Overlays are not associated with fewer failures. The figures are a correlation. They do not show that an overlay causes failures.
Text with too little color contrast. It appeared on 71.5% of the 12,682 small-business sites, and it was the most common failure in every industry measured.
Automated checks of the page structure screen readers rely on found a best-practice structure issue on at least one page of 86.5% of 12,682 small-business sites. 40.5% had at least one page missing a main content region or a top-level heading, twice the national-brand rate of 19.5%. These are best-practice structure issues that affect screen reader navigation. They are not WCAG failures, and they are not counted in the failure figures.
The HHS Section 504 rule (45 CFR 84.84) requires WCAG 2.1 AA for web content of recipients of HHS federal financial assistance. A May 2026 extension set the dates to May 11, 2027 for recipients with 15 or more employees and May 10, 2028 for recipients with fewer than 15. It is not in force yet. This is not legal advice.
What overlays and widgets change at runtime, the limits accessibility organizations describe, the 2025 FTC order against accessiBe, and the evidence-first alternative.
How to do screen reader testing on a website with NVDA and VoiceOver: the commands that matter, an 8-step test, what to listen for, and common mistakes.
A step-by-step method for testing a site with the keyboard alone: focus order, visible focus, traps, skip links and custom controls.
The three layers of accessibility testing — what each one catches, what each one misses, and why a program needs all three.
Check one public page free, with no account. To see the barriers real visitors hit, Monitor is free too, with no card and no expiry.