How Accessible Are Local Business Websites? We Tested 2,868 Homepages
Most of them fail. In a November 2026 test of 2,868 US local business homepages with axe-core, 86.4% had at least one automatically detectable WCAG 2 A or AA failure. Low-contrast text (61.0%) and links with no readable name (41.8%) were the most common. Sites using an accessibility overlay failed at the same rate as sites without one.
- 86.4%of local business homepages fail at least one automated WCAG check. Only 13.6% pass them all.
- 61.0%have text with too little contrast to read comfortably.
- 41.8%have links with no readable name, usually social or phone icons.
- 10.0%stop visitors from zooming in on a phone.
- 7.4%load an overlay widget, and they fail at the same rate: 86.7% against 86.4% without one.
Thousands of US businesses are sued every year over websites that people with disabilities cannot use. UsableNet, which tracks these cases, counted 5,114 ADA digital accessibility lawsuits in 2025 across federal courts and New York and California state courts, and 1,416 of them named a company that already had an accessibility widget on its site (UsableNet 2025 year-end report). Most published data on website accessibility looks at big sites. The best-known study, the WebAIM Million, tests the home pages of the top one million websites (WebAIM). We wanted the picture for the restaurant, salon or contractor down the street.
We took the 2,868 independent local business homepages from our AI readability audit and ran the open-source axe-core engine, version 4.14.0, on each one in a full desktop browser on November 8, 2026. This report covers what it found. It is not legal advice, and an automated scan cannot tell you whether a particular site meets the law.
How many local business websites fail automated accessibility checks?
86.4%. Only 13.6% of homepages passed every automated check axe-core runs for WCAG 2.0, 2.1 and 2.2 at levels A and AA. The median homepage had 7 failing elements, and 25.2% had at least one problem axe rates as critical.
A failing element is one thing on the page that breaks a rule: a single grey caption, one icon link, one image. Most sites have a handful, and 15.9% have 25 or more. The count says nothing about how hard a page is to use, though. One unlabeled “Book now” button can block a screen reader user completely, while 30 faint footer captions are an annoyance.
These numbers are a floor. axe-core is built to avoid flagging things that are not real problems, and Deque, the company behind it, found in its own audit data that automated testing caught about 57% of accessibility issues by volume (Deque). Keyboard traps, confusing reading order and unclear instructions only show up when a person tests the page. A site that passes every automated check can still be hard to use.
Which accessibility problems are most common on local business websites?
Low-contrast text (61.0%) and links with no readable name (41.8%) lead by a wide margin. Next come images without alt text (17.3%), tap targets that are too small (15.3%), embedded frames with no title (13.4%) and pages that block zooming on phones (10.0%).
Low-contrast text
Light grey text on white, or white text laid over a busy photo. WCAG asks for a contrast ratio of at least 4.5 to 1 for normal text (W3C). People with low vision, older customers and anyone reading a phone outdoors are the first to lose it.
Links with no readable name
Often social media icons, a phone icon or a map pin with no text behind it. A screen reader announces “link” and nothing else, so the visitor cannot tell Instagram from the booking page (W3C). The fix is a few words of hidden label text on each icon.
Tap targets that are too small
WCAG 2.2 added a minimum size of 24 by 24 CSS pixels for things people tap, or enough space around smaller ones (W3C). Tiny footer links and tightly packed icon rows are common causes.
Zoom blocked on phones
A setting in the page code, user-scalable=no or maximum-scale=1, stops visitors from pinching to zoom. 10.0% of homepages had it. WCAG expects text to work at 200% zoom (W3C), and for someone who cannot read small print this one line makes the whole site unusable on a phone.
Frames with no title and images with no alt text
Embedded Google Maps, booking widgets and video players sit in frames. Without a title, a screen reader user hears “frame” and has to guess. Images need a short text alternative when they carry information, such as a photo of the menu or a sign with the opening hours.
How do local business websites compare with the web’s biggest sites?
On the six failure types the WebAIM Million tracks, local business homepages showed fewer automatically detected problems than the top one million home pages. The two studies use different testing engines, so treat this as a rough comparison.
| Failure type | Local business homepages (axe-core) | Top 1M home pages (WebAIM, WAVE) |
|---|---|---|
| Low-contrast text | 61.0% | 83.9% |
| Missing alternative text | 17.3% | 53.1% |
| Missing form input labels | 4.4% | 51.0% |
| Empty links | 41.8% | 46.3% |
| Empty buttons | 5.0% | 30.6% |
| Missing document language | 5.2% | 13.5% |
| Any detected failure | 86.4% | 95.9% |
WebAIM tests the rendered page with its own WAVE engine, which applies different rules from axe-core, and its sample is the busiest sites on the web: the average top home page carried 1,437 elements in its 2026 report. Because the engines and the samples differ, the lower local figures do not mean local businesses are doing well. They do suggest the fixes are within reach, since a small site has fewer pages and templates to correct.
Do accessibility overlays fix local business websites?
Not by this measure. 7.4% of sites loaded an accessibility overlay widget, most often UserWay on 92, AudioEye on 61, accessiBe on 41 and One Click Accessibility, a WordPress plugin, on 16 sites. Sites with an overlay failed automated checks at the same rate as sites without one: 86.7% against 86.4%.
An overlay is a script that adds a toolbar or tries to repair a page in the browser. We ran our checks after the page and its overlay had finished loading, so any repairs the overlay made were counted. The failure rate barely moved. Low-contrast text was less common on overlay sites (54.5% against 61.5%). Unnamed links were about as common (43.6% with an overlay, 41.7% without). Images without alt text were more common on overlay sites (22.7% against 16.8%).
Overlays were far more common on some builders: BentoBox (45.0% of 60 sites) and Duda (27.4% of 117 sites). When an overlay arrives with a platform, a theme or a plugin, the owner may not know it is there.
Regulators have taken note of overlay marketing. In April 2025 the Federal Trade Commission finalized an order requiring accessiBe to pay $1 million and barring it from claiming its automated product can make any website WCAG-compliant without evidence to support the claim (FTC). UsableNet’s count of 1,416 lawsuits in 2025 against companies already using a widget points the same way.
Which website builders produce the most accessible local business sites?
Among platforms with at least 50 sites in our sample, BentoBox sites failed least often (65.0% failed at least one check) and GoDaddy sites most often (96.8%). Much of this comes from templates. Colors, footer icons and the zoom setting usually arrive with the theme and stay on the live site unless someone changes them.
| Platform | Sites | Failed a check | Median failing elements | Low contrast | Unnamed links | Zoom blocked | Overlay loaded |
|---|---|---|---|---|---|---|---|
| WordPress | 1,069 | 95.1% | 11 | 72.7% | 54.1% | 17.9% | 6.9% |
| Other or custom | 695 | 83.6% | 6 | 56.7% | 26.6% | 8.8% | 6.9% |
| Squarespace | 397 | 70.8% | 3 | 52.1% | 34.0% | 0.5% | 3.5% |
| Wix | 268 | 78.7% | 3 | 57.5% | 35.1% | 0.4% | 2.2% |
| GoDaddy | 124 | 96.8% | 9 | 54.0% | 62.1% | 7.3% | 4.0% |
| Duda | 117 | 84.6% | 4 | 52.1% | 47.0% | 0.9% | 27.4% |
| Webflow | 67 | 94.0% | 8 | 70.1% | 67.2% | 0.0% | 7.5% |
| BentoBox | 60 | 65.0% | 1 | 6.7% | 16.7% | 0.0% | 45.0% |
| Square Online / Weebly | 33 | 93.9% | 5 | 54.5% | 21.2% | 54.5% | 0.0% |
| Shopify | 24 | 95.8% | 10 | 66.7% | 33.3% | 12.5% | 0.0% |
We identified platforms from the files each page loads. “Other or custom” covers hand-built sites and builders we did not fingerprint. A builder’s numbers reflect its templates and its customers’ choices together, so a well-built site on any platform can pass, and a poorly chosen theme can fail on all of them.
Which local industries have the most accessibility problems?
House cleaning services had the highest failure rate (93.5%) and restaurants the lowest (75.5%). Low-contrast text showed the widest gap between them.
Low-contrast text appeared on 79.6% of house cleaning homepages and 46.0% of restaurant homepages.
What should a local business fix first?
Name your icon links, raise text contrast, and allow zoom. These three cover most of what automated checks flagged on local business homepages, and none needs a redesign.
- Give every icon link a name. Social icons, phone and map icons need a text label a screen reader can read. Most builders have an “alt text” or “accessible name” field for icons.
- Darken light text. Grey on white and white over photos are the usual culprits. A free contrast checker will tell you if a pair of colors passes.
- Let people zoom. Ask whoever built the site to remove
user-scalable=noandmaximum-scale=1from the viewport tag. - Describe meaningful images. A sentence of alt text for photos that carry information. Decorative images can stay empty.
- Title embedded maps and booking widgets. A frame title such as “Map to our shop” is enough.
- Test it yourself. Press Tab to move through the page without a mouse, then try your phone’s built-in screen reader (VoiceOver or TalkBack) on the homepage and booking page. Ten minutes of this finds problems no scanner can.
Accessibility work overlaps with the basics in our AI readability audit: clear link text, real text instead of text inside images, and labelled content help screen readers and AI crawlers alike. If you are planning a new site, what pages a website needs is a good place to start.
How was this audit done?
We tested 2,868 homepages from our earlier sample of independent local businesses in 20 US cities with axe-core 4.14.0 in headless Chromium on November 8, 2026, running every rule tagged for WCAG 2.0, 2.1 and 2.2 at levels A and AA.
Sample
The sites are the independent restaurants, hair salons, general contractors, auto repair shops and house cleaning services that loaded normally in our first audit, drawn from Google Maps results in 20 US cities. Chains, franchises and businesses without their own website were excluded there. 14 homepages could not be tested this time and are left out.
Method
- Engine. axe-core 4.14.0, rules tagged wcag2a, wcag2aa, wcag21a, wcag21aa and wcag22aa. We recorded violations only, not items axe marks for manual review.
- Browser. Headless Chromium at 1366 by 900 pixels. Each page was given time to finish loading, including third-party scripts, before the scan.
- Overlays. Detected from the scripts and network requests each page made, matched against the published script hosts of the main overlay vendors.
- Failure groups. To compare with WebAIM we grouped axe rules: color-contrast as low contrast; image-alt and related rules as missing alt text; label and select-name as missing labels; link-name as empty links; button-name and input-button-name as empty buttons; html-has-lang and html-lang-valid as missing language.
Limits
- Homepages only, desktop size only, one point in time.
- Automated checks find only part of the problems a person with a disability would hit. Treat every figure here as a minimum.
- The comparison with WebAIM mixes two engines and two very different samples.
- Platform detection relies on file signatures, so some builders fall into “Other or custom.”
Can I use this data?
Yes. The aggregated results are free to reuse under the Creative Commons Attribution 4.0 licence. Credit AltHealth Works and link to this page.
Download the aggregated data (CSV), with sample sizes and 95% confidence intervals. We do not publish results for individual businesses.
AltHealth Works (2026). How Accessible Are Local Business Websites? An automated WCAG audit of 2,868 US local business homepages. https://althealthworks.com/research/local-business-website-accessibility-2026
Sources
- The WebAIM Million: 2026 report on the accessibility of the top 1,000,000 home pages · WebAIM
- 2025 Year-End Digital Accessibility Lawsuit Report · UsableNet
- FTC Approves Final Order Requiring accessiBe to pay $1 Million · Federal Trade Commission, April 22, 2025
- Automated testing study identifies 57 percent of digital accessibility issues · Deque Systems
- Web Content Accessibility Guidelines (WCAG) 2.2 · W3C
- Understanding Success Criterion 1.4.3: Contrast (Minimum) · W3C
- Understanding Success Criterion 2.4.4: Link Purpose (In Context) · W3C
- Understanding Success Criterion 2.5.8: Target Size (Minimum) · W3C
- Understanding Success Criterion 1.4.4: Resize Text · W3C
- Guidance on Web Accessibility and the ADA · ADA.gov
- axe-core accessibility engine · Deque Systems
More research and guides
- Can AI Read Local Business Websites?Our first audit: phone, address and hours, AI crawler access, speed and platforms.
- What Pages Does a Website Need?Five pages, what to write on each one, and a checklist before you publish.
- How to Get a Website That Looks the Way You WantTemplates, custom design, and checking the result.