No. An overlay mainly changes how the page looks. Problems in the code, such as missing labels, a broken focus order or a heading structure that is not marked up, still have to be fixed in the site itself.
The short answer: An accessibility overlay is a script that adds a toolbar on top of your pages. It can change colors, enlarge text or stop animations. Some overlays also try to patch the page code with scripts, but they cannot reliably make a whole site accessible. Screen readers and keyboards depend on the real structure of the site. That is why accessibility overlay problems show up in audits even on sites that already have a widget installed. Real accessibility is built into headings, buttons, forms, labels and focus order.
Every few weeks a new client sends us the same message. “We installed an accessibility widget, so we should be fine, right?” Then we open the site with a screen reader.
Over 9+ years of manual audits, the pattern is consistent. The widget adds a floating button with font sizes, contrast modes and a readable font switch. Some widgets also try to patch things automatically, but the real problems in the templates are still there. And that structure is what assistive technology actually depends on.
An accessibility overlay, sometimes called an accessibility widget or an accessibility plugin, is a third-party script. You add a line of code that loads from an external server, or you install it from a CMS plugin directory. Installation is simple. That does not mean the site is accessible.
The overlay offers visual adjustments: color changes, pausing animations, highlighting headings and links, or enlarging text to 200%. Some of these adjustments can help some users. The trouble starts when the widget is treated as the accessibility solution itself.
People who rely on screen readers, switch devices or voice control already have their own tools. They configured them carefully. They don’t need a second toolbar on your page. They need your buttons to be buttons, your forms to have labels and your menus to open from the keyboard.
These come straight from the issues we keep finding when we test real sites, including sites that already have a widget running.
An overlay does not really know why an image is on the page. Good alt text is written for the context. A product photo needs a different description from a chart, and a purely decorative image should have an empty alt so screen readers skip it. We often see auto-generated phrases like “image of a person” or “graph”. For someone using a screen reader, that is not information.
Keyboard navigation needs a logical order, a visible focus, and a way to reach every control. If focus jumps around or disappears, the user gets stuck. A display toolbar does not reliably fix that order.
We often see date pickers, menus and pop-up windows that open from the keyboard but never let focus out. Tab, Shift+Tab and Escape do nothing. A user without a mouse is trapped. An overlay script cannot reliably understand and repair that kind of custom behavior. A modal dialog should keep focus inside it while it is open and let the user close it from the keyboard. That has to be built into the component.
A clear heading hierarchy helps a screen reader understand the page. A blind user jumps between headings the way a sighted user scans with their eyes. When headings are just bold text, the page becomes one long block. Skipped heading levels are not always a failure on their own, but a confusing hierarchy makes the page much harder to scan. Changing colors or font size does not create that structure.
Fields, labels, instructions and error messages have to be clear in the code. A contact form or a signup form that does not work with a keyboard stops the user at the exact moment that matters most to the business. Error messages that appear only in red, required fields marked with an asterisk the screen reader skips, and focus that jumps to the top after submit, all sit in the form logic. An overlay sits outside it.
Complex components such as menus, sliders and dialogs need their own manual check and their own fix in the source. We have also tested overlays whose toolbar was hard to reach by keyboard, or that changed the reading order. On mobile, the extra layer can hide content, slow the page down or take over keyboard navigation. A comprehensive check I ran with users with disabilities showed that they sometimes avoid these widgets: the menu is unclear, it is hard to use, or it forces them to use the overlay even when they do not want to.
WCAG is the W3C’s Web Content Accessibility Guidelines. In our audits, we have not seen an overlay that meets those guidelines on its own.
An overlay can change how the page looks, and some try to patch the code automatically. It cannot reliably detect and fix missing alt text, a broken heading structure, a screen reader problem or a custom component that is not a real button. Automated tools are limited here. Those issues are found by hand, with a keyboard and with assistive technology.
There is also no requirement to install an overlay in order to meet WCAG. You can, and I recommend that you do, make the site accessible in the code without adding an external widget.
It can. Most overlays load extra JavaScript and CSS on every page. The browser has to download and run them before the page is fully ready.
We tested this live with PageSpeed on a test WordPress site, same URL. Without an accessibility overlay, the performance score was 90. After we installed a free overlay from the WordPress plugin directory, the score dropped to 55. Results differ from site to site, but a slower site can lose visitors who will not wait for the page to load. So you pay twice: the site is still not accessible, and it also became slower.
The good news is that the fixes are usually smaller than people expect. Most of what we correct comes down to a handful of patterns repeated across templates. Fix the template once and many pages improve together. We walk through the basics in our guide on how to make your website accessible.
If you are building a new site, bring in an accessibility expert while the first templates are still being designed. It saves development hours later and helps the site stay at the same level as it grows.
Automated scanners are useful for catching contrast issues and missing attributes. They cannot tell you whether a checkout flow makes sense with a screen reader. In our audits, we test the main templates with a keyboard, with NVDA or JAWS on desktop, and with VoiceOver and TalkBack on mobile. We try to fill in forms, open menus and reach every button, the way a user with a disability would.
This is what actually changes the experience. Headings get a real hierarchy. Interactive elements become real, native controls where possible. Images get alt text that matches their purpose. ARIA is used only where HTML cannot do the job. We work directly with your developers, or our experts make the changes.
After the fixes, we test again. Same devices, same flows. Then you publish an accessibility statement that reflects what was actually done on your site, not a generic template copied from a widget vendor. Short training for content editors, and a periodic review, keep the site at WCAG 2.2 AA as new pages and campaigns go live.
Yes. If you want a contrast toggle or a larger text option after the site is accessible in the code, that can stay as a convenience feature. Treat it as an extra layer on top of an accessible codebase, not as a replacement for one.
Pick the three journeys that matter most on your site: maybe contact, signup and checkout. Try each one using only the Tab and Enter keys. If you get stuck, an assistive technology user will too.
Want a clear picture of where your site stands? Our experts handle this end to end. You can see everything we do on our accessibility services page.
No. An overlay mainly changes how the page looks. Problems in the code, such as missing labels, a broken focus order or a heading structure that is not marked up, still have to be fixed in the site itself.
No. There is no requirement to install one. The reliable way to meet WCAG is to fix the code, test with a keyboard and with assistive technology, and publish a statement that matches the site.
Yes. Once the code works, a simple contrast or text-size control can stay as an extra option. It should not be the thing you rely on for accessibility.
Keep using the site, and start a manual audit of your main templates and user journeys. Fix what the audit finds in the source code, re-test, and update the accessibility statement so it describes the real site. Then decide whether the widget is still useful as a convenience feature.
This website uses essential cookies to operate and, with your consent, statistics cookies (Google Analytics). For more details, see our Privacy Policy.