The accessibility widget in the corner of the screen, and what it does not do

What an overlay changes on the page and what it leaves untouched

A vendor emailed this morning offering a single line of code that makes the site compliant for a small monthly fee. It sounded like the easiest fix you'd found all year.

An overlay widget sits in a corner of the page and offers a menu of adjustments: bigger text, higher contrast, a dyslexia-friendly font, sometimes a screen-reader mode of its own. All of that happens visually, in the browser, on top of the page that's already there.

It does not touch the actual HTML underneath. The missing labels, the broken heading order, the images with no description are all still exactly as broken as they were before the widget loaded.

The overlay changes what the page looks like. It does not change how the page is built. That distinction is the entire question this post is trying to answer.

Ask what the widget's own installation instructions require. If the answer is a single script tag and nothing else, that's a strong hint about how deep the fix actually goes.

Why screen reader users are the loudest critics of these tools

The people likeliest to actually rely on assistive technology are also the loudest critics of overlay widgets. That's worth sitting with for a moment.

Screen reader users report overlays that override their own device settings. They report broken keyboard navigation that was working fine before the widget arrived. New obstacles, stacked on top of old ones.

A tool marketed as a fix for accessibility, built without much input from people who use assistive technology daily, tends to solve the problem its marketing team imagined. Not the one that actually exists on a real page, for a real visitor.

Some visitors have taken to disabling these widgets the moment a page loads, going back to whatever their own device already had set up. That reaction alone says something about how well the product works.

The recurring fee against a fixed piece of build work

An overlay is sold as a subscription. A monthly fee, for as long as the widget stays on the site, with no natural end point.

The alternative is different in shape. Fixing the actual HTML — the labels, the heading structure, the color contrast — is a fixed piece of work with a beginning and an end.

One keeps billing indefinitely. The other, done properly once, stays done until the next redesign touches it. Over several years, that difference in shape adds up to a meaningfully different total.

A fixed piece of work also has something a subscription doesn't: a person you can name, who did the work, and who you can call if something breaks later.

Cancel the widget subscription and the site goes back to exactly how it was before. Finish the actual build work and the improvement stays, subscription or not.

What the sales page means by compliant and what it cannot mean

No accessibility standard has a single pass-or-fail switch that a script can flip. The guidelines are a set of detailed criteria covering structure, contrast, keyboard access, and dozens of other specifics.

A widget layered on top of broken markup cannot satisfy criteria that depend on the markup itself. That's not an opinion about the product. It's how the underlying standard is written.

When a sales page says a widget makes a site compliant, it's describing a surface-level adjustment. Not a rebuild of what's underneath the surface. Read the actual claim carefully, not the headline above it. A sales page written by a marketing team is not the same document as the standard it's claiming to satisfy.

Sites carrying an overlay have still been sued, sourced to UsableNet

UsableNet, which tracks federal website-accessibility litigation, has recorded lawsuits filed against sites that were running an overlay widget at the time of the filing. Carrying the tool did not prevent it.

That single fact undercuts the central promise in overlay marketing — that the widget itself acts as a legal shield. It may be many things to many buyers.

A reliable defense against a claim is not one of them, according to what UsableNet has actually recorded across the cases it tracks. Worth reading directly, not taking on faith from a vendor's own pitch.

The narrow cases where a widget is a defensible stopgap

There's a narrow, honest case for one of these tools. Picture a site that's mid-rebuild, with real remediation already scheduled and contracted, using a widget as a temporary bridge while the actual work is underway.

Even then, it's a stopgap sitting on top of a real plan. It is not a replacement for one.

Buying the tool instead of fixing the underlying build is the version that leaves the site exposed, and the visitor still locked out of using it properly.

Even in that narrow case, the widget is a placeholder with an expiration date, not a permanent answer bought once and forgotten about.

Put a date on the contract for the real remediation work. A stopgap with no end date has a way of quietly becoming the permanent plan by accident.

This sits inside the wider build discipline that accessibility actually is — what accessibility means before any of this gets legal is worth reading first. The litigation numbers themselves, and what they do and don't tell you about your own risk, are covered properly at the current filing data, sourced to the same tracker.

Finding out what your site actually does badly, widget or not, starts with testing it yourself. And if the list of problems turns out longer than a DIY pass can fix, a paid audit is the other option worth pricing out.

More from the blog

All posts