ADA Monitor

Buyer's guide

Overlay widgets vs. source remediation.

Two very different things get sold as "website accessibility." This page explains what each actually does, where each genuinely helps, and what neither can do — including the limits of our own approach.

last reviewed: august 2026 · sources linked inline

The short answer

An overlay is JavaScript that adjusts a page inside the visitor's browser. Your site's code is untouched; the adjustment is re-applied on every page load, and only while the script runs.

Source remediation changes the underlying code — for a Shopify store, the theme — so pages are built correctly before they reach anyone's browser.

The honest summary: overlays are quick to install and can offer real visitor conveniences, but they cannot fix everything and they do not change what is actually wrong with your store. Source remediation is slower and requires approval, but the fix persists, works without JavaScript, and shows up for search engines and assistive technology alike. Neither one makes a site fully accessible on its own, because no automated tool can.

What an overlay actually does

You add one script tag. On each page load it inspects the rendered page and applies changes — adding ARIA attributes, guessing at labels, generating alt text, adjusting contrast — and usually offers visitors a toolbar of display preferences.

Where overlays genuinely help

  • Visitor display controls. Letting someone raise contrast, enlarge text, or increase spacing on demand is a real convenience, and some visitors like having it.
  • Speed. Installation takes minutes and needs no developer.
  • Some genuine coverage. Certain mechanical problems — a missing landmark, an unlabeled control — can be patched at runtime.

Where they fall short

  • The underlying code is still wrong. Change themes, and every issue returns, because nothing was ever fixed.
  • They depend on JavaScript running. If the script fails, is blocked, or loads slowly, the page reverts to its unmodified state.
  • Guessed labels can mislead. An automatically generated label that describes the wrong thing can be worse for a screen-reader user than no label, because it sounds authoritative.
  • Many people who rely on assistive technology dislike them. That feedback is worth taking seriously; it comes from the users the tool claims to serve.

What the regulators and standards bodies have said

Three public sources are worth knowing before you buy anything in this market.

The FTC fined an overlay vendor $1 million

In April 2025 the FTC finalized an order requiring accessiBe to pay $1 million. The agency alleged that claims an automated AI product could make any website WCAG-compliant were false, misleading, or unsubstantiated, and that connected promotional content appeared independent without adequate disclosure.

Read that precisely. The action was about the claims made for the product, not a ruling that overlays are illegal or worthless. The lesson applies to every vendor in this category, us included: the promise is what gets you in trouble.

An overlay is not a legal shield

Accessibility.com's report on 2025 website accessibility lawsuits recorded 1,432 cases and found that 37% of the sites sued had an overlay installed.

Be careful how you read that number. It does not show overlays cause lawsuits. Businesses that already suspect they are exposed are more likely to install one, so the two are entangled. What it does show is that having an overlay did not prevent those businesses from being sued.

No automated tool can do the whole job

The W3C's guidance on selecting accessibility evaluation tools states plainly that tools cannot automatically check all accessibility criteria, that human judgment is required, and that no tool alone can determine whether a site meets a standard.

This applies to us as much as to anyone else. It is why our product routes some findings to a person instead of pretending to fix them.

Side by side

Comparison of overlay widgets and source remediation across nine practical dimensions
  Overlay widget Source remediation
What changesThe page in the browserYour theme code
Survives a theme changeIssues returnFix stays in the theme
Works without JavaScriptNoYes
Visible to search enginesNot reliablyYes
Setup effortMinutesReview and approval per change
Visitor display controlsUsually includedNot applicable
Evidence of what changedVaries by vendorA diff you can read
Handles design judgmentNoNo — needs a person either way
Fixes third-party app codeSometimes, at runtimeNo — belongs to the vendor

Comparison of the two approaches as categories. Individual products differ; check what any specific vendor actually does rather than assuming its category's behaviour.

What neither approach can do

Being straight about this is more useful than another feature list.

  • Judge whether alt text is meaningful. A tool can tell that alt text is missing. Only a person can tell whether "product image 4" describes anything useful.
  • Tell you whether a checkout flow makes sense to someone using a screen reader end to end.
  • Make design decisions. If your brand colour fails contrast, something has to change, and that is a business decision.
  • Fix a third-party app's code. A support chat widget or reviews app belongs to that vendor. It can be flagged and escalated, not rewritten safely.
  • Guarantee compliance. Conformance is not the same as legal compliance, and no automated product can promise either.

How to choose

Rather than picking a category, ask any vendor these five questions. The answers separate the market faster than any feature grid.

  1. What exactly did you change, and can I see it? "AI fixed 40 issues" is not an answer. A diff is.
  2. What could you not fix, and who owns it now? A vendor that reports only successes is hiding the rest of your list.
  3. What happens when I change my theme? This is where runtime patches and real fixes diverge sharply.
  4. Do you guarantee compliance? If yes, walk away — that is the exact claim the FTC acted on.
  5. Can I approve changes before they go live? Anything touching your storefront should be your call.

Where ADA Monitor sits

We do source remediation for Shopify themes, and we are bound by the same limits described above.

What we do: monitor your storefront as it changes, prepare eligible fixes in an unpublished copy of your theme, show you the exact diff, and publish only under approval rules you configure. What we can't fix automatically gets an owner and a next step — a design decision routed to your team, a third-party issue routed to that vendor — instead of silence.

What we will never tell you is that this makes your store compliant. It gives you a maintained, documented programme: issues found, fixes verified, decisions recorded. That is the substance regulators look at when they ask what a business is actually doing.

We publish the rules we hold our own marketing to, including the claims we refuse to make.

Questions people actually ask

Are accessibility overlays bad?

Not inherently. An overlay is a piece of JavaScript that adjusts a page in the visitor's browser. Some of what overlays offer is genuinely useful — a contrast or text-size control that a visitor chooses to turn on is a real convenience. The problem is what gets claimed for them. In April 2025 the FTC finalized an order requiring overlay vendor accessiBe to pay $1 million, alleging its claims that an automated AI product could make any website WCAG-compliant were false, misleading, or unsubstantiated. The tool is not the issue; promising conformance it cannot deliver is.

What is the difference between an overlay and source remediation?

An overlay leaves your site's code unchanged and patches the page in the visitor's browser each time it loads. Source remediation changes the underlying code — for a Shopify store, the theme itself — so the page is built correctly before it ever reaches a browser. The practical difference is durability and reach: a source fix persists, works without JavaScript, and is visible to search engines, while an overlay's effect exists only while its script is running.

Will an overlay protect me from an accessibility lawsuit?

No tool can promise that, and you should be wary of any vendor that does. Accessibility.com's report on 2025 website accessibility lawsuits recorded 1,432 cases and found that 37% of the sites sued had an overlay installed. That figure does not show overlays cause lawsuits — businesses that already believe they are at risk are more likely to install one — but it does show an overlay is not a shield.

Can automated tools make a website fully accessible?

No, and this is true of every automated approach including ours. The W3C states that evaluation tools cannot automatically check all accessibility criteria, that human judgment is required, and that no tool alone can determine whether a site meets an accessibility standard. Automated testing finds many real barriers quickly and cheaply. It cannot tell you whether alt text is meaningful, whether a flow makes sense with a screen reader, or whether a design decision is right.

Which approach should a Shopify store choose?

If you want the underlying problems actually fixed and a record of what changed, choose source remediation. If you want visitor-facing display preferences such as contrast or text sizing, an overlay offers those, and some merchants run both. What matters most is not the category but whether the vendor tells you plainly what it did, what it could not do, and who approved each change.

Sources

Figures are reported with the tracker, scope, and year named, and are not combined across sources. Statements about any specific product reflect that vendor's public claims or the public record cited above.

See your own store

Stop guessing which category you need.

A free scan reads your public storefront and tells you what's automatically preparable, what needs a decision, and what belongs to a third party — in plain language.