Comparison
Accessibility overlays, compared with the alternatives
An overlay widget does not make a website accessible, and installing one has not stopped businesses being sued. It can still do two or three useful things, and this page says what they are before it says anything else.
We sell scheduled scanning with a dated record, so we are not a neutral party. The way to read a comparison written by a vendor is to check whether it admits where it loses. This one does, in its own section below, and every number on the page names its source.
What is documented
What an overlay is genuinely good at
Starting here because it is the part most vendor comparison pages skip, and skipping it is how you end up with a page nobody trusts.
- It installs in minutes. One script tag, no developer, no deployment, no access to the codebase. For an owner with no technical help at all, nothing else on this page is that easy.
- Some of the controls are real. Larger text, more line spacing, a high-contrast palette, stopping animation, a bigger cursor. Those settings help some people, and a visitor who does not know their browser or operating system already offers them may find the toolbar useful.
- It comes off as fast as it went on. Removing the script tag reverses the change completely, which is not true of most decisions you make about a website.
That is the honest case. It is a presentation layer, and as a presentation layer it works. The problem is what gets claimed on top of it.
What an overlay cannot reach
An overlay runs in the visitor's browser, after your page has loaded, and it can only adjust what it can see. The failures that generate complaints and demand letters live in the markup your server sent, and changing them means changing the page.
- An image with no alt text, or with alt text that describes nothing useful. A script can guess at the picture; it cannot know why you put it there.
- Heading levels used for size rather than structure, so the document outline a screen reader announces is meaningless.
- Form fields with no programmatically associated label, so the field is announced as "edit text" and nothing else.
- Keyboard traps, focus that vanishes, and modal dialogs that do not return focus.
- Reading order that makes sense visually and no sense in sequence.
- A custom component whose ARIA role promises behaviour the JavaScript does not deliver.
This is also why courts have declined to treat an installed overlay as evidence that a site is accessible, and why settlements have required the widget to be taken back off. The widget is not the site.
Side by side
Five ways to approach the same problem
| Approach | Changes the page? | Finds source problems? | Produces dated evidence? | Cost |
|---|---|---|---|---|
| Overlay widget | No, it adjusts presentation in the browser | No | No | A subscription set by the vendor |
| Free scanner, run by hand | No, it reports | Yes, the 30 to 40% a machine can see | Only if you save the output yourself | $0 |
| Scheduled scanning with a record (us) | No, it reports | Yes, the same 30 to 40% | Yes, that is the entire product | $49 to $249 a month, or $149 once |
| Manual audit by a specialist | No, it reports | Yes, including everything a machine cannot judge | Yes, a report dated on the day | Quoted per site, the most expensive row here |
| Remediation by a developer or agency | Yes, this is the only row that fixes anything | Depends who you hire | Your commit history, if you keep it | Quoted per project |
Read the table as a sequence rather than a menu. Almost every site needs the last row eventually, and the rows above it exist to tell you what to put in it.
Where the other options beat us
A comparison page that only flatters its author is an advert. Here is the part that costs us sales.
Established vendors in this category
The larger companies selling accessibility products have things we do not: staffed support during business hours, manual audit teams who test with real assistive technology, remediation services that actually change your code, track records measured in years, and enough customers to give you references. If you need somebody to take the work off your hands, we are not that, and no amount of scanning will make us that.
The free tools
WAVE, Lighthouse and axe DevTools are free, instant, need no account, and run in your own browser so nothing leaves your machine. On a single page, today, they will tell you essentially what our scanner tells you. If a one-off check is all you need, use them and keep your money. We compare ourselves with them in more detail on the audit tools page.
Manual auditors
A specialist working through your site by hand finds the 60 to 70 percent that no software reaches, can tell you whether your alt text is any good, and can produce the conformance documentation that procurement departments ask for. If you have the budget for exactly one thing, buy that, not us.
Us
We are a small operation. We do not fix your site, we do not perform manual audits, we do not test with screen readers on your behalf, and we cannot tell you whether you are compliant with anything. Our scanner uses axe-core, which is open source and which several free tools also use, so we are not claiming to detect more than they do. What we add is that it runs on a schedule, keeps every finding and every fix with the date it happened, and exports that history as a dated PDF. That is one narrow thing. It is genuinely hard to reconstruct after the fact, which is why it is worth paying for, and it is worth nothing at all if you do not need it.
What we would actually tell a friend to do
- Run a free scanner on your busiest three pages this afternoon. It costs nothing and takes ten minutes.
- Fix what it finds. A large share of automated findings are contrast, missing labels and missing alt text, and most of that is an afternoon of work.
- If there is any budget, spend it on a manual audit by a specialist before you spend it on any subscription, ours included.
- Start keeping a dated record from the day you begin, whatever you use to keep it. The date is the part you cannot manufacture later.
- Do not install an overlay expecting it to change your legal position. On the documented evidence, it does not.
If you already have an overlay installed
There is no need to panic and no need to rip it out this morning. Two practical things are worth doing. Test your site with the widget switched off, because that is the site everyone else is arguing about. And stop counting it as work done, because the underlying page has not changed and the record of what you have fixed is currently empty.
See what is on your own page
One page, no account, no card. The report names each failure, where it is, and which order to work through them. It also says out loud which parts of WCAG it could not check.
Scan any page, free
Questions
Things people ask about overlays
Will an overlay stop me being sued?
There is no honest way to promise that about any product, ours included. What is on the record is that more than 800 businesses were sued in 2023 and 2024 while running an overlay, and that courts have declined to treat an installed widget as evidence that a site is accessible.
Whether anything helps in your particular situation is a question for a lawyer, and we are not one.
Why was accessiBe fined?
In April 2025 the Federal Trade Commission imposed a $1,000,000 penalty over claims that the company's software could make any website compliant with accessibility standards. That is the documented action, and it is the only company we name on this page, because it is the only one with a published federal enforcement outcome behind the claim.
We are not going to speculate about any other vendor. If a vendor tells you their product delivers compliance, the FTC's position on that claim is now a matter of public record and you can weigh it yourself.
Are overlays actively harmful to disabled users?
Some users and some accessibility practitioners report that overlays interfere with their own assistive technology, which already provides similar settings. Others use the controls and find them helpful. We have not run a study, so we are not going to give you a percentage.
The claim we are confident making is narrower and better documented: an overlay does not change the markup that causes the failures, so the failures are still there when the widget is switched off.
You sell a scanner. Is this page just a sales pitch?
Partly, obviously. The test of whether it is only that is whether it tells you when not to buy from us, and the sections above tell you to spend your first budget on a manual audit and your first afternoon on a free tool. Both of those are honest advice that costs us money.
The limits of this
Automated testing detects roughly 30-40% of WCAG success criteria. It cannot determine whether a site is legally compliant, and it is not a substitute for manual review and assistive-technology testing. This report is a record of what automated checks found on the date shown - nothing more, and nothing less.
Nothing on this page is legal advice and we are not a law firm. No software, ours or anyone else's, can make a site compliant or protect it from a claim. Anyone telling you otherwise is describing the thing the FTC took action over.