Accessibility monitoring tools compared: overlays, manual audits, and automated scanning
When businesses start looking at web accessibility, they typically encounter three categories of products: overlay scripts, manual expert audits, and automated scanning tools. Each addresses a different part of the problem, at a different price point, with different trade-offs. This page explains what each approach does and where each one fits.
| Feature | Overlay scripts | Manual expert audits | Automated scanning |
|---|---|---|---|
| What it covers | Runtime DOM patches on detected issues | Full WCAG evaluation including screen reader and keyboard testing | Approximately 30-57% of WCAG criteria, reliably and programmatically |
| How it works | JavaScript added to existing pages; modifies the accessibility tree after page load | A certified specialist tests the live site manually | Automated rules engine tests pages on a schedule |
| Ongoing monitoring | Runs on each page load, but does not track changes or regressions | No -- point in time; a new audit is needed after major changes | Yes -- rescans on a schedule, reports what is fixed, new, or regressed |
| Fix guidance | None -- patches the surface, leaves the underlying code broken | Detailed remediation report | Per-issue: WCAG criterion, affected elements, HTML-level fix |
| Typical cost | $49-300/month for a subscription widget | $5,000-50,000+ depending on site size and scope | $250-750/month for ongoing monitoring (BarrierScan) |
| FTC action | accessiBe settled for $1 million over deceptive compliance claims (2024) | Not applicable | Not applicable |
| When to use it | We do not recommend it (see below) | Legal filings, VPATs, Section 508, thorough pre-launch review | Ongoing code-level monitoring as the site ships changes |
Overlay scripts
Overlay scripts, products like accessiBe, UserWay, and the overlay tier of AudioEye, inject a JavaScript snippet into existing pages and attempt to modify the accessibility tree after the page renders. The premise is that a single script installation replaces the need for code-level fixes.
The technical problem is foundational. Screen readers build their model of a page from the DOM and accessibility tree at load time. A JavaScript patch that fires afterward can conflict with what a screen reader has already parsed, creating behavior that is more disorienting than the original issue. This is why many screen reader users disable overlay scripts entirely: the interference with their tools is often worse than the underlying barriers the overlay was trying to fix.
Overlays are also limited to what automated analysis can detect. The issues they miss are the same issues any automated tool misses, roughly 40-70% of WCAG criteria, including screen-reader-specific navigation problems, custom interactive components, and anything that requires human judgment to evaluate. Unlike manual audits or scanning tools, overlays do not tell you what those issues are: they cannot surface problems they cannot detect, and they patch what they can detect at runtime rather than giving you a list of what to fix in code.
In 2024, the FTC took enforcement action against accessWidget, Inc. (maker of accessiBe), resulting in a $1 million settlement centered on deceptive claims that the product would bring websites into compliance with the Americans with Disabilities Act. Courts have also declined to treat overlay installation as a defense in ADA web accessibility lawsuits. If a plaintiff demonstrates a barrier on your site, the presence of an overlay script has not been found to change the outcome.
BarrierScan does not use or sell overlay scripts, and we do not recommend them as an accessibility strategy.
Manual expert audits
A manual audit involves a trained specialist, typically certified by the International Association of Accessibility Professionals (IAAP) or equivalent, who tests a site using a screen reader, keyboard, and systematic WCAG review. They identify both the issues that automated tools detect and the issues that require human judgment, and they produce a detailed remediation report.
This is the most thorough option and the only one that covers the full range of WCAG criteria. For organizations pursuing VPAT documentation, Section 508 compliance for federal contracts, or preparing for legal review, a manual audit is the appropriate tool. The documentation it produces is suitable for legal proceedings in a way that automated scan reports are not.
The practical limitations are cost and time. Manual audits typically run from $5,000 to $50,000 or more depending on site complexity. More importantly, an audit reflects the site at a point in time. E-commerce sites change constantly: new product pages, seasonal campaigns, third-party widget integrations, developer releases. A single audit conducted in January does not cover what ships in March.
For sites that need formal documentation or are facing litigation, a manual audit is the right investment. For monitoring a site as it evolves month over month, a manual audit alone does not provide that coverage.
Automated scanning tools
Automated scanners test pages against a defined rule set and flag violations that can be detected programmatically. What automated scanning reliably catches: missing image alt text, unlabeled form fields, unlabeled buttons and links, color contrast failures, structural ARIA errors, missing page language declarations, and a range of keyboard accessibility issues.
The coverage boundary is well-established in accessibility research. Automated tools identify approximately 30-57% of WCAG success criteria violations. This is not a product limitation specific to any one tool; it is the boundary of what can be evaluated without human judgment. Issues that require understanding context, like whether alt text is descriptive enough for a complex diagram, or whether a data table is interpretable, require a person.
For e-commerce sites, the issues that automated tools catch reliably, missing alt text, unlabeled controls, form field labels, contrast failures, are also the issues most frequently cited in ADA web litigation. You can address the majority of litigation exposure and a majority of real user barriers through automated-tool-identified fixes alone.
One-time scans vs. ongoing monitoring
Free one-time tools like the axe DevTools browser extension, Google Lighthouse, and WAVE let you run a scan against a specific page. These are useful for spot-checking work in progress or verifying that a specific fix did not introduce a new issue.
Ongoing monitoring is a different use case. It rescans the full site on a schedule, compares each scan against the previous one, and surfaces what changed: what was fixed, what regressed, and what is new. For a store that ships developer changes regularly, this is the difference between a one-time snapshot and a continuous view of the accessibility state of the codebase.
What BarrierScan does
BarrierScan scans every page of your site on a schedule, not just the homepage. We prioritize what we find by the issue patterns most commonly cited in ADA web litigation, and we give your developers the specific WCAG criterion, the affected elements, and the HTML-level fix for each issue. On an ongoing subscription, each scan produces a diff report: what is fixed, what regressed, and what is new since the last scan.
We are honest about what automated scanning covers. Our scans catch roughly the same 30-57% of issue types that any automated tool covers. We are not a substitute for a manual audit if you need VPAT documentation or are preparing a legal defense. What we are: a cost-effective way to find and monitor the issues that matter most in your actual codebase, with fix guidance your developers can use the same day.
We do not use overlays and do not recommend them. Every finding we report is a real issue in your code with a real code-level fix. Nothing is patched at runtime and left broken at the source.
Which approach fits your situation
If your primary concern is ongoing code quality and monitoring as your site ships changes: automated scanning is the most cost-effective tool. Start with a scan, fix the findings, and establish a baseline you can track over time.
If you are preparing VPAT documentation, pursuing a federal contract, or responding to litigation: a manual audit from a certified specialist is the appropriate next step. BarrierScan is not a substitute for that.
If you have an overlay script installed and are considering whether to keep it: the FTC enforcement record and the litigation record are both clear. Addressing the underlying code is the lower-risk path.
The free homepage scan is a practical starting point regardless of which direction you are heading. It returns the critical and serious issues on your homepage in about 60 seconds, with the specific HTML fix for each one. No account required.