What Overlay Widgets Actually Do
Overlay widgets load as a third-party JavaScript tag and try to fix your WCAG violations at runtime — without touching your code. An overlay might inject aria-label onto unlabeled buttons, apply CSS filter adjustments to boost apparent contrast, or surface a floating toolbar that visitors activate to enter a "screen reader mode."
From a sales deck this looks like zero-dev-time compliance. In a courtroom it looks different: the widget cannot prove the fix worked. It fires after your page loads, so the remediation exists only in the visitor's current browser session. No audit log, no before-and-after geometry measurement, no evidence the patched element met the criterion at the moment a disabled user interacted with it. That absence is what plaintiff attorneys now surface in discovery.
The Criteria They Cannot Evaluate
WCAG 2.2 added five AA criteria that require real-browser layout measurement. Focus Not Obscured (2.4.11) fails when a fixed header covers a focused element — detectable only by tabbing through a live Chromium instance and comparing bounding rects. Target Size Minimum (2.5.8) requires checking rendered pixel dimensions across viewports. An overlay that fires after initial paint cannot evaluate either. For the full breakdown, see our WCAG 2.2 New Requirements guide.
Plaintiff attorneys know overlays do not satisfy WCAG. They grep your page source for the overlay script, confirm it is present, then run the same scanner they would use on any site. The overlay runs in the visitor's browser — not in the plaintiff's headless instance. Your violations appear in both.
The ADAflags Evidence-Only Scan Path
ADAflags runs five stages of evidence gathering on every scan, covering WCAG 2.1 AA and the five new 2.2 AA criteria. Each stage produces raw evidence rows and never a conformance claim the runner did not actually verify. For the lawsuit-risk context that frames these criteria, see our ADA Website Lawsuit Risk guide.
Phase 0 — axe-core Baseline
Phase 0 runs axe-core against a JSDOM parse of your HTML. It catches the static-analyzer failures — missing alt text, unlabeled form fields, insufficient contrast — that drive most ADA demand letters.
Phase 1A, 1B, 1C — Real-Chrome Geometry Runners
Phase 1A (Focus Not Obscured, 2.4.11) opens headless Chromium, tabs through up to 50 elements per viewport, and records whether any position: fixed or position: sticky rect overlaps the focused rect. Phase 1B (Target Size Minimum, 2.5.8) runs axe's experimental target-size rule across mobile and desktop at zoom 1 and 2. Phase 1C (Dragging Movements, 2.5.7) verifies any drag-only UI pattern has a single-pointer alternative. All three persist raw evidence with conformanceInference: null and isConformanceClaim: false.
Phase 2 — Accessible Auth and Error Prevention
Phase 2 evaluates 3.3.8 Accessible Authentication (login cognitive tests without alternatives) and 3.3.4 Error Prevention (checkout or account deletion confirmation/reversal paths).
Every ADAflags result row carries conformanceInference: null and isConformanceClaim: false. We report what real Chrome observed — the same class of evidence a court-appointed accessibility expert would gather — and never assert a conformance pass on criteria that require human judgment for a final determination.
Compare the Overlay Vendors
Each of the three largest overlay providers markets WCAG conformance claims and has been sued by users whose assistive technology failed to interact with the injected DOM:
ADAflags vs. AccessiBe →
How AccessiBe's overlay toolbar holds up against WCAG 2.2 geometry rules.
ADAflags vs. UserWay →
UserWay's injected focus indicators against real-Chrome target-size measurements.
ADAflags vs. AudioEye →
AudioEye's "industry-leading" claims against measured focus obscuration.
Build Your Defensible Evidence Trail
ADAflags' free scan attaches real-Chrome geometry evidence to every WCAG 2.1 and 2.2 result. No overlay, no estimates, no conformance claim we didn't actually verify.
Run My Free Scan →Frequently Asked Questions
Can an overlay widget satisfy WCAG 2.2 AA?
No. WCAG 2.2 specifically tightens focus visibility and target size — the overlay toolbar directly affects both. Courts have repeatedly held overlays do not satisfy WCAG 2.1 AA; the new 2.2 AA criteria make the argument even harder.
What makes ADAflags scans "evidence-only"?
Every result row carries conformanceInference: null and isConformanceClaim: false. The scan reports what real Chrome observed and never asserts a conformance pass on criteria that need human judgment.
How often should I scan my site?
At minimum, any time you deploy new content, add third-party scripts, or change navigation or checkout. ADAflags' Starter ($9.99/mo) and Pro ($29/mo) plans rescan monthly and email the deltas so regressions are caught early.
Related reading: ADA Website Lawsuit Risk · WCAG 2.2 New Requirements · vs. AccessiBe · vs. UserWay · vs. AudioEye · Starter ($9.99/mo) and Pro ($29/mo).