How A11y Owl Tests Your Website
A11y Owl combines real browser testing, vision AI analysis, and keyboard interaction checks to find accessibility issues that automated DOM-only scanners miss. Here is how it works.
1. How the Scan Works
When you submit a URL, A11y Owl performs a multi-step analysis of your page:
Real Browser Rendering
We load your page in a real Chromium browser using Playwright. This means we render JavaScript, execute dynamic content, and interact with your page exactly as a real visitor would. No static HTML parsing — we test the live page.
Screenshot Capture
We capture full-page screenshots of your site as rendered in the browser. These screenshots are analyzed by our vision AI to detect visual accessibility issues that cannot be found through DOM inspection alone, such as low-contrast text rendered on background images, small or unclear interactive elements, and visually confusing layouts.
Keyboard Interaction Testing
We simulate keyboard-only navigation through your page, testing tab order, focus indicators, keyboard traps, and interactive element accessibility. This replicates the experience of users who rely on keyboard navigation instead of a mouse.
Vision AI Analysis
Screenshots and page data are analyzed by Google Gemini vision AI, which evaluates the page from a visual and structural perspective. The AI identifies issues based on WCAG 2.2 criteria, maps each issue to specific elements, and generates actionable code fixes.
Report Generation
Results are compiled into a structured report with annotated screenshots, WCAG criterion references, severity ratings, and (in the Full Report) AI-generated code fixes for every issue found.
2. WCAG 2.2 Criteria We Check
A11y Owl tests against WCAG 2.2 Level AA success criteria. The following are the primary areas we evaluate:
Perceivable
- 1.1.1 Non-text Content (alt text)
- 1.3.1 Info and Relationships
- 1.3.2 Meaningful Sequence
- 1.4.1 Use of Color
- 1.4.3 Contrast (Minimum)
- 1.4.4 Resize Text
- 1.4.11 Non-text Contrast
- 1.4.12 Text Spacing
Operable
- 2.1.1 Keyboard
- 2.1.2 No Keyboard Trap
- 2.4.1 Bypass Blocks (skip nav)
- 2.4.3 Focus Order
- 2.4.6 Headings and Labels
- 2.4.7 Focus Visible
- 2.5.5 Target Size
- 2.5.8 Target Size (Minimum)
Understandable
- 3.1.1 Language of Page
- 3.2.1 On Focus
- 3.2.2 On Input
- 3.3.1 Error Identification
- 3.3.2 Labels or Instructions
Robust
- 4.1.1 Parsing
- 4.1.2 Name, Role, Value
- 4.1.3 Status Messages
3. Compliance Risk Score
The Compliance Risk Score is a proprietary metric that rates your website's ADA and WCAG 2.2 AA lawsuit exposure on a scale of 0 to 100.
How the score is calculated
The Compliance Risk Score considers multiple weighted factors, including:
- Number and severity of issues found (critical issues have a higher impact than low-severity issues)
- Types of issues detected (keyboard traps and missing focus indicators are weighted more heavily because they disproportionately affect assistive technology users)
- Coverage of key WCAG criteria tested
- Presence or absence of essential accessibility features such as skip navigation, proper heading hierarchy, and form labels
- Visual accessibility signals detected by vision AI, including contrast and touch target sizing
A higher score indicates lower lawsuit risk and better accessibility compliance. A score of 0 means significant accessibility barriers were detected; a score of 100 means no issues were found during the scan.
4. AIO Score (AI Search Visibility)
The AIO (AI Optimization) Score measures whether AI systems like ChatGPT, Google AI Overview, and Perplexity can find, understand, and cite your website content. It uses a 5-factor evaluation model:
Semantic Structure
Evaluates whether your page uses proper semantic HTML elements (header, nav, main, article, section, footer) instead of generic div elements.
JSON-LD Structured Data
Checks for the presence and quality of JSON-LD markup that helps AI systems understand your page content and relationships.
Heading Hierarchy
Analyzes whether your headings follow a logical H1, H2, H3 hierarchy that enables AI systems to parse content structure.
Answer-Ready Content
Evaluates whether your content is structured in a way that AI systems can extract direct answers from (clear definitions, lists, tables).
Schema.org Coverage
Checks for Schema.org markup that provides structured data about your business, products, services, and FAQs.
Each factor is scored individually and combined into an overall AIO Score from 0 to 100. A higher score means your content is more likely to be discovered and cited by AI systems.
5. Confidence Scoring
Because A11y Owl uses AI analysis, each identified issue includes a confidence indicator. This reflects how certain our system is that the issue is a genuine accessibility barrier.
High Confidence
The issue is clearly identifiable and well-defined. Examples: missing alt text on images, missing form labels, absence of skip navigation links.
Medium Confidence
The issue is likely present but may depend on context. Examples: color contrast that is close to the threshold, heading hierarchy that may be intentional.
Lower Confidence
The issue requires human review to confirm. Examples: reading order concerns based on visual layout, potential keyboard interaction issues that depend on custom JavaScript behavior.
6. Limitations
No automated tool can catch every accessibility issue. A11y Owl has the following known limitations:
What A11y Owl does not check
- Multi-page user flows (we scan one page at a time)
- Content behind authentication walls or login screens
- Video captions and audio descriptions (WCAG 1.2.x)
- Complex interactive widgets that require specific user input sequences to trigger
- Cognitive accessibility issues that depend on subjective user experience assessment
- Mobile-native app accessibility (we test web pages only)
- Third-party embedded content (iframes, embedded widgets) may have limited analysis
- Real assistive technology compatibility (we simulate, not test with actual screen readers)
According to the DHS/GovTech study, approximately 57% of WCAG issues require some form of manual testing. No automated scanner catches everything. A11y Owl catches more than DOM-only scanners because of vision AI, but a complete accessibility audit still benefits from manual expert review.
7. A11y Owl vs. Manual Audits
A11y Owl is designed as your first line of defense, not a replacement for manual audits. Here is how we compare:
| A11y Owl | Manual Audit | |
|---|---|---|
| Speed | Under 60 seconds | 1-4 weeks |
| Cost | Free (scan + full report) | $5,000 - $25,000+ |
| Code fixes included | Yes (AI-generated) | Recommendations only |
| Vision-based testing | Yes | Yes |
| Keyboard testing | Yes (automated) | Yes (manual) |
| Multi-page flows | Single page per scan | Full site |
| Screen reader testing | Simulated | Real devices |
| Best for | Quick scan, ongoing checks | Full certification |
We recommend using A11y Owl as your first step: scan your site, fix the issues found, and then pursue a manual audit if you need full certification or are responding to a legal demand. For ongoing compliance, regular A11y Owl scans help catch regressions before they become problems.
Need a full manual audit?
Our parent company, Logixtecs, provides professional accessibility audits and remediation services. Start with A11y Owl, then let our team handle the rest.