EZUD Continuous Accessibility Monitoring
Continuous Accessibility Monitoring
An accessibility audit is a snapshot. The moment it is complete, the product starts drifting. New features introduce new barriers. Content updates add images without alt text. Third-party scripts inject inaccessible components. Continuous accessibility monitoring detects these regressions before they accumulate into a compliance crisis.
Why Monitoring Is Necessary
Research from WebAIM’s annual Million analysis consistently finds that the most common accessibility issues on the web are basic, detectable violations: missing alt text, low contrast, empty links, missing form labels, and missing document language. These issues recur because they are constantly reintroduced through routine content and feature changes.
Without monitoring:
- Issues fixed during remediation quietly return.
- New team members, unfamiliar with accessibility standards, introduce new barriers.
- Content managed through a CMS is published without accessibility review.
- Third-party integrations degrade accessibility without anyone noticing.
Monitoring Layers
CI/CD Pipeline Checks
Integrate automated accessibility testing into the build pipeline so that every deployment is evaluated:
- axe-core (Deque): Integrate via @axe-core/playwright, cypress-axe, or jest-axe to run accessibility assertions as part of the test suite.
- Pa11y CI: Run against a list of URLs on every build. Fail the build if critical violations are detected.
- Lighthouse CI: Google’s Lighthouse includes accessibility scoring and can be configured as a CI check.
These checks catch regressions before code reaches production. See open-source accessibility tools for tool details.
Scheduled Scanning
Not all changes go through CI/CD. Content updates, CMS-managed pages, and third-party script changes require scheduled scans:
- Pa11y Dashboard: An open-source web-based tool that runs scheduled accessibility scans and displays results in a dashboard.
- axe Monitor (Deque): Commercial tool that crawls sites on a schedule and tracks accessibility health over time.
- Siteimprove Accessibility: Commercial platform with scheduled scanning, issue tracking, and reporting.
- Custom scripts: For organizations with specific needs, axe-core can be scripted to crawl pages on a cron schedule and output results to a database.
Run scheduled scans at least weekly. For high-change sites (news, e-commerce), daily scans may be appropriate.
Manual Spot Checks
Automated monitoring catches detectable violations but misses the 60-70% of issues that require human judgment. Supplement automated monitoring with periodic manual spot checks:
- Screen reader navigation of new features and content
- Keyboard testing of interactive components
- Focus management verification for dynamic content
Assign these checks to accessibility champions or rotate among team members.
User-Reported Issues
Users who encounter accessibility barriers are a valuable monitoring source. Ensure:
- The accessibility statement includes a prominent, accessible feedback mechanism.
- User reports are triaged through the same bug process as internally discovered issues.
- Response times are defined and met.
Alert and Escalation
Configure monitoring to alert the right people:
- Critical violations (e.g., form without labels, keyboard trap): Alert the team immediately via Slack, email, or PagerDuty.
- New violations introduced in a release: Alert the release owner and accessibility program manager.
- Trend degradation (conformance score dropping over consecutive scans): Alert program management and governance stakeholders.
Define escalation paths for situations where alerts are not addressed within defined timelines.
Metrics from Monitoring
Monitoring data feeds directly into accessibility metrics and dashboards:
- Page-level pass/fail rate: Percentage of monitored pages with zero critical automated violations.
- Issue trend over time: Total issue count by severity, tracked weekly or monthly.
- Regression rate: Number of previously fixed issues that have recurred.
- Mean time to detection: How long new issues exist before monitoring catches them.
- Mean time to resolution: How long issues persist after detection.
Choosing a Monitoring Strategy
The right monitoring approach depends on your organization’s context:
| Organization Size | CI/CD Checks | Scheduled Scanning | Manual Spot Checks |
|---|---|---|---|
| Small team, single product | Essential | Weekly | Monthly |
| Medium, multiple products | Essential | Weekly per product | Biweekly per product |
| Enterprise, many products | Essential | Daily for high-change, weekly for others | Monthly rotation |
Combine all three layers for comprehensive coverage. No single layer is sufficient.
Connecting to the Broader Process
Continuous monitoring is the sustain phase of the inclusive design process and the final phase of the accessibility roadmap. It ensures that the investment made in remediation, training, and governance is maintained over time.
Key Takeaways
- Accessibility regresses without continuous monitoring as content, features, and integrations change.
- Layer CI/CD checks, scheduled scanning, manual spot checks, and user-reported issues for comprehensive coverage.
- Configure alerts for critical violations and trend degradation with clear escalation paths.
- Feed monitoring data into accessibility metrics dashboards for organizational visibility.
- No single monitoring layer is sufficient; combine automated and manual approaches.
Sources
- https://www.deque.com/axe/ — Deque axe-core and axe Monitor for CI/CD integration and scheduled accessibility scanning
- https://pa11y.org/ — Pa11y and Pa11y Dashboard for open-source scheduled accessibility monitoring
- https://webaim.org/projects/million/ — WebAIM Million annual analysis showing that basic violations recur across the web without monitoring
- https://www.w3.org/WAI/test-evaluate/ — W3C WAI evaluation resources for combining automated and manual monitoring approaches