Diagnose inconsistencies in issue counts

Issue counts in accessibility tests can vary because of website changes, dynamic content, scan settings, or rule updates. Find the cause and remediate it.

Issue counts reported by BrowserStack Accessibility Testing can differ between two scans of what looks like the same website or workflow. Most of these differences have a known cause, and many of them are expected rather than a defect.

Each section below summarizes one cause. Open the linked topic for the full root cause, the affected products, and the remediation steps.

Website updates

If you or your team deployed changes to content, HTML, CSS, or scripts between two scans, the issues detected can change. Those changes might have fixed old issues or introduced new ones, so a different count is an expected outcome.

For the full explanation and remediation steps, see Inconsistencies due to website updates.

Differences across products

Workflow Analyzer captures real-time DOM states, including dynamic and user-interacted content, so it might detect more issues than a static website scan. A difference between reports from two different products is expected.

For the full explanation and remediation steps, see Inconsistencies across accessibility products.

Differences between Workflow Analyzer scans

Two Workflow Analyzer reports of apparently the same workflow can differ for three reasons. The workflow varied slightly, a page state change aborted the scan, or the extension didn’t finish post-processing after you saved the report.

For the full explanation and remediation steps, see Inconsistencies between Workflow Analyzer scans.

Scan settings

Settings for WCAG version, Advanced Rules, Best Practices, and Potential issue control which rules run. If two scans use different settings, a difference in issue count is expected.

For the full explanation and remediation steps, see Inconsistencies due to scan settings.

Hidden issues

Hidden issues are excluded from the issue count, and unhiding an issue adds it back. If you hid or unhid issues between two scans, the counts differ.

For the full explanation and remediation steps, see Inconsistencies due to hidden issues.

Page load and network failures

Pages that fail to load are not scanned, so 404 errors, timeouts, network drops, and VPN restrictions all reduce the number of issues reported.

For the full explanation and remediation steps, see Inconsistencies due to page load and network failures.

Spectra™ Rule Engine updates

The Spectra™ Rule Engine is updated regularly to improve issue detection. Scans run on different engine versions can report a different number or type of issues.

For the full explanation and remediation steps, see Inconsistencies due to Spectra™ Rule Engine updates.

Dynamic and personalized content

Ads, popups, carousels, and cookie consent banners can render differently on each run. Geolocation-based content, A/B test variations, and session-based personalization change what the scanner sees.

For the full explanation and remediation steps, see Inconsistencies due to dynamic and personalized content.

Other causes

If none of the causes above explain the difference, the scans might need a deeper analysis. Submit both scans for manual review through the support escalation process.

For the full explanation and remediation steps, see Investigate other causes of inconsistencies.

Related topics

• Choose the right accessibility tool
• Accessibility Score calculation
• Severity classification
• Impact of rule engine customization