ℹ️ Scope of this lens: This dashboard uses automated Lighthouse accessibility audits mapped to WCAG 2.2 guidelines to identify issues at scale. Full WCAG 2.2 conformance requires a dedicated accessibility assessment.
The Accessibility Lens is a monitoring and prioritization tool, not a WCAG conformance tool. It detects issues that can be tested automatically in rendered web pages, including problems involving HTML structure, ARIA attributes, accessible names, labels, contrast, and language.
Use it to find recurring issues, compare page groups, and focus remediation work on the templates and sections where it will have the greatest impact.
Accessibility glossary
Name | Description | Technical definition |
WCAG 2.2 | The Web Content Accessibility Guidelines published by the W3C. They provide a framework for making web content more accessible to people with disabilities. | WCAG 2.2 is organized around four principles and 13 guidelines. Its testable Success Criteria are assigned to conformance levels A, AA, or AAA. |
WCAG principles | The four foundations of WCAG: Perceivable, Operable, Understandable, and Robust. | Perceivable content can be presented in ways users can perceive. Operable interfaces can be navigated and used through supported input methods. Understandable information and interactions are clear and predictable. Robust content can be interpreted reliably by user agents, including assistive technologies. |
Success Criterion | A testable WCAG requirement, such as 1.1.1 Non-text Content. | Each Success Criterion belongs to a guideline and has a conformance level. Automated checks can provide evidence for some requirements, but they do not necessarily test the complete criterion. |
Accessibility score | A Lighthouse accessibility score for a page, displayed from 0 to 100. | A weighted average of the scored Lighthouse accessibility audits that apply to the page. It is not the percentage of WCAG 2.2 Success Criteria satisfied. |
Principle score | An Oncrawl score for one WCAG principle: Perceivable, Operable, Understandable, or Robust. | Oncrawl groups the applicable, scored audits mapped to each principle and calculates a weighted score for that group. |
Audit | An individual automated accessibility check, such as checking whether images have alternative text. | Each audit has a stable audit ID. Scored Lighthouse audits return pass or fail at page level and have different weights based on their potential impact. One audit may map to more than one WCAG Success Criterion or principle and therefore appear on several rows in the audit table. |
Criticality | An indication of how severely a failed check may affect users. | Oncrawl displays four levels: critical, serious, moderate, and informative. Higher-impact scored audits carry more weight in the score. |
Pages audited | The share of fetched pages for which accessibility audit data is available. | Pages with an accessibility score divided by the total number of fetched pages represented by the audited and not-audited groups. |
Applicable pages | The pages on which a specific audit had something relevant to evaluate. | Some audits return not applicable when the required elements or conditions are absent. For example, an image-related audit may not apply to a page without images. |
Not applicable (N/A) | The audit did not produce a scored result for the page, usually because it had nothing relevant to evaluate. | The page-level export leaves Score empty for this result. The audit is excluded from the page score calculation, so it neither improves nor reduces the score. |
For the complete standard, see Web Content Accessibility Guidelines (WCAG) 2.2.
Explore the Accessibility Lens
The summary contains the following charts:
Chart | What it shows | How it is calculated |
Average Accessibility Score | The average overall accessibility score across pages with audit data, displayed from 0 to 100. | The average of the page-level Lighthouse accessibility score for pages where |
Pages audited | The share of fetched pages with accessibility audit data. The gauge distinguishes Pages audited from Pages not audited. | Pages with an accessibility score divided by all fetched pages represented by the audited and not-audited groups. |
Total accessibility checks | The number of distinct automated accessibility checks reported across the audited pages. | Distinct audit IDs are counted once, even when an audit is reported on many pages or mapped to several WCAG Success Criteria. This is a count of audit types, not a count of WCAG 2.2 Success Criteria passed. |
ℹ️ Average Accessibility Score is a weighted Lighthouse audit score. It must not be interpreted as the percentage of WCAG 2.2 Success Criteria satisfied.
Accessibility score ranges by Page Group
[Screenshot placeholder — Accessibility score ranges by Page Group]
This chart compares score distributions across your Oncrawl page groups. Use the tabs to switch between:
Global Score
Perceivable
Operable
Understandable
Robust
For each page group, the stacked column shows the percentage of pages in each range:
Poor: 0–49
Needs Improvement: 50–89
Good: 90–100
The line shows the average score for the selected global or principle score within each page group. This lets you compare both the distribution and the average without treating one unusually high or low page as representative of the whole group.
Only pages with the selected score available are included in that tab.
WCAG principle scores
[Screenshot placeholder — WCAG principle score]
Each principle card displays the average page-level score for pages where that principle score is available:
Perceivable score: automated checks mapped to the Perceivable principle, including text alternatives, captions, contrast, and visual presentation.
Operable score: automated checks mapped to the Operable principle, including keyboard access, navigation, focus behavior, labels, and input interactions.
Understandable score: automated checks mapped to the Understandable principle, including language, labels, predictable behavior, and input assistance.
Robust score: automated checks mapped to the Robust principle, including valid ARIA and name, role, and value requirements.
The cards provide a shortcut to the corresponding principle overview. These values are averages of Oncrawl's page-level principle scores; they are not separate native Lighthouse category scores.
ℹ️ WCAG 2.2 Success Criterion 4.1.1 Parsing is obsolete and was removed from WCAG 2.2. The Robust score therefore focuses on the automated checks mapped to the current Robust requirements and related compatibility checks.
Accessibility audits
[Screenshot placeholder — Accessibility audits]
The Accessibility audits tables provides the detailed results behind the scores. Audit results are organized using:
The audit ID and title
The mapped WCAG principle
The mapped WCAG 2.2 Success Criterion
The WCAG conformance level
The audit criticality
The page-level result
The page-level export represents the result as:
1: the page passed the audit
0: the page failed the audit
Empty: the audit did not produce a scored result for the page and is treated as not applicable
Use the table to move from a principle or score to the specific automated checks and affected pages.
A single audit can map to more than one WCAG Success Criterion or principle. In this case, the same audit ID appears on several rows so that each mapping can be explored. These rows represent one Lighthouse audit result and do not cause the audit to be counted or scored multiple times in the overall Lighthouse score.
An automated audit mapped to a Success Criterion tests a detectable aspect of that criterion. Passing the audit does not prove that the complete Success Criterion has been satisfied.
How scores are calculated
Oncrawl renders the selected pages and runs Lighthouse accessibility audits against them.
As described in Chrome's Lighthouse accessibility scoring documentation, the Lighthouse accessibility score is a weighted average of accessibility audits. Each scored audit returns a pass or fail result for the whole page. If one relevant element fails a rule, the page fails that audit, even when other elements pass it. Audit weights are based on axe user-impact assessments.
The Lighthouse accessibility score for an individual page is the weighted average of its applicable scored audits:
Page score = Σ (audit weight × audit result) ÷ Σ applicable audit weights
Audits marked not applicable are excluded from the calculation. Manual audits and unscored best-practice checks also do not affect the Lighthouse accessibility score.
The Average Accessibility Score displayed in Oncrawl is the average Lighthouse accessibility score across pages where a11y_score is available.
The four principle cards average the corresponding page-level Oncrawl principle score across pages where that score is available. They help you compare the four principles but do not represent native Lighthouse category scores.
ℹ️ Accessibility scores summarize automated audit results. They do not represent the percentage of WCAG 2.2 Success Criteria satisfied and cannot establish WCAG conformance.
For details about audit weighting, see Lighthouse accessibility scoring.
How WCAG mappings affect the audit table
The WCAG mapping is used to organize audit results by principle, Success Criterion, and conformance level. It does not change the underlying Lighthouse result or cause an audit to contribute more than once to the overall Lighthouse score.
The supplied page-level export contains 47 distinct audit IDs represented by 50 mapping rows. The difference comes from three audit IDs that each map to two WCAG Success Criteria or principles. This illustrates why the number of rows in an export may be higher than the Total accessibility checks value.
The audit set and mappings can evolve with the Lighthouse version used by Oncrawl. A page-level export is also not a master list of every possible audit: the results and applicable checks depend on the page being analyzed.
Set up the Accessibility Lens
The Accessibility Lens uses Lighthouse data collected during the crawl. Before launching the crawl:
Open the Crawl JS section of your crawl profile.
Enable JavaScript crawling.
Under JS audits, enable accessibility audits.
Save the crawl profile and launch the crawl.
JavaScript audits require JavaScript crawling. Each analyzed URL consumes 5 crawl tokens in total, including the crawl, JavaScript rendering, and audit. This is a total cost, not an additional five tokens on top of the JavaScript crawl.
If both Core Web Vitals and accessibility audits are enabled, the total remains 5 crawl tokens per analyzed URL.
💡 Tip: Define a representative scope that includes your important templates and user journeys. Depending on your objective, relevant pages may include both indexable and non-indexable URLs.
FAQ
Is the Accessibility Lens a WCAG certification or conformance tool?
Is the Accessibility Lens a WCAG certification or conformance tool?
No. It covers the technical layer that Lighthouse can test automatically and maps relevant checks to WCAG 2.2 guidelines.
WCAG conformance requires evaluating all applicable requirements at the claimed level across complete pages and processes. Many requirements need human judgment or interaction testing. A dedicated accessibility assessment is therefore required to establish conformance.
Why does the number of applicable pages vary between audits?
Why does the number of applicable pages vary between audits?
Each audit has its own target and applicability conditions. An image-related audit may apply only to pages containing images, while a table-related audit may apply only to pages containing relevant table structures. The denominator therefore varies by audit.
Why are some results marked N/A?
Why are some results marked N/A?
N/A means that the audit had nothing relevant to evaluate on that page. The result is excluded from the page score, so it neither helps nor hurts the score.
Can a page pass every automated check and still have accessibility problems?
Can a page pass every automated check and still have accessibility problems?
Yes. Automated checks can detect whether an image has alternative text, for example, but they may not determine whether that text communicates the image's purpose accurately.
Automated testing is effective at finding many structural and technical problems at scale. Human review is still needed to evaluate meaning, context, complete keyboard journeys, usability with assistive technologies, and other requirements that cannot be assessed reliably through automation alone.
What does a failing-audit count of 17 mean for a page?
What does a failing-audit count of 17 mean for a page?
It means that the page failed 17 distinct audit rules. It does not count every affected element.
For example, if several elements fail the same duplicate-ID audit, that still counts as one failed audit rule for the page. Review the audit details to see the individual affected elements.