Value
Automated testing is a valuable tool; it can quickly find most occurrences of certain issues, with a thoroughness that would be very difficult to achieve manually.
Most of these issues are related to coding practices; for example, automated testing can quickly check if all image content (such as <img> elements) have text alternatives, which are needed by screen reader users.
Limitations
Unfortunately, automated testing can only detect 10 to 40% of accessibility issues; furthermore, automated tools can lead to false negatives (not detecting an issue), as well as false positives (reporting something that’s not actually a problem).
Since accessibility practices are designed to meet the needs of real people, most issues need to be manually tested by a human. For example, while automated testing can detect if text alternatives are present, only a person can check if these text alternatives are properly written and actually deliver the information they are supposed to. You can learn more about this on our text alternatives page.
Resources
- Building the most inaccessible site possible with a perfect Lighthouse score by Manuel Matuzo (2019)
- aXe testing tool by Deque
- Lighthouse testing tool by Google
- Siteimprove Accessibility Checker Chrome extension by Siteimprove
- Siteimprove, a service that provides analytics, including accessibility, on deployed sites