Why does a test fail? In Cypress Dashboard terms - test flake (can be because of the test or the system) - top failures (it retried n times, still didn't work) - real Service/App failure - environment instability - system dependency failure There can be 3 different kinds of failures. Without test retries you will not be able to differentiate between those real failures, and the lack of test quality. You must have test retries enabled so that you can handle important failures while working on improving test quality; it is the key to isolating the two efforts.