The point in executing performance tests is less a green or red outcome than the insights gathered from the runs. The test reports and the application's behavior are collected during the test runs. These collections enable us to gain experience and discover trends in performance.

While the performance tests can be executed standalone, they ideally run continuously as part of a Continuous Delivery pipeline. It's already helpful to gain these insights without needing to affect the outcome of the pipeline step. After some metrics have been gathered, engineers can consider defining a performance run as failed if the measured performance indicated a major drop from the usual expectations.

This matches the idea of continuous improvement or in this case avoiding responsiveness deterioration.