What is an A/A test and why is it needed?

A/A test is an experiment where two groups of users receive the same version of a product. Its goal is to verify the correctness of the A/B testing system, ensure that metrics and statistics work correctly, and that random differences between groups do not affect the results.

Example of use: before launching a real A/B test, an A/A test is conducted to ensure that the tool does not show false differences and that the user sample is evenly distributed.