Growth experiment bounds

Growth experiments test different approaches to acquisition, activation, or retention. However, not every growth tactic should be tried just because it might work. Some experiments can harm users, violate their trust, or create negative experiences even if they improve your metrics.

Establishing experiment bounds means defining what you won't test regardless of potential results. This might include not experimenting with deceptive patterns, not testing tactics that exploit vulnerable users, or not running experiments that significantly degrade the experience for a test group. Your bounds should reflect your values and the long-term relationship you want with users.

Before running growth experiments, consider the user experience for all groups, potential unintended consequences, and whether short-term gains justify any negative impact. Document your ethical guidelines for experimentation and review tests against these standards. Strong growth teams know when to stop an experiment that's working numerically but creating problematic user experiences or violating trust.