You're evaluating a product growth experiment on a user-facing surface with clear day-of-week and monthly usage patterns. The team wants to know how seasonal traffic swings should change the way you design and read the test.
How would you think about the impact of seasonality on a product growth experiment?