You notice that a key product metric moved, and the team wants to know whether it reflects a real change or a recurring calendar pattern. You have historical daily data and need to separate normal seasonality from an unusual shift.
How would you investigate whether a metric moved because of a seasonality pattern?