Small experiments beat big bets when you are exploring formats. Ship a prototype, measure for a week, then decide.

Document what you learned even when you stop.