“I tend to have like the pillars — the channels that are working where I continue investing massively in this channel — usually 30% will go to just testing new things, and in the meantime I will expand my budget like starting to explore new channels knowing that for some of them it's going to take longer.”

Reserve 30% of paid budget permanently for testing new channels

Concrete portfolio rule: keep pillar channels funded, but always carve out roughly 30% of paid spend for testing new ones. This guards against platform dependency and creative fatigue while compounding learning across the funnel — so when a pillar saturates you already have the next winner warming up.