"Duplicating a winner restarts learning. Don't do it."

"Duplicating a winner restarts learning. Don't do it."

Half true. People say duping (copying a working ad set into a new one) always nukes your data. It doesn't.

The dupe DOES enter learning. But it inherits your account-level + pixel signal, so it often exits in hours, not days.

Here's the move when you want more spend on a winner:

— Dupe the ad set, keep the SAME budget, don't 2x it
— Launch off-peak (night) so it learns on cheaper traffic
— Watch hour 0-12 CPA, not the lifetime number
— Kill it only if it's 1.5x your target by hour 24

The reset is real but small. Stop treating it like a death sentence. Try it on ONE winner this week.