kill fatigue on a winner instead of retiring it

an ad that worked and then stopped did not become wrong. it became familiar. the shape was right — the open, the turn, the beat the offer lands on — and the only thing that changed is that the people you are buying have already seen it.

shelving it throws away the one thing you had proof of. starting over means paying to test a structure from zero, on a new idea, when you already bought that answer once.

the grid stays. the cuts stay where they were. the camera does what it did. the words land on the same beats. one thing moves.

on a rebuilt ad the room is not one of the three. it wears the look of the reel you forwarded, so there is nothing there to move. what it has instead is the voice: an ad with nobody on screen can be recast into a different one without a face ever appearing in it, which is the cleanest single variable in the whole product.

change the face and the room together and a lift tells you nothing: you have two candidate explanations and no way to separate them. change one and the result has a name you can spend against.

it is the least glamorous rule on this page and it is the one that makes the rest of it worth anything.

it cannot see your account. it does not know your spend, your cost per result or how long anybody watched, and it will not pretend to. what it can do is hold everything still except the one thing you asked it to change, and hand back a video where that is genuinely true. the reading of the result is yours.