One client names the producing agency inside every ad name. So for three months our creative and their incumbent agency's ran in the same account, for the same buyer, under the same attribution. 212 of ours against 370 of theirs, compared line for line.
Every agency says its ads perform. Almost none can show it in the same account, under the same buyer, against the incumbent. This client's naming convention made that comparison possible, so we ran it.
Purchase campaigns only, Meta-attributed, straight out of the client's own ad account. No cherry-picked window: every ad both agencies shipped between early May and late July 2026.
The client's media buyer launched both sets of ads. We touched neither the spend nor the targeting, which is what makes the comparison clean.
Every ad we ship is graded on what it cost per purchase. The scoreboard here is the same one we keep on every account.
Our ads returned 16.69x against the incumbent's 4.2x, roughly four times the return. 2.97% CTR against 1.85%, so 61% more people clicked ours. Same account, same buyer, same three months.
Every story on this site ran on the same four moves. The story changes. The engine does not.
shows what already worked and exactly where it died.
lets us brief against what is winning right now.
gives us the customer's words, so we never invent an angle.
so winners compound into the next round.
