The third month of a paid programme tells you everything
Month one is setup. Month two is luck. Month three is the first read you can actually trust.
Everyone wants to judge a paid account in week two. It’s the worst possible time to judge it.
Why the first two months lie
Month one is structural: account restructure, tracking fixes, creative in flight. Fixing what a conversion even means usually comes first, and it is where most of the wasted spend hides. Month two often catches a tailwind or a headwind that has nothing to do with the work, and Google’s own bid strategies carry a learning period after significant changes that makes early reads doubly unreliable. Month three is the first clean read, when the changes have had time to settle and the noise has averaged out.
What we look for
Contribution, not ROAS in isolation. Whether the incremental spend is buying incremental customers or just re-buying ones you’d have got anyway. And whether the account can scale without the efficiency collapsing, which is the exact shape of the ROAS lift we hold up as proof: 3.5 to 4.5 that held while the account grew, not a screenshot week.
Judging on that clock is built into how we run paid media, with a three-month review where either side can walk. If your current agency is still selling you week-two screenshots, come and get a straight answer.