campaign-comparisonlisted

# Campaign comparison Small accounts generate pause-and-scale decisions weekly from data that cannot support one. This skill is mostly about refusing. ## Compare at the level the budget is set **Judge campaigns against campaigns.** Below-campaign breakdowns in a small account are almost always too thin to carry a decision, and the level that has a budget is the level where a decision changes anything. If you genuinely need an ad-group or ad-level verdict, check the conversion count at that level first. Usually that answers it. ## Pull more than one window A single window is a single sample. Pull the same metric over a short, a medium and a long window before any irreversible call. | Pattern across windows | Reading | |---|---| | Consistent across all three | Trust it | | Recent much worse, long-run fine | Could be noise, seasonality, or a real new problem — investigate, do not pause | | Recent much better, long-run poor | Do not scale on this alone. Most such moves regress | | Wild disagreement | You do not have enough data. Say so and stop | The long window is the one people skip, and it is the one that prevents pausing a campaign for having an ordinary bad fortnight. ## Cost per conversion, not conversion rate Conversion rate compares differently-sized things and hides the cost of getting there. Cost per acquired customer is what the budget actually buys, and it is comparable across campaigns that do completely different things. Where value varies a lot between c