Most advertising reports answer the wrong question. They tell you how many people saw an ad, how many clicked, and what each of those clicks cost — and stop precisely where the interesting part begins. None of that tells you whether the money came back. The fix is not a better dashboard; it is deciding, before a campaign runs, which single number counts as a result, then refusing to be distracted by the rest.
Start by writing down what a customer is worth to you over a year. Without that number every other number floats. With it, a cost per acquisition stops being an abstract benchmark and becomes a simple question: are we paying less than we get back, and how confident are we in the gap?
“If you cannot say what a customer is worth, you cannot say whether an ad was expensive. Every measurement problem we are asked to fix starts there.”
The second step is boring and unavoidable: get one source of truth. When the ad platform, the analytics tool and the finance spreadsheet disagree, teams default to whichever number is most flattering. Pick the conservative one and make it the only one anybody quotes.
From there the reporting almost writes itself: spend, results, cost per result and value returned, per channel, on one page.
None of this requires expensive tooling. It requires agreeing what counts as a win before the money is spent, and being willing to turn off the campaigns that never get there — including the ones somebody senior is fond of.
Written by the NEHIM strategy team. We publish what we learn running campaigns and building the products behind them.
3 Comments
Finally an article that says the quiet part out loud. We killed two campaigns after running this exercise.
The single-source-of-truth point is the one our team keeps failing. Sending this to our analyst.
Would love a follow-up on how you handle attribution across offline sales.