ROI gets fuzzy when every team brings a different scoreboard. Pick the money, the action that leads to the money, and the cost of getting that action. Then ignore the rest until those three are trustworthy.
Three lines, in this order
- What did it cost, including the making and the media?
- What commercial action did it create?
- What was that action worth, even if the estimate is rough?
Separate learning from scaling
A test can be worth it even when the ROI is poor, if you learned something you will use. A scale campaign cannot hide behind learning. Say which one you are running before it launches, and report it that way after.
Put the creative next to the number. If the winning idea is invisible in the report, the next campaign will not repeat it. The point of measurement is a better next decision, not a fuller dashboard.