1. Define value
In a good roas program, define value before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
2. Define eligible spend
In a good roas program, define eligible spend before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
3. Align scope and currency
In a good roas program, align scope and currency before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
4. Choose attribution boundaries
In a good roas program, choose attribution boundaries before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
5. Account for margin and reversals
In a good roas program, account for margin and reversals before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
6. Set maturity rules
In a good roas program, set maturity rules before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
7. Calculate the baseline
In a good roas program, calculate the baseline before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
8. Model scenarios
In a good roas program, model scenarios before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
9. Compare marginal return
In a good roas program, compare marginal return before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.
10. Scale or stop
In a good roas program, scale or stop before advancing. Document the hypothesis, responsible owner, input evidence, accepted output, deadline and stop condition so the decision can be reproduced.
The output of this good roas step should be understandable to a reviewer who did not create the campaign or page. That discipline reduces hidden assumptions and improves future iteration.