CAC
Spend divided by the customers the campaign truly acquired.
Marketing analytics is not finding the channel with the prettiest ROAS. It is deciding which spend creates valuable customers that would not have arrived otherwise.
The important question is not “who got credit?” but what value did the spend actually create?
“Paid Social looks amazing on first-order ROAS. Should we move budget there?”
First-order attribution can reward channels that capture low-quality or already-converting customers.
A better analysis connects spend to acquisition, conversion, margin, retention and incremental lift.
Spend divided by the customers the campaign truly acquired.
Revenue relative to spend, with clear attribution rules.
Repeat rate, margin and retention after acquisition.
What happened because of marketing, not merely after exposure.
Attribution is an accounting rule; incrementality is a causal question.
Evaluate whether Paid Social budget should be increased.
Use:
spend, attributed orders, new paying customers, contribution margin,
30-day repeat rate, 90-day customer value.
Report:
CAC, first-order ROAS, contribution ROAS, repeat rate and customer value.
Explicitly state attribution assumptions.
Do not claim incrementality from observational attribution.
Suggest an experiment or holdout design if causal lift is required.Is the attribution window explicit?
Are new and returning customers separated?
Does revenue become contribution or profit anywhere?
Are channels compared on customer quality?
Could branded/direct demand be receiving paid credit?
Is incrementality distinguished from attribution?
Measure the cost of a meaningful acquired customer.
Compare downstream retention and margin.
State the rule and its limitations.
Use experiments or holdouts for incremental lift.
Whenever a channel looks efficient, ask what happens to those customers after the first conversion.