Break-even CPA answers, “How much can I spend to acquire this order before modeled contribution reaches zero?” Target CPA answers a stricter question: “How much can I spend and still keep the contribution I want?” Mixing the two produces acquisition targets that are either unnecessarily conservative or quietly unprofitable.
Define the two ceilings from the same cost stack
break-even CPA = revenue − all non-ad variable coststarget CPA = break-even CPA − target contribution per acquired orderThe non-ad variable-cost stack should include product cost, fulfillment, seller-paid shipping, payment/platform fees and an expected return-loss allowance. The target contribution is a management choice: the amount you want left after acquisition to help cover fixed costs, operating profit and risk.
Worked example: a $60 order
Suppose net order revenue is $60. Product cost is $18, shipping subsidy is $6, fulfillment is $3, payment and platform fees total $2.10, and expected return loss is $1.90.
pre-ad contribution = 60 − 18 − 6 − 3 − 2.10 − 1.90 = $29.00The break-even CPA is therefore $29. If you require $10 of contribution after acquisition, target CPA is $19.
| Acquisition cost | Contribution after acquisition | Interpretation |
|---|---|---|
| $15 | $14 | Above the $10 target |
| $19 | $10 | At target CPA |
| $24 | $5 | Profitable but below target |
| $29 | $0 | Break-even |
| $34 | −$5 | Negative modeled contribution |
Convert CPA ceilings into ROAS guardrails
ROAS can be derived from the same order economics:
required ROAS = order revenue ÷ allowed CPAAt $60 revenue and $29 break-even CPA, break-even ROAS is about 2.07. At a $19 target CPA, target ROAS is about 3.16. The exact threshold changes whenever price, discount, product cost, returns, shipping or fees change.
Stress-test before changing bids
- Recalculate after a promotion or average-order-value shift.
- Use a higher expected-return loss for cohorts with worse return behavior.
- Separate new-customer and returning-customer economics when acquisition goals differ.
- Model platform-attributed CPA and blended acquisition cost as different views, not interchangeable truths.
- Set a target contribution buffer rather than running permanently at break-even.
Decision rule
Use break-even CPA as the hard economic ceiling and target CPA as the operating guardrail. If actual acquisition cost stays between the two, the order may still be contribution-positive but is missing the profit target; that is a deliberate tradeoff, not automatically a successful campaign.
Use the Break-even ROAS Calculator with your own cost stack, then compare the result with the $60 worked example.
Use the thresholds to interpret, not just label, a campaign
The downloadable table retains the $29 pre-ad contribution, $60 revenue and $10 target. At a $24 CPA, realized order ROAS is 2.50 and contribution is $5. Calling that campaign profitable is incomplete: it is positive before fixed overhead, but misses the target by $5 per acquired order. At 100 comparable orders the modeled shortfall is $500.
When target contribution equals $29, allowed CPA is zero. Required ROAS is not a useful finite number in that case. When target contribution exceeds $29, no nonnegative advertising cost can meet it. Improve non-ad economics or revise the target rather than forcing a positive bid out of an impossible equation.
Keep conversion and cost windows aligned
Choose the order cohort, attribution window, currency and treatment of returns before evaluating the threshold. Dividing today's spend by today's purchases can mismatch the time at which customers clicked and bought. A platform CPA and a blended store acquisition cost are different measurements; neither should silently replace the other.
For a controlled budget decision, record the measured CPA, the threshold calculation and which cost assumptions are still provisional. Recheck after a promotion, a different product mix, a carrier-cost change or a mature return cohort. A threshold is a model boundary, not a promise that changing a bid will generate orders at that price.