The short answer

A good ROAS is one that supports the economics and purpose of the advertising you are measuring. A revenue ratio alone cannot establish that. Start with the costs behind the attributed orders, then decide how much contribution must remain after ads.

Find your own floor

For an order-level model, subtract non-ad variable costs from revenue. Divide revenue by the amount remaining to calculate break-even ROAS. This floor covers only the costs included; it is not automatically a whole-business profit target.

Break-even ROAS = revenue ÷ (revenue − non-ad variable costs)

Illustrative inputs: two offers each sell for $100. Offer A has $50 in non-ad variable costs; offer B has $75. Their break-even ratios are $100 ÷ ($100 − $50) = 2× and $100 ÷ ($100 − $75) = 4×. At 3× ROAS, advertising costs $100 ÷ 3 ≈ $33.33 per order. Offer A retains about $16.67 after ads; offer B loses about $8.33 before any excluded overhead.

Separate a floor from a useful target

If offer A also needs $20 per order for overhead and a desired surplus, its ad allowance falls to $50 − $20 = $30. Its required ROAS becomes $100 ÷ $30 ≈ 3.33×. The $20 is an explicit planning choice in this example, not an industry standard. Changing that requirement changes the target.

Compare like with like

A retargeting campaign reaching previous customers and a prospecting campaign reaching new buyers answer different questions. Compare their customer mix, attribution window, refunds, and contribution as well as ROAS. Neither result proves how many sales advertising caused. Also check scale: a high ratio on a tiny spend may contribute less money than a lower but profitable ratio at larger volume.

Make a decision you can explain

  1. Write down the revenue and cost definitions.
  2. Calculate break-even using the current product mix.
  3. Add the contribution requirement for your decision.
  4. Compare mature campaign data against that threshold.
  5. Review cash availability and uncertainty before increasing spend.

If the underlying data is incomplete, label the result provisional. A benchmark cannot supply missing cost or attribution information.

Set your own threshold

  1. Gather: Collect current order revenue, variable costs and the contribution you need to retain.
  2. Produce: Compare mature actual ROAS with the threshold on the same revenue basis.

Decision boundary: Do not copy a target from a business with different margins.

Next reading

Examples are illustrative, not client results or benchmarks. The calculations use the inputs shown; real-world decisions depend on your records and assumptions. Report a correction.