Formulas used by this calculator
- Contribution before media = retained revenue − included variable costs
- Contribution margin (%) = contribution before media ÷ retained revenue × 100
- Pre-overhead break-even ROAS = retained revenue ÷ contribution before media
- Media allowance = contribution before media − contribution to retain
- ROAS required for the reserve = retained revenue ÷ positive media allowance
Worked example
Illustrative retained revenue per order: $100. Included variable costs: $60. Contribution to retain: $10.
Contribution before media is $40, or 40%. Pre-overhead break-even ROAS is 100 ÷ 40 = 2.5×.
Keeping $10 leaves a $30 media allowance. The corresponding ROAS threshold is 100 ÷ 30, approximately 3.33×. Calculations use unrounded values; displayed ratios are rounded.
Build the cost definition first
Use retained revenue for a representative order after the discounts and refund treatment included in your model. Exclude taxes collected on behalf of others. If expected refunds are already reflected in revenue, do not subtract the same adjustment again as a cost.
Include the variable costs that belong to serving the order: product, fulfillment, payment charges, shipping subsidies, and any other relevant cost. Advertising goes outside that field because the calculator is solving for the amount available to spend on media.
The Small Business Administration explains contribution margin and distinguishes variable costs from fixed costs in break-even analysis. This calculator applies that contribution concept to an advertising allowance; it is not a complete business profit-and-loss model. U.S. Small Business Administration: contribution and break-even analysis.
Distinguish media break-even from the amount you need to retain
At the pre-overhead threshold, the modeled order covers included variable costs and media while leaving nothing for overhead or another reserve. Calling that business profitability would hide costs the calculation has not covered.
Enter a retained contribution amount when the order needs to support overhead, non-media acquisition costs, or another operating objective. Make the allocation explicit and avoid counting the same cost in both the variable-cost field and the reserve. The reserve is your assumption, not a recommended margin supplied by the tool.
If contribution is positive but the requested reserve consumes all of it, the model cannot support positive media spending. If the reserve exceeds it, the negative allowance shows the shortfall. Revisit the commercial assumptions instead of interpreting that negative number as a budget.
Compare the threshold with a compatible revenue measure
A platform may report gross attributed sales while this calculator uses retained revenue. The two ROAS figures are not directly interchangeable. Reconcile the numerator before concluding that the campaign is above or below the modeled threshold.
This model uses one representative order. Product mix, discounts, shipping zones, and returns can change the average. Compare meaningful cohorts when one average conceals a segment with very different economics, and retain the assumptions with the downloaded worksheet.
A threshold is not a prediction of what a bidding system can achieve. Nor does being above it prove the ads caused incremental sales. Use it as one input to a budget review alongside attribution quality, demand, cash timing, and operational capacity.
Common questions
Why is break-even ROAS the inverse of contribution margin?
When margin is expressed as a fraction of retained revenue, that fraction is the amount available for media at pre-overhead break-even. Dividing revenue by that allowance produces 1 divided by the margin fraction. A 40% margin therefore corresponds to 1 ÷ 0.40 = 2.5×.
Should I use gross margin instead?
Only if its included costs match the decision. A gross-margin figure that omits fulfillment, payment fees, or other relevant variable costs can overstate the amount available for advertising. Build the cost definition rather than relying on a label.
Does this include repeat purchases?
No. It models one order. Use realized customer cohorts and a separate payback scenario if the business deliberately relies on repeat contribution to recover acquisition costs.