Start with margin
Calculate revenue after discounts and refunds, then subtract variable costs such as product, fulfilment and payment fees. If an order contributes 40 before advertising, 40 is the approximate break-even acquisition cost before other costs. Include management and creative costs when assessing the programme.
Start with what the order contributes
For ecommerce, subtract product cost, payment processing, variable shipping, expected returns and other relevant variable costs from revenue. That result is contribution before advertising. Compare it with the cost to win the order. For lead generation, work from qualified lead to signed job using measured close rates and contribution per customer. Do not assume that every lead or every attributed purchase represents new business.
- Assess the orderSubtract costs that vary with each sale
- Find acquisition costInclude the campaign expenses you need to judge
- Compare and reconcileCheck returns, duplicates and CRM outcomes
Follow a lead through to a sale
For a lead campaign, estimate the share that qualifies and the share of qualified leads that buys. If ten leads cost 250 and one customer contributes 400 before marketing, the campaign has 150 left before overhead. Small samples can swing sharply, so review a longer period.
Calculate a break-even threshold you can monitor
Suppose a 100-dollar order contributes 35 dollars before advertising. The break-even acquisition cost is 35 dollars if no other campaign-specific costs apply. Revenue divided by ad spend gives ROAS. At 35% contribution margin, break-even revenue ROAS is roughly 2.86 before other campaign-specific costs. If testing, production or agency fees belong in the analysis, track them separately or include them explicitly in the model.
Compare reported returns with actual orders
Reconcile platform reports with transaction IDs and CRM outcomes. Distinguish new from returning customers and delayed purchases from immediate ones. Adjust attribution windows and refund treatment consistently before comparing campaigns.
Reconcile platform reports with actual business outcomes
Match ad costs with order or CRM data and adjust for refunds, duplicates and qualified leads. Use a timeline long enough for purchases and cancellations to settle. Evaluate brand terms separately and check the sensitivity of your conclusion to different attribution windows. When you can, compare areas or periods with a controlled shift in spend to understand incremental effect.
Why a strong ROAS can still lose money
A store reports 400 dollars of attributed revenue on 100 dollars of ad spend, or 4× ROAS. If the orders contribute 20% before ads, the contribution is only 80 dollars. The campaign is short 20 dollars before additional campaign costs. The next step is to improve margin, reduce acquisition cost or test a different offer.
Build two versions of the model when repeat purchases matter. Start with contribution on the first order. Then show a repeat-purchase estimate backed by a net revenue cohort and a defined window. Include the risk that repeat orders may arrive later or less often than expected. If campaigns appear profitable only in an optimistic model, keep the pilot small until observed margins and repeat orders support the higher acquisition cost.
| Metric | Formula or input | Decision supported |
|---|---|---|
| ROAS | Attributed revenue divided by media spend | How efficiently spend converts to recorded revenue |
| CPA | Relevant spend divided by acquisitions | Whether the acquisition cost fits contribution |
| Profit contribution | Contribution before ads minus relevant campaign costs | Whether the work creates value under stated assumptions |
Common questions
Should I set target ROAS equal to break-even?
Build in room for uncertainty, operating costs and the profit your business needs. Validate the target against actual margins by product or service.
How do I estimate a defensible cost per lead?
Multiply contribution per won customer by the measured share of leads that convert into won customers. Adjust for selling and fulfilment costs where relevant.
Why should branded campaigns be reported separately?
People searching for your brand may already be ready to buy. Separating branded traffic keeps a high reported ROAS from masking performance on campaigns intended to acquire new customers.
What to check next
- Break-evenCalculate contribution before media costs
- Lead economicsFollow enquiries to closed sales
- ReportingRemove refunds and deduplicate orders




