ROAS vs profit: measure contribution margin and MER instead

Four problems, each of which can turn a "profitable" 4.0 ROAS into a loss.

It is attributed, not actual. Meta's 7-day click, 1-day view window and Google's 30-day click window each claim conversions the other also claims, plus conversions that would have happened anyway. Summed platform revenue typically runs 150 to 250 percent of actual revenue for US ecommerce brands. A 4.0 platform ROAS is often a 2.0 to 2.5 true ROAS.

It is revenue, not margin. A $100 order at 4.0 ROAS cost $25 in ads. If the product costs $40, shipping $9, payment fees $3 and packaging $2, you have $21 left before overhead. Change the product cost to $55 and the same ROAS loses money.

It ignores discounts and returns. Platforms report gross order value at the moment of purchase. A 20 percent promo code and a 15 percent return rate take a third of that away.

It rewards existing customers. Retargeting and branded campaigns post high ROAS because they catch people who were already going to buy. Allocating by ROAS shifts budget toward the bottom of the funnel until new-customer volume dries up and the whole account decays.

Break-even ROAS

Definition: 1 divided by your contribution margin rate before marketing.

Contribution margin rate before marketing = (net revenue minus cost of goods minus variable fulfillment, payment and shipping costs minus returns allowance) divided by net revenue.

Example: net revenue $100, cost of goods $40, fulfillment and shipping $9, payment fees $3, returns allowance $5. Contribution before marketing is $43, or 43 percent. Break-even ROAS is 1 / 0.43 = 2.33. Any campaign under 2.33 true ROAS loses money on first purchase.

Work this out per product category, because a store's blended margin hides the fact that some categories cannot be advertised profitably at all.

Contribution margin after marketing

Definition: net revenue minus cost of goods minus variable costs minus total marketing spend (media, agency and tool fees, creative production). Express it in dollars and as a percentage of revenue.

This is the number that should go to the CEO and the one to set targets against. Healthy US direct-to-consumer brands typically target 10 to 25 percent contribution margin after marketing during growth phases, higher when optimizing for cash.

Marketing efficiency ratio (MER)

Definition: total revenue divided by total marketing spend for the period, using finance's revenue number, not platform numbers.

MER cannot be gamed by attribution because it uses the bank-account version of revenue. Pair it with break-even: if break-even ROAS is 2.33, your MER must exceed that just to cover variable costs, and must exceed roughly 3.0 to deliver a 20 percent contribution margin. Track MER weekly and trend it over 13 weeks.

Some teams also track new-customer MER or acquisition MER: first-order revenue divided by prospecting spend. It isolates the cost of growth from the revenue of the existing base.

A US home goods brand spends $120,000 a month. Platform dashboards report $540,000 attributed revenue, a 4.5 ROAS. Finance reports $360,000 in total revenue, of which $90,000 is from returning customers who did not click an ad.

• MER: 360,000 / 120,000 = 3.0.
• Contribution margin before marketing: 45 percent, so $162,000.
• Contribution margin after marketing: 162,000 minus 120,000 = $42,000, or 11.7 percent of revenue.
• Break-even ROAS: 1 / 0.45 = 2.22.

The 4.5 ROAS suggested the brand could scale aggressively. The 11.7 percent contribution margin says it is profitable but has little room, and a 15 percent increase in spend with flat MER would roughly halve profit. Those are different decisions from the same month's data.

1. Get the cost inputs from finance monthly. Cost of goods by SKU or category, average shipping cost per order, payment processing rate, packaging, and a returns allowance based on the trailing 90 days. Put them in a shared sheet with a version date.
2. Calculate break-even ROAS by category. Publish it to everyone who touches budgets.
3. Set up a weekly MER tracker. Revenue from the commerce platform or ERP, spend from finance or a consolidated spend sheet. Not from the ad accounts, which miss agency fees and creative costs.
4. Report contribution margin after marketing monthly, alongside MER and new-customer count. Put it on the executive dashboard; we describe the full set in marketing dashboard KPIs.
5. Pass profit to the platforms. Instead of sending order value as the conversion value, send margin-adjusted value: order value multiplied by category margin rate, or a fixed profit value per product. Google Ads and Meta can then optimize toward profit. For a Shopify store this is usually a configuration or small app change; see Shopify and ecommerce.
6. Separate new and returning customers. Tag campaigns by objective and exclude existing customers from prospecting with customer-match audiences. Measure prospecting on first-order contribution and expected 12-month value, not on immediate ROAS.
7. Run incrementality tests on the two biggest channels each year to calibrate how much platform ROAS over-claims. A geo holdout costing 5 to 10 percent of a channel's monthly budget usually pays for itself in the first budget decision it changes.

• Budget up when MER is above target and the marginal contribution of the last 10 percent of spend is positive. Check marginal, not average: a campaign at 3.5 average ROAS can be at 1.8 on its last dollars.
• Budget down when contribution margin after marketing drops below the floor for two consecutive weeks and no product, pricing or seasonal explanation exists.
• Judge prospecting on 60 to 90 days of cohort value, not the week it ran.
• Judge retargeting and brand search on incrementality, and cap them at a share of spend (often 15 to 25 percent) rather than letting ROAS pull budget into them.
• Never compare channels on platform ROAS. Compare them on deflated ROAS or test results.

For lead-generation businesses, replace revenue with closed-won revenue from the CRM, and contribution with gross profit on those deals. Cost per lead is the ROAS of lead-gen: easy to improve while destroying profit, because cheap leads are usually bad leads.

• Using platform ROAS to report to the board. It is a marketing-internal signal.
• Setting one ROAS target for the whole account. Categories differ; objectives differ.
• Forgetting fees in spend. Agency fees, tools and creative production often add 15 to 30 percent to media cost.
• Treating returns as someone else's problem. A 20 percent return rate in apparel changes break-even ROAS dramatically.
• Chasing ROAS by cutting prospecting. It works for six weeks and then new-customer volume collapses.
• Ignoring cash timing. Contribution margin is an accrual view; a brand with 90-day inventory cycles can be profitable and out of cash.

We run paid media as part of a commercial system, not a dashboard competition, through discover, plan, build, improve.

Discover. We reconcile the last six months of platform-reported revenue against finance, gather the cost inputs and calculate true break-even ROAS by category. Most clients learn in this step that at least one category has never been profitable to advertise.

Plan. We agree the targets: contribution margin floor, MER target, new-customer volume, and the share of spend allowed in retargeting and brand. We define what each campaign is for and how it will be judged.

Build. We restructure accounts by objective, implement margin-adjusted conversion values, set up customer-match suppression, build the MER and contribution tracker and connect it to the executive dashboard. Creative testing is set up so that winners are judged on contribution, not click metrics. This sits within performance marketing and ecommerce growth.

Improve. Weekly MER review, monthly contribution review, quarterly incrementality tests to recalibrate deflators. Budgets move by the decision rules above, not by whichever platform reported the best week.

Our Performance Growth Sprint is a representative engagement. The client's paid media had been scaled on platform ROAS, and the account looked excellent while profit did not move. We reset measurement around finance's revenue and the real cost structure, restructured campaigns by objective, and ran creative testing judged on contribution. Spend decisions moved from the ad platforms' numbers to the business's numbers, which is the pattern every engagement of this kind follows.