An account I reviewed last quarter had an 11x blended ROAS. The client was thrilled. The agency was thrilled. The dashboard was green across every column. And the business was bleeding cash on the average order.
This is not a cautionary tale about fraud or misattribution. The tracking was clean. The math was correct. The problem was that ROAS, even at 11x, answers the wrong question. It tells you how much revenue came back per dollar of ad spend. It does not tell you whether that revenue left any profit after you paid for the product, shipped it, processed the return, and handed VAT to the government.
The Gap Between Platform Truth and P&L Truth
The account in question was an ecommerce brand with a healthy mix of brand and nonbrand campaigns. As reported this week, the blended 11x figure meant ad spend was roughly 9% of reported conversion value. That sounds efficient. It is efficient, if you stop at the platform view.
The platform view, though, included VAT in the conversion value. It included revenue from orders that would later be returned. It did not subtract product cost, fulfillment, payment processing fees, or the clearance markdown that moved stuck inventory. Once all of those costs were layered in, the average order was underwater.
Brand campaigns were running at roughly 18x ROAS. Nonbrand campaigns were closer to 3x. The blended number looked strong because brand traffic, which would have arrived anyway, was doing the heavy lifting. The incremental result from nonbrand was far weaker than the dashboard suggested.
Why High ROAS Can Still Mean Negative Contribution
The formula most marketers use is simple: Revenue ÷ Ad Spend = ROAS. A 4x means four dollars back for every dollar spent. The formula is correct. The interpretation is where things break.
Break-even ROAS is the minimum return at which an ad actually pays for the product it sells. The calculation is Selling Price ÷ Margin After Fulfillment. If you sell a product for $100 and your margin after product cost and fulfillment is $67, your break-even ROAS is 1.5. Anything above that makes money. Anything below loses money, regardless of how impressive the multiple looks.
The 11x account had a break-even ROAS around 2.8 once you included all variable costs. The nonbrand campaigns at 3x were barely clearing the bar. After returns (which ran higher on paid acquisition than on organic), several product categories were actually below break-even. The blended number masked the problem because brand campaigns, with their 18x return, were subsidizing the loss.
Industry benchmarks show median Meta ROAS across roughly 35,000 ecommerce brands at 1.93x and median Google Shopping at 3.68x. Those figures vary by category and attribution methodology, but they illustrate the point: a 3x nonbrand ROAS is not exceptional. It is roughly average. And average, for a business with thin margins and high return rates, can mean losing money on every incremental order.
The Brand Traffic Problem
Brand campaigns are the silent distorter of blended metrics. A customer who already knows your name, who typed your brand into Google or clicked a retargeting ad after browsing your site, is not the same as a customer you acquired cold. The first would likely have converted anyway. The second is the one your ad spend actually created.
When brand and nonbrand are blended into a single ROAS figure, the high-intent traffic inflates the number. The 11x account was a textbook case. Strip out brand, and the picture changed. The campaigns that were supposed to drive growth were barely breaking even, and in some product categories, they were not.
This is not an argument against brand campaigns. Protecting your brand terms from competitors has value. But counting that traffic as proof of ad efficiency is like counting your salary as investment income. The money was coming in either way.
What Changed After the Restructure
The account was restructured with a different objective: contribution per order, not ROAS. Campaigns were evaluated on whether they cleared break-even after all variable costs, not on whether they hit a ROAS target that looked good in a slide deck.
The results were counterintuitive if you only watched the dashboard. Blended ROAS fell. The green numbers got smaller. But contribution per order turned positive. Cash arrived during the selling season instead of at clearance, when margins collapse. The business stopped subsidizing growth with margin it did not have.
The specific changes were not exotic. Nonbrand budgets were reallocated toward product categories with higher margin after fulfillment, which meant lower break-even thresholds and more headroom for profit. Campaigns targeting high-return customer segments were paused. The team stopped optimizing for conversion value and started optimizing for contribution, which required piping margin data into the measurement stack.
The Model Your CFO Needs
If you are reporting ROAS to your board without a break-even threshold next to it, you are reporting a number that cannot be interpreted. A 4x ROAS is meaningless without knowing whether 4x is above or below the floor. A 10x ROAS is meaningless if the conversion value includes VAT and pre-return revenue that will never hit the bank account.
The model is not complicated. It requires three inputs: selling price, product cost, and fulfillment cost. From those, you calculate margin after fulfillment. Divide selling price by that margin, and you have your break-even ROAS. Compare actual ROAS to break-even ROAS, and you have a number that means something.
At the store level, use full margin after fulfillment for budget planning. At the product or campaign level, stop at margin after product cost if you cannot allocate per-order shipping to a single SKU. Pick the right level for the decision and stay consistent.
The Two-Week Pilot
If your blended ROAS looks healthy but cash flow does not, run this diagnostic. Pull your top five campaigns by spend. For each, calculate break-even ROAS using actual margin after fulfillment, not gross margin, not contribution margin from the finance deck, but the margin that accounts for product cost, shipping, and payment fees. Compare actual ROAS to break-even. Flag any campaign where the gap is less than 20%.
Then pull return rates by acquisition source. If paid acquisition has materially higher returns than organic, your reported conversion value is overstated. Adjust the break-even calculation to reflect net revenue after returns.
Two weeks of this analysis will tell you whether your 11x ROAS is a growth engine or a cash incinerator. The dashboard will not tell you. The P&L will.
