11x ROAS, Yet Losing Money Per Order

High ROAS metrics can mask severe cash burn, creating a dangerous illusion of profitability for ecommerce businesses. An 11x return on ad spend (ROAS) appears to be a monumental victory, often cited by founders to investors and displayed prominently in agency reports. However, this single metric can obscure a reality where the business is quietly destroying capital with every transaction. The discrepancy lies not in the calculation itself, but in the context surrounding it. While “it depends” is a cliché, it is also the most accurate assessment of whether high ROAS signals health or impending financial collapse.

The Illusion of Profitability

Consider an account that reported an 11x blended ROAS while simultaneously losing money on every single order. This was not a result of fraudulent activity or bad luck; it was a structural failure of how performance was measured. The previous agency presented these figures as proof of success, yet the finance director within the company knew the truth: the contribution margin per order was deeply negative.

The numbers presented here are composite data based on real-world patterns, adjusted only to protect client anonymity. By breaking down the math, we can see exactly how a positive ROAS figure coexists with negative profit. For a typical £100 apparel order, the reported conversion value includes VAT and is recorded before returns are processed. This is standard behavior for most Shopify and GA4 setups, which do not automatically adjust for post-sale reversals.

• Reported conversion value: £100.00Despite the dashboard showing a massive return, the company loses £0.39 on every £100 order. The marketing dashboard and the P&L describe the same transactions, but only one reflects the actual legal and financial obligations of the business. A 28% return rate is normal for fashion, and a 63% COGS rate reflects heavy promotional trading. These conditions peak during sales periods when discounted stock converts well, making the account look its best financially while operating at its worst marginally.

The Blended Metric Trap

The danger of the 11x figure is compounded by blending. In this case, brand campaigns were achieving roughly 18x ROAS, absorbing the majority of the budget. Meanwhile, non-brand campaigns sat around 3x. The headline 11x was simply a weighted average of demand the brand already owned and a modest amount of incremental work.

Brand searches represent users who have already decided to purchase. They would likely convert through organic or direct channels regardless of paid intervention. Therefore, the true incremental return was significantly lower than 11x. The account was taking credit for sales it did not cause, at margins it did not generate. This is the default state of accounts optimized solely for platform-reported revenue. The reporting is not lying; it is answering a narrow technical question while stakeholders treat it as the answer to a broad commercial one.

The Hidden Cost of Efficiency

A common correction is to shift from revenue-based bidding to profit-based bidding (POAS), calculating values net of VAT, returns, COGS, and fulfillment. While this aligns ads with contribution, there is a critical risk: optimizing too tightly for efficiency can strangle volume.

When you set a tight efficiency target, Smart Bidding buys only the cheapest, most certain conversions. Volume drops. For retailers holding physical stock, this is catastrophic. Stock is a depreciating asset. A fashion brand pays for inventory months before selling it. Every week a unit sits unsold, its recovery value falls, and markdowns deepen. If an account tuned for maximum efficiency sells fewer units, working capital is converted into boxes sitting in a warehouse.

Consider a SKU with eight weeks left in season: 1,000 units at an £18 cost and £45 RRP.

• Tight Efficiency Target: Sells 350 units at a strong margin. Leaves 650 units to be cleared at 70% off, mostly below cost.
• Relaxed Target (4x ROAS): Sells 850 units at a lower per-unit margin. Leaves just 150 units to clear.

The second path produces a worse ROAS but generates more total contribution and brings cash back into the business sooner. The “efficient” setting was actually the expensive one because it failed to liquidate inventory effectively. The account in our example lost money twice: on every order made, due to poor margins, and on every order missed, due to overly restrictive bidding.

Strategic Realignment

To fix this, businesses must implement four key changes to their targeting and management strategies.

First, rebuild conversion value as contribution. Send data net of VAT, expected returns, COGS, and fulfillment to Google. Until the algorithm sees margin, targets are being negotiated in a fictional currency.

Second, separate brand from non-brand analysis. Brand harvest rates tell you nothing about acquisition efficiency. Grade non-brand campaigns on their own numbers, ideally focusing on new-customer contribution.

Third, assign one job per SKU. Each product should be run for profit, volume, or cash recovery before the season ends. It cannot do two jobs at once because the bidding strategies conflict. Reassigning each SKU’s objective weekly against the P&L forces alignment between marketing and finance teams.

Fourth, decide the cash question explicitly. For stock-holding retailers, target setting is a weekly commercial decision regarding margin, sell-through, and cash recovery. Sometimes a less profitable sale is the smarter move if it frees up capital. This decision belongs to the business leadership, not to a static tROAS target set months ago.

When we restructured similar accounts, blended ROAS fell. Separating brand, rebuilding values, and setting volume targets against sell-through made the reporting look uglier. However, the business became healthier. Contribution per order turned positive, and cash flow improved as revenue arrived during the season rather than during clearance. If your account posts a number that looks impressive, verify what it measures. A high ROAS on the wrong catalog is often just a list of opportunities the account declined to pursue.