Return on Ad Spend (ROAS)

Return on ad spend is the revenue attributed to advertising divided by the cost of that advertising, usually expressed as a ratio or multiple. It is the dominant performance measure in digital advertising because it is simple to compute, available in every platform's reporting, and can be optimized toward automatically by bidding systems.

Its simplicity is also its principal weakness, since it measures revenue rather than profit. A campaign returning four times its spend is excellent for a business with high gross margin and poor for one selling at low margin, where the same ratio loses money on every sale. Businesses with meaningful returns, shipping costs, payment fees, or cost of goods need a margin-adjusted version of the measure to make sensible decisions, and organizations that manage to a raw revenue ratio frequently scale campaigns that are quietly unprofitable.

The second structural issue is that platform-reported figures are correlational and generous. Each platform counts conversions it can associate with its own exposure under its own attribution window and model, so the sum of platform-reported revenue routinely exceeds actual revenue by a wide margin. The figures also credit conversions that would have occurred without the advertising, which is most pronounced in retargeting and branded search, precisely the activity that reports the strongest performance. Comparing platform-reported ratios across channels is therefore comparing numbers produced by different, self-interested methodologies.

The third issue is that it says nothing about incremental volume. A high ratio is easy to achieve by restricting spend to the most responsive audiences, and a business optimizing purely for the ratio will systematically underinvest in growth, since expanding reach necessarily lowers average efficiency. The economically correct question is usually not how to maximize the ratio but what the marginal return on the next increment of spend is, and whether that marginal return still exceeds the required threshold. Most businesses could profitably accept a lower average ratio in exchange for materially more volume, and many do not, because the average is what gets reported.

Time horizon compounds all of these. Short attribution windows undercount considered purchases, while the measure entirely ignores repeat purchase and lifetime value, so a campaign acquiring loyal customers at a modest first-order return may be far more valuable than one generating one-off sales at a high one. Businesses with meaningful repeat rates need to evaluate on customer value rather than on first transaction revenue.

A more decision-useful framing for most businesses replaces the ratio with a contribution target expressed per acquired customer or per order. Setting a maximum acceptable cost per acquisition, derived from gross margin and the expected value of the customer relationship, gives buying teams a clear constraint that already accounts for profitability, and it removes the incentive to protect an average ratio by restricting reach. It also makes different channels directly comparable, since a cost per acquisition figure means the same thing everywhere while a revenue ratio depends on the average order value of whoever each channel happens to reach.

Used sensibly, the ratio remains a reasonable tactical signal for in-flight optimization within a channel, where relative movements are informative even if absolute levels are inflated. Strategic allocation, though, requires incrementality experiments and aggregate modelling to establish what advertising actually causes, which is normally the province of data analytics, with the resulting targets and thresholds set by growth management and executed through marketing services.