What is Marginal ROAS and Why Does It Matter More Than Average ROAS?

Average ROAS tells you how past spend performed on average. Marginal ROAS (mROAS) tells you what the next pound of spend will earn. For budget decisions, you need marginal ROAS. Average ROAS can be high even when you are wasting money.

Your paid search campaign reports a 4x ROAS. Your board is happy. But what does that number actually tell you? It says that across all the spend you put in, you got four times back. It does not tell you whether adding another 10,000 to that campaign will earn 40,000 or 8,000. That is what marginal ROAS tells you, and it changes every budget decision you make.

The difference between average and marginal ROAS

Average ROAS divides total revenue by total spend across a period. If you spent 100,000 and generated 400,000 in revenue, your average ROAS is 4x. Marginal ROAS asks a different question: what happens to revenue if you increase spend by one more unit? It measures the return on the incremental pound, not the whole pound.

Because of diminishing returns, marginal ROAS is almost always lower than average ROAS. The early spend in any channel reaches the most receptive customers at the lowest cost. Later spend reaches progressively less-interested audiences at higher cost. If your average ROAS is 4x, your marginal ROAS might be 1.5x or even below 1x.

Why average ROAS leads to bad decisions

A team running on average ROAS will keep pouring budget into a channel as long as the headline number looks good. They see 4x ROAS on the dashboard and approve more budget. But if the marginal ROAS on that extra budget is 0.8x, they are actively destroying value. Every extra 10,000 they spend returns only 8,000.

• Average ROAS hides the performance of your most recent, most expensive spend.
• It does not account for the fact that your early, cheap impressions drove most of the revenue.
• It rewards channels that have been running longest, not channels that are most efficient now.
• It gives no signal about where the saturation point is.
• It makes scaling decisions feel safer than they are.

How to calculate marginal ROAS

The practical way to estimate marginal ROAS is through a Marketing Mix Model. A well-built model produces a response curve for each channel, showing how revenue changes as spend increases. The slope of that curve at your current spend level is your marginal ROAS. Steeper slope means higher mROAS. A flat curve means you are close to saturation.

Without a model, you can approximate it with a simple experiment. Increase spend in a channel by 20 percent for four weeks and measure the change in revenue (ideally against a holdout region). Divide the incremental revenue by the incremental spend. That ratio is your marginal ROAS for that spend level.

A channel with a 6x average ROAS but a 0.9x marginal ROAS should receive less budget, not more. A channel with a 2x average ROAS but a 3x marginal ROAS should receive more. Average ROAS gets this exactly backwards.

Using marginal ROAS to set budgets

To optimise your budget using mROAS, rank your channels by current marginal return. Shift spend from the lowest mROAS channel to the highest until the marginal returns equalise across all channels. At that point, a pound moved from any channel to any other channel earns the same return. You have found the efficient frontier for your current total budget.

Talking to your team about marginal ROAS

Channel managers will push back because their headline ROAS looks great. Frame it this way: we want to know what the next pound earns, not what past pounds averaged. The question is not how well this channel performed last quarter. The question is what happens if we add 10,000 today. The answer to that question should drive the decision.