Marginal ROAS: Calculation, Purpose & Potential Pitfalls

Every seasoned chef knows there's a point at which adding more salt stops improving the dish. The first pinch transforms the flavor. The second refines it. But somewhere around the fifth, you've ruined dinner. The hard part isn't knowing that too much salt is bad. Even home chefs know that. But knowing exactly when you've crossed the line is harder to figure out. Marginal ROAS promises to solve that problem for your ad spend: add one more dollar, measure what comes back, and you'll know whether to keep going. It sounds precise. It sounds useful. And in theory, it is.

In practice, though, understanding marginal ROAS (return on ad spend) means grappling with some meaningful limitations before you build your marketing budget strategy around it. Knowing what the metric captures, where it falls short when it comes to diminishing returns assumptions, and what a more actionable alternative looks like can make a real difference in how you invest and scale your marketing campaigns across channels.

Key takeaways

• Marginal ROAS measures the incremental revenue generated by an additional dollar spent, which is different from average ROAS, which reflects total revenue divided by total spend
• While the concept is useful for understanding diminishing returns in theory, most real budget decisions don't happen one dollar at a time, which limits how directly actionable the metric is
• Marginal ROAS is only as reliable as the underlying ROAS figure it's built from; if platform-reported ROAS misattributes upper funnel impact to lower funnel campaigns, the marginal calculation inherits that error
• The assumption that returns always diminish as spend increases isn't universally true; many campaigns have more headroom than standard models suggest
• Scenario-based forecasting gives marketers a more practical answer to the same underlying question: if we shift budget across these specific campaigns, what does our projected return look like?
• Campaign-level measurement gives you the granularity needed to act on efficiency signals; channel-level averages tend to flatten the variation that actually matters for informed decisions

What is marginal ROAS?

Marginal ROAS is a way of measuring the return on ad spend at the margin, meaning what happens when you increase your advertising spend by one additional unit, typically one dollar.

The marginal ROAS formula is straightforward: divide the incremental revenue generated by the incremental ad spend used to generate it.

If you invest $1 more in a campaign and generate $4 more in revenue, your marginal ROAS at that point is 4. The goal of calculating marginal ROAS is to understand how returns change as spend increases, and whether you're still getting efficient incremental return on each additional dollar you put in.

The idea is that this tells you something about your marketing efforts that average ROAS can't.

How marginal ROAS differs from average ROAS

Average ROAS is a look back. You take total revenue attributed to a campaign or channel, divide it by total ad spend, and you have a summary of past performance. It's useful context, but it tells you what already happened across your entire investment, not what the next dollar will do.

Marginal ROAS tries to be forward-looking. A campaign can have a strong average ROAS while already past the point of efficient scaling across its channels. Conversely, a campaign with modest average performance might still have significant room to grow if it hasn't been pushed near its ceiling yet. Unlike average ROAS, the marginal version is meant to capture where you are on that curve right now, which is why (the theory goes) understanding marginal ROAS matters for any marketer trying to make informed decisions about where to invest next. That distinction is where marginal returns and budget allocation decisions start to intersect.

Why marginal ROAS is hard to use in practice

Understanding the concept is far simpler than using it to make real budget decisions. Marketing teams run into a few consistent friction points when they try to apply marginal ROAS to real planning work, and it's worth knowing what those are before leaning on the metric too heavily.

Budgets don't move one dollar at a time

The marginal ROAS framework is theoretically elegant but practically awkward. When marketing managers are deciding whether to increase a campaign's budget, they're not asking what a single additional dollar returns. They're asking whether putting an additional $10,000 or $20,000 into a specific campaign over the next 30 days is worth it. That's a scenario question. Answering it well requires forecasting across a realistic spend range, not calculating an efficiency rate at a single point on a curve. The math behind marginal ROAS was designed for a type of precision that most actual budget decisions don't call for.

It's only as good as the ROAS underneath it

This is the limitation that doesn't get discussed enough: marginal ROAS is a derivative metric. It's calculated on top of your existing ROAS figures, which means if those figures are off, your marginal calculations are off too.

Platform-reported ROAS is frequently distorted by marketing attribution models that favor lower funnel campaigns. When a brand runs an upper funnel awareness campaign on YouTube or Meta, that spend drives interest that often converts later through branded search, direct traffic, or organic channels. But last-click and platform attribution models assign the conversion credit to whichever touchpoint was closest to the purchase, typically a retargeting or branded search campaign. The upper funnel campaign that created the demand gets little or no credit.

The result is that lower funnel ROAS looks artificially strong and upper funnel ROAS looks artificially weak. When you then try to calculate marginal ROAS on either campaign type, you're doing precise math on imprecise inputs. Scaling decisions built on top of that will keep compounding the same misattribution.

Diminishing returns aren't as universal as the model assumes

Marginal ROAS is built on the assumption that marginal returns decline as ad spend increases. That's the standard model, and it's intuitive. Diminishing returns are a real phenomenon; the question is when they actually kick in. The actual relationship between ad spend and revenue varies by campaign, channel, creative, audience saturation, and season. Prescient's research into saturation assumptions in marketing mix modeling has found that many campaigns show linear or near-linear returns well beyond what conventional models predict, meaning brands sometimes hit artificial ceilings in their tools rather than real ones in the market. (This article will be updated with a link to the research paper once it's peer-reviewed and published.) Treating diminishing returns as a constant can lead to capping investment prematurely on campaigns that still have meaningful headroom, and the data often tells a different story when you look carefully.

A more useful question: What does a budget scenario return?

If marginal ROAS isn't the most practical frame for real decisions, what is? The same underlying question, just asked differently.

Thinking in scenarios instead of increments

Rather than estimating the return on the next marginal dollar, a more actionable question is: if we shift budget allocation across these specific campaigns by this amount, what does our projected return look like? That question produces an answer a team can act on. It accounts for where each campaign sits on its saturation curve, reflects realistic ad spend levels across channels rather than theoretical increments, and produces a projected ROAS tied to a plan. It also gives marketers a way to maximize investment in the campaigns most likely to drive profitable growth, rather than chasing a marginal rate in isolation. The data behind a well-built scenario forecast reflects how returns change across different spend levels, which is really what marginal ROAS is trying to surface in the first place.

Why campaign-level data matters here

Scenario analysis only works if your measurement is granular enough to support it. Channel-level ROAS averages flatten the variation between campaigns that are saturated and campaigns that still have room to grow. Getting to campaign-level efficiency data is what lets marketers identify where a budget shift will actually move the needle across channels, where they can maximize ad spend and incremental return, and where additional investment will hit a ceiling quickly. That granularity is also what makes profitable growth possible over time, because you're not chasing average performance across marketing channels; you're finding the specific campaigns worth scaling. It's also what makes it possible to give upper funnel campaigns fair credit for the data-supported role they play in driving demand, rather than letting lower funnel campaigns absorb revenue they didn't generate on their own.

Where Prescient comes in

Prescient's approach to budget optimization is built around scenarios, not marginal rates. Through our Optimizer, brands select the campaigns they want to include, set a desired spend amount, and get campaign-level allocation recommendations that predict an optimal revenue and ROAS outcome, without requiring a single dollar in budget increases. Each recommendation is paired with confidence scores so marketers can weigh the suggested allocation against their risk tolerance. Because Prescient measures at the campaign level rather than rolling everything up to the channel, those recommendations reflect real variation in efficiency rather than averages that obscure it. Marketers can maximize the incremental return on their ad spend by seeing exactly where the next dollar should go, not just what their overall channel average looks like.

Prescient also accounts for halo effects, the incremental revenue that upper funnel campaigns drive through organic search, branded search, direct traffic, and retail channels, so the ROAS figures feeding into any optimization recommendation are built on a more complete picture of what each campaign is actually doing across all your channels. If you're ready to see what scenario-based budget planning looks like with measurement you can trust, book a demo.

FAQs

What is marginal ROAS?

Marginal ROAS measures the return generated by an incremental increase in ad spend, or what your return would be for an additional dollar spent. It's calculated by dividing the additional revenue produced by the additional amount spent to produce it. Unlike average ROAS, which summarizes overall past performance, marginal ROAS is meant to reflect efficiency at the current spend level and help determine whether increasing spend is worthwhile.

What's the difference between marginal ROAS and average ROAS?

Average ROAS divides total attributed revenue by total ad spend and reflects how a campaign or channel has performed overall. Marginal ROAS focuses specifically on what additional spend returns at the margin. The distinction matters because a campaign with a strong average ROAS can still have declining returns on new investment, while a campaign with modest average performance might still be a good candidate for scaling if it hasn't reached its efficiency ceiling.

What is a good marginal ROAS?

There's no universal benchmark. A good marginal ROAS depends on your margins, your customer lifetime value, and your current business goals. At a minimum, you'd want marginal ROAS to exceed your break-even threshold, meaning additional spend is at least paying for itself. Beyond that, what qualifies as a strong marginal return varies by brand and by campaign type.

What does it mean when marginal ROAS is below 1?

A marginal ROAS below 1 means an additional dollar of ad spend is generating less than a dollar in revenue, so each incremental investment is losing money at the margin. This is generally a signal to stop scaling that campaign at its current spend level, though it's worth investigating whether the underlying ROAS figures are accurate before drawing conclusions about where to reallocate budget.