ROI vs. ROAS: Which Metric Should You Use?

ROI vs. ROAS is one of the most common sources of confusion in marketing finance, because both measure return and both are expressed relative to spend. ROAS, return on ad spend, divides the revenue attributed to advertising by the amount spent on it. ROI, return on investment, divides the profit earned by the amount invested. The difference is that ROAS ignores what it cost to deliver the revenue, and so a campaign can show a ROAS of 4.0 and still lose money. This guide defines both metrics, shows with a worked example how the same ROAS can mean a profit or a loss depending on margin, calculates break-even ROAS, explains when each metric is appropriate, and covers incrementality and the mistakes that lead teams astray. It is part of business analytics for finance and operations, follows how to calculate ROI for a business project and leads to marketing ROI and the ROI of a software purchase. Marketing budgets are set in budgeting and forecasting.

Definitions

ROAS = revenue attributed to advertising / advertising spend. A ROAS of 4.0 (or 400 percent) means $4 of revenue for each $1 of spend. It is a gross measure, easily read from ad platforms.

ROI = (profit attributable to the investment - cost of the investment) / cost of the investment. For advertising, "profit attributable" means the contribution margin on the sales the advertising caused, before deducting the ad spend. ROI is a net measure and requires data on margins, costs and attribution.

A worked example

A brand spends $20,000 on a campaign and the ad platform attributes $80,000 of revenue to it (hypothetical). The ROAS is 80,000 / 20,000 = 4.0. Whether that is good depends on the margin. If the contribution margin on the products sold, after product cost, shipping, payment fees and returns, is 40 percent, the contribution is 0.40 × 80,000 = $32,000. Subtracting the ad spend gives a profit of $12,000, and the ROI is 12,000 / 20,000 = 60 percent. If the contribution margin is only 25 percent, the contribution is $20,000, exactly equal to the spend, and the ROI is zero. At 20 percent, the contribution is $16,000, and the campaign loses $4,000, an ROI of -20 percent.

The same ROAS of 4.0 yields a loss, a break-even and two different profits. A team that celebrates a ROAS of 4.0 without knowing its margin is guessing.

Break-even ROAS

Break-even ROAS is the ROAS at which contribution equals ad spend: 1 / contribution margin. At a 40 percent margin it is 1 / 0.40 = 2.5. At 25 percent it is 1 / 0.25 = 4.0. At 60 percent it is 1 / 0.60 = 1.67. Any ROAS above the break-even point earns a profit on the first purchase, and any ROAS below it loses money unless later purchases make up the difference. Give campaign managers a target ROAS by product line, calculated from margin, instead of a single number for the business. Mixed margins matter: a campaign that sells a 60 percent margin product and a 20 percent one has a blended break-even between the two, which depends on the mix.

When to use ROAS

• Campaign optimization. Comparing ads, audiences and bids within a platform, where the margin is similar.
• Daily monitoring. ROAS is available quickly and is a reasonable early signal.
• Setting bidding targets, using break-even ROAS adjusted for the desired profit.

When to use ROI

• Budget decisions between marketing and other investments.
• Reporting to the CFO or board, where profit matters more than revenue.
• Comparing channels with different cost structures, such as paid media and a creator program with fees and product costs.
• Evaluating campaigns that include non-media costs, such as agency fees and production.

Allowing for lifetime value

For a business where customers buy repeatedly, first-purchase ROI understates the return. Suppose a campaign acquires customers at a cost of $50 each, with a first-order contribution of $30. The first-purchase ROI is (30 - 50) / 50 = -40 percent. If the average customer places two more orders over the next year, each worth $30 of contribution, the lifetime contribution is $90, and the ROI becomes (90 - 50) / 50 = 80 percent. Make this adjustment only with evidence of repeat behavior, using cohort data, and discount for the time delay. See customer acquisition cost.

Attribution and incrementality

Both ROAS and ROI depend on how much revenue is credited to the advertising. Platforms tend to claim credit for sales that would have happened anyway, such as purchases by existing customers who clicked a brand ad. The honest question is incremental: how much revenue occurred because of the ads, compared with a world without them? Holdout tests, geographic experiments and conversion lift studies measure this. If a test shows that only half the attributed revenue is incremental, the true ROAS is half the reported figure. In the example above, a platform ROAS of 4.0 with 50 percent incrementality is a true ROAS of 2.0, which is below break-even at a 40 percent margin (2.5). See incremental ROI and marketing attribution models.

Common mistakes

1. Treating ROAS as profit.
2. Using one target ROAS for products with different margins.
3. Ignoring non-media costs in ROI.
4. Trusting platform attribution without any incrementality check.
5. Comparing ROAS across platforms that attribute differently, such as click-based and view-based.
6. Ignoring repeat purchases when they are material, or assuming them without data.
7. Optimizing ROAS to the point of shrinking the business. Pushing for the highest ROAS can mean spending only on the cheapest, most loyal customers, and ceasing to grow.

Prospecting versus retargeting: why blended ROAS hides problems

A single blended ROAS can conceal a mix of very good and very bad spending. Suppose a brand spends $20,000, split between two campaigns (hypothetical). Prospecting, aimed at new audiences, takes $12,000 and is credited with $24,000 of revenue, a ROAS of 2.0. Retargeting, aimed at people who have already visited the site, takes $8,000 and is credited with $48,000, a ROAS of 6.0. The blended result is $72,000 / $20,000 = 3.6. With a 40 percent contribution margin, break-even ROAS is 2.5. Prospecting contributes 0.40 × $24,000 = $9,600 against spend of $12,000, a loss of $2,400. Retargeting contributes 0.40 × $48,000 = $19,200 against $8,000, a profit of $11,200. Together, the profit is $11,200 - $2,400 = $8,800, an ROI of 8,800 / 20,000 = 44 percent.

A naive reading says to cut prospecting, which has a ROAS below break-even, and move the money to retargeting. But retargeting can only reach people who have visited, and prospecting is what brings them. Cutting prospecting shrinks the audience that retargeting depends on, and the retargeting ROAS will fall as the pool dries up. Retargeting also tends to claim credit for sales that would have occurred anyway. The right response is to evaluate the two campaigns as a system, using a holdout test on retargeting and a longer view of prospecting, which may bring customers who buy again.

Setting a target ROAS from a desired return

If you want a particular ROI, you can work backward to a target ROAS. The formula is: target ROAS = (1 + desired ROI) / contribution margin. For a 40 percent margin and a desired ROI of 20 percent, the target is 1.20 / 0.40 = 3.0. Check: at a ROAS of 3.0, each $1 of spend brings $3 of revenue and $1.20 of contribution, a profit of $0.20 per $1, or 20 percent. At a 25 percent margin, the same 20 percent ROI would need a ROAS of 1.20 / 0.25 = 4.8. Giving campaign managers a target ROAS by product line, derived this way, aligns day-to-day bidding decisions with the profit goal.

Blended measures at company level

Platform ROAS is only one view. Many finance teams also track the marketing efficiency ratio (MER): total revenue divided by total marketing spend across all channels, without attribution. If total revenue is $1,500,000 and marketing spend is $300,000, MER is 5.0. MER avoids the attribution argument, because it doesn't try to assign revenue to channels, but it can't tell you which channel works, and it moves with seasonality, pricing and organic demand. Use it as a trend check next to the channel-level ROAS and ROI: if platform ROAS is stable and MER is falling, the platforms may be claiming more of the same sales.

Practical guidance

1. Agree on one definition of attributed revenue (window, click or view-based) and use it consistently.
2. Compute contribution margin by product line and keep it current.
3. Publish break-even and target ROAS by line.
4. Report ROI at least quarterly, including non-media cost.
5. Run incrementality tests on the largest or least certain channels at least once a year.
6. Review the relationship between ROAS and scale: ROAS usually falls as spend rises, so the average ROAS overstates the return on the next dollar. See incremental ROI.

FAQs

What is the difference between ROI and ROAS?

ROAS divides attributed revenue by ad spend. ROI divides net profit by cost, so it accounts for product margin and other costs.

What is a good ROAS?

One that exceeds break-even ROAS for your margin, which is 1 divided by the contribution margin, plus enough to cover other costs and the profit you want.

Which should I report to the CFO?

ROI or contribution after marketing cost, because it shows profit. ROAS can accompany it as an operating metric.

How do I calculate break-even ROAS?

Divide 1 by your contribution margin. A 40 percent margin gives a break-even ROAS of 2.5.

Does ROAS include repeat purchases?

Only if the attribution window captures them. Platform ROAS usually reflects a short window, so lifetime value needs a separate analysis.