What Is Return on Ad Spend?
Return on Ad Spend, abbreviated ROAS, measures the revenue earned for every dollar spent on advertising. The math is simple. Divide ad revenue by ad spend. A campaign that earned 30,000 dollars from 10,000 dollars in spend has a 3.0 ROAS, sometimes written as 300%. ROAS is the cleanest single number for evaluating paid media performance and the metric most paid programs report on weekly because the calculation is fast and the comparison across campaigns is straightforward.
It is also one of the most misused metrics in marketing. ROAS based on first order revenue alone misses lifetime value. ROAS based on last click attribution underrates upper funnel work that closed nothing directly but helped close everything indirectly. The number itself is fine. What teams do with it tends to be where things go wrong, because ROAS is sensitive enough to the underlying assumptions that the same campaign can look like a winner under one calculation and a loser under another.
Why Is ROAS Different From ROI?
Because ROAS uses revenue and ROI uses profit. A campaign with a 4.0 ROAS sounds great until you remember the gross margin is 25%, which means the campaign barely broke even after cost of goods. ROAS is a top line ratio. ROI is a bottom line ratio. Both are useful, but they answer different questions and should not be used interchangeably. Most marketing reports that confuse the two end up celebrating campaigns that are quietly losing money or pausing campaigns that are quietly making it.
The right ROAS target depends on margin, fixed costs, and customer lifetime value. A subscription business with high LTV can run a 1.5 ROAS profitably because the second purchase is where the money is made. An ecommerce store on tight margins might need a 4 to 5 ROAS to actually pay back. The brand that sets a single ROAS target across every campaign without accounting for margin and LTV makes worse budget decisions than the brand that thinks through the math per channel.
What ROAS Should You Target by Channel?
Direct response ecommerce on paid social and paid search typically needs 3 to 5 ROAS to clear margins and operating costs. Subscription businesses and SaaS often run 1 to 2 ROAS on first order because the LTV math justifies it. Brand and awareness campaigns will look weak by ROAS because they were never designed to close the sale on their own. Use brand lift studies, branded search volume, and assisted conversions instead. Retargeting routinely produces 5 to 15 ROAS, but the audience is already warm so the high number is partly recapturing existing demand rather than creating new demand.
The right ROAS target is your number, calibrated to your category, your margin, your LTV, and your funnel structure. Generic benchmarks are useful as sanity checks. Your own historical data is more useful for actually setting targets. The brands that hit healthy ROAS reliably usually run their math monthly and adjust targets as conditions shift, rather than setting one target at the start of the year and defending it regardless of what the data is telling them.
What Are the Common Mistakes Teams Make With ROAS?
The most common is reporting ROAS on first order revenue when LTV justifies a different math. A 1.5 ROAS on first order looks bad until the cohort data shows that customers from that channel return three times in the first year. The second mistake is mixing attribution models within a single ROAS report. Calculating ROAS for paid social under last click and ROAS for SEO under first click then comparing them makes the data noise rather than signal. Pick one attribution model and apply it consistently across every channel comparison.
The third mistake is judging awareness campaigns by ROAS. Awareness exists to make later channels convert better. Direct ROAS on awareness ads will always look weak because they rarely close the sale on their own. Brands that defund awareness based on ROAS reports usually watch their other channels degrade six months later and never quite figure out why. The fourth mistake is celebrating high ROAS on retargeting without accounting for cannibalization. Some retargeting conversions would have happened anyway. Incremental lift studies, when feasible, separate the conversions retargeting actually drove from the ones it just claimed credit for.
How Do You Use ROAS Honestly in Practice?
Calculate ROAS using net revenue, not gross. Subtract returns, refunds, discounts, and chargebacks before measuring. Match the time window to the buying cycle. Long sales cycles need longer windows. Factor in lifetime value, not just first order revenue, especially for subscription and considered purchase categories. Use a single attribution model across every channel comparison, and disclose the model used in every ROAS report. Pair ROAS with cost per acquisition, conversion rate, and lifetime value to get a complete picture of unit economics.
For a real world look at the math behind one of the most common paid platforms, read are Google Ads worth it. We track ROAS by channel, campaign, and customer cohort inside Google Ads Management and Analytics, with the integrated paid program running inside our Growth and Acquisition solution. For related concepts, see ROI, CPA, Customer Lifetime Value, LTV to CAC Ratio, and Attribution Model. The bottom line: ROAS is a useful number when calculated honestly and a misleading one when it is not. Use it well and budget decisions get sharper. Use it badly and you defund the wrong things.