ROAS Meaning: What the Number Measures—and What It Leaves Out
ROAS, or return on ad spend, measures the conversion value attributed to advertising for each unit of ad cost. A ROAS of 5.0 means the measurement system credited five dollars of conversion value for every dollar counted as ad spend. The same result may appear as 500%.
That answer is useful, but it contains two traps. “Conversion value” may mean sales revenue, estimated margin, or a value assigned to a lead. “Ad spend” may include only media charges, leaving creative work, agency fees, sales labor, and fulfillment elsewhere. The ratio is easy to calculate. Its meaning depends on what went into it.
Google Ads defines ROAS as total conversion value divided by total spend. That makes ROAS a performance ratio under a particular tracking setup—not a profit figure, a cash-flow measure, or proof that the ads caused every conversion they received credit for.
How to calculate and read ROAS
The basic formula is:
ROAS = attributed conversion value ÷ advertising cost
To express it as a percentage:
ROAS percentage = attributed conversion value ÷ advertising cost × 100%
Suppose a campaign records $50,000 in attributed sales and $10,000 in media spend. Its ROAS is $50,000 ÷ $10,000 = 5.0, or 500%. This is an illustrative calculation, not a benchmark. The defensible reading is: “The configured measurement system attributed $5 in sales to the campaign for each $1 of media spend.”
It would be wrong to say that the campaign made $5 of profit per dollar spent. Nothing in the calculation has yet deducted the cost of producing or delivering what was sold.
Read the ratio and percentage as the same result
A dashboard can display ROAS as a multiple, a percentage, or a currency-per-currency ratio. These expressions are equivalent:
- 5.0 ROAS
- 5:1 ROAS
- 500% ROAS
- $5 in attributed value per $1 of ad spend
The percentage format causes avoidable confusion because 500% can sound like a 500% profit margin. It is not. It is simply the numerator divided by the denominator, multiplied by 100. A 100% ROAS is 1.0: the campaign recorded one unit of conversion value for each unit of ad cost. Whether that is commercially acceptable depends on the margin and costs behind the conversion value.
The time window matters too. A weekly report may count spend immediately while some conversions arrive days later. Google advises advertisers evaluating Target ROAS to exclude the most recent conversion-delay period, because incomplete conversion data can make recent performance look weaker than it eventually will. Extending the window can also introduce repeat purchases or later conversions, so two ROAS figures are comparable only when their date and attribution rules match.
Identify what the numerator and denominator contain
ROAS does not require the numerator to be booked revenue. Google’s guidance allows advertisers to assign static or transaction-specific values to conversions and describes value-based bidding around measures such as sales revenue or profit margins. A purchase might carry its order value, while a demo request might carry a modeled value based on the likelihood and economics of a later sale.
This creates a sharp distinction between two reports that both say “400% ROAS.” One might mean $4 of attributed sales for each $1 of media spend. The other might mean four assigned lead-value points per dollar. The arithmetic matches; the business meaning does not.
Before comparing campaigns, write down the metric contract in one sentence: the eligible conversion, its value basis, the cost included, the attribution model, the lookback window, the currency, and the treatment of cancellations or other adjustments. For the denominator, “ad spend” ordinarily refers to the advertising cost included by the reporting system. If a decision also depends on production, agency, technology, or sales costs, add them in a separate economic view instead of quietly calling a broader cost base platform ROAS.
Do not compare two ROAS figures until you know that their conversion value, cost scope, attribution rules, and time window describe the same thing.
Why a strong ROAS can still support a bad decision
The numerator in a revenue-based ROAS sits near the top of the economics. Profit sits much lower. The SEC’s guide to financial statements explains that gross profit is net revenue after cost of sales, while operating expenses are deducted later. A standard revenue ROAS makes none of those deductions except the ad spend in its denominator.
Return to the illustrative campaign with $50,000 in attributed sales and $10,000 in ad spend. If those sales carry a 30% gross margin before advertising, they produce $15,000 of gross profit before ad cost. Subtract the $10,000 in advertising, and $5,000 remains before any other acquisition or operating costs. The dashboard still shows 5.0 ROAS. The economic cushion is far smaller than the headline ratio suggests.
That is why “good ROAS” has no universal value. For revenue-based ROAS, an approximate starting threshold can be derived from the gross margin before advertising:
Break-even revenue ROAS before other costs = 1 ÷ gross margin rate
At a 30% gross margin, that produces 1 ÷ 0.30 = 3.33, or 333%. At a 70% gross margin, it produces about 1.43, or 143%. These are simplified break-even points before agency fees, payment costs, sales commissions, support, overhead, required profit, refunds, or the cost of waiting for cash. They are not campaign targets.
A finance-approved target therefore needs more than the platform ratio. It needs the actual value basis, the costs relevant to the decision, and the return the business requires. If the numerator already represents contribution rather than revenue, the threshold must be derived from that different numerator; reapplying the revenue-margin formula would count margin twice.
ROAS also answers a different question from several metrics that often appear beside it:
| Metric | Numerator or outcome | Denominator | What it can tell you | What it cannot settle alone |
|---|---|---|---|---|
| Attributed ROAS | Conversion value credited under tracking rules | Ad spend | Value credited per ad-cost unit | Profit, cash timing, or causal lift |
| ROI | Profit or gain under a declared scope | Investment under that scope | Return relative to the investment counted | Which ads received conversion credit |
| Payback | Cumulative realized contribution over time | Acquisition investment to recover | When the investment is recovered | How attribution assigned the original sale |
| Incremental ROAS | Estimated additional conversion value caused by the intervention | Ad spend | Causal value per ad-cost unit under the study design | Full profit after uncounted costs |
The difference between attributed and incremental ROAS is especially important when a campaign reaches people who were already likely to buy. Google Ads attribution models determine how conversion credit is distributed across eligible ad interactions. Changing the model can move credit among campaigns or keywords even when the underlying customer purchases do not change.
Incrementality asks a harder question: what happened because the advertising ran? In Google’s user-based Conversion Lift measurement, conversions from a treatment group that saw ads are compared with conversions from a control group held back from seeing them; incremental ROAS divides incremental conversion value by total ad spend. The result is an estimate tied to that experiment and its uncertainty. Changing an attribution window is not a substitute for a counterfactual.
Cash timing creates another separation. A B2B campaign can record a high value for a signed contract while invoices will be collected over months. An ecommerce campaign can record the order before the return period closes. ROAS can be correct under the reporting rules and still say nothing about when cumulative realized contribution repays the acquisition outlay. A cohort view that follows spend, collections, delivery costs, refunds, and churn over time is the appropriate bridge to payback.
The practical reading process has four layers:
- Validate the reported ratio. Recalculate attributed conversion value divided by ad spend for the same period, currency, and campaign scope. A mismatch often points to filters, delayed conversions, or different included actions.
- Translate value into economics. Reconcile the attributed value with order, billing, or CRM records, then apply the margin and cost rules relevant to the decision. The result may be contribution, but it is not automatically net profit.
- Add the recovery timeline. Follow the acquired cohort until cash and realized contribution recover the acquisition cost. This prevents predicted or contracted value from masquerading as immediate cash.
- Match evidence to the decision. Attributed ROAS can guide reporting and platform optimization under stable measurement rules. A budget question about what additional spend will cause may require an experiment or another credible causal design.
This layered reading preserves what ROAS does well. The metric can reveal which campaigns receive more value per ad dollar under a consistent setup, and it can support value-based bidding. It simply cannot carry the whole investment case by itself.
The next number to put beside ROAS
If a budget review has room for only one economic companion to ROAS, use contribution after ad cost under a clearly stated cost scope. It forces the discussion below revenue without pretending that one dashboard ratio has become a full profit statement. For businesses with long sales or collection cycles, add cohort payback next.
The judgment should follow the decision: use attributed ROAS to understand credited performance, contribution to test whether the unit economics work, payback to see when the money returns, and incrementality when the question is whether the advertising created additional value. Asking one ratio to do all four jobs is how a precise calculation produces a weak decision.
Frequently asked questions
What is Target ROAS?
Target ROAS is a value-based bidding setting in which an ad platform adjusts bids while trying to achieve an average conversion-value-to-cost ratio. Google notes that individual conversions can fall above or below the target and that setting the target too high may limit traffic. A 500% target asks the system to pursue about $5 in configured conversion value per $1 of ad spend; it does not guarantee that result or establish that 500% is profitable.
How often should ROAS be checked?
Match the review cadence to the conversion delay and the volume of data rather than assuming daily data is final. A fast ecommerce purchase cycle may support frequent operational checks, while a B2B pipeline with delayed offline conversions needs a longer settled window. Mark the newest incomplete period separately, and compare mature cohorts or equivalent date windows before changing budgets.
Can ROAS from different ad platforms be added together?
Adding platform-reported conversion values can double-count one customer when multiple platforms claim credit for the same journey. A blended ROAS is usable only after spend is combined and conversion value is deduplicated against an authoritative order, billing, or CRM record under one attribution rule. Summing the individual ROAS percentages is mathematically wrong; calculate total deduplicated value divided by total included spend.
Is ROAS the same as CPA?
CPA divides ad cost by the number of attributed actions, while ROAS divides attributed value by ad cost. CPA treats two conversions as equal unless they are separated into different actions; ROAS can distinguish them when they carry different values. When every conversion has the same trustworthy value, either metric may support optimization, but neither supplies margin or causal evidence on its own.
Should refunds reduce ROAS?
For an economic assessment, refunded or cancelled sales should not remain as realized value. Whether a platform report reflects them depends on its conversion-adjustment setup and reporting timing. Keep the dashboard figure labeled as reported, then maintain an adjusted view that reverses cancelled value consistently; otherwise campaigns with different return rates can appear comparable when their retained economics are not.