What is ROAS? A practitioner's guide
Written by Leadscale on October 02, 2025
Return on Ad Spend (ROAS) is revenue divided by ad spend. A 4.5x ROAS means £4.50 came back for every £1 in. Four versions matter, and they answer different questions. Gross ROAS compares channels. Net ROAS (gross × margin) tells you whether you’re making money. Break-even ROAS (1 ÷ margin) names the floor. LTV-adjusted ROAS extends the time horizon. In B2B, the same formula needs reading honestly: revenue arrives months after the campaign, not the same day.
What ROAS actually measures (and what it doesn't)
ROAS measures revenue attributable to an ad campaign, divided by the cost of that campaign. The output is a ratio (4.5x, 2.1x, 0.8x) or the same number expressed as a percentage (450%, 210%, 80%). The ratio is the more common reading in B2B because it makes the break-even line visible at a glance: anything below 1.0x means revenue did not cover spend.
The numerator is the revenue that can credibly be tied to the campaign. The denominator is the spend on the campaign, including media costs and any direct ad-platform fees. Production costs sit outside the standard reading unless the team has agreed otherwise in writing.
What ROAS does not measure: profit. The revenue side is gross revenue, not contribution margin. A 4.0x gross ROAS at a 20% margin returns 80p per £1 spent after the cost of delivering the product or service. The campaign is destroying contribution on every pound.
What ROAS also does not measure: incrementality. The number tells you what arrived against the spend, not what would have arrived without it. Some share of any campaign’s attributed revenue would have come anyway. ROAS does not separate that share out.
And what ROAS does not measure cleanly in B2B is timing. A campaign that runs in March can generate revenue in October. The standard month-on-month ROAS view fails on buying cycles that run several months from first touch to closed-won.
The four versions of ROAS
Each variant answers a different question. Three are arithmetic on the same two inputs; the fourth (LTV-adjusted) adds a time horizon. Gross ROAS is the version most teams default to and report up the chain. It is one of four readings worth knowing.
| Variant | Formula | When to use it |
|---|---|---|
| Gross ROAS | Revenue / Ad spend | Comparing channels. Spotting trends. Ignores margin. |
| Net ROAS | (Revenue × Gross margin) / Ad spend | The profitability lens. Needs an accurate margin figure. |
| Break-even ROAS | 1 / Gross margin | The floor below which net contribution from the campaign turns negative. |
| LTV-adjusted ROAS | (Revenue × Gross margin × LTV multiplier) / Ad spend | SaaS and subscription. B2B with measurable expansion. Only when the LTV multiplier is grounded in retention data. |
Gross ROAS is the version most teams report. It is the right reading for channel comparison and for spotting trend breaks, because it strips margin out and lets you compare two channels on a clean revenue-per-spend basis. It is the wrong reading for “are we making money”, because it pretends margin doesn’t exist.
Net ROAS is the reading finance wants. Multiply gross ROAS by the margin and the result is the contribution-per-pound-spent number. The same campaign at 4.0x gross looks healthy. At a 25% margin it returns 1.0x net, break-even, no contribution. At a 75% SaaS margin it returns 3.0x net, comfortably positive.
Break-even ROAS is the floor underneath both. One divided by margin. At a 25% margin, 4.0x. At a 50% margin, 2.0x. At a 75% margin, 1.33x. The “is this campaign making money?” question is answered against break-even, not against a benchmark.
LTV-adjusted ROAS is the long-cycle variant. Multiply revenue by the LTV multiplier and the margin, then divide by spend. It is the right reading for SaaS retention or B2B expansion. It is the wrong reading for one-off purchases. Where the LTV multiplier is an unsupported guess, it is also the version that produces the most flattering and least defensible number on a slide.
A worked example, in numbers
A B2B SaaS team spends £10,000 on a search campaign in Q2. The campaign generates £45,000 in first-year contracted revenue, attributed back through the company’s CRM. The gross margin on that revenue is 75% (the team checked with finance before reporting). Internal data on customers who converted on similar campaigns shows a 3.0x LTV multiplier over three years.
The four numbers:
- Gross ROAS: 45,000 / 10,000 = 4.5x. Every pound of spend returned four pounds fifty in first-year revenue.
- Net ROAS: (45,000 × 0.75) / 10,000 = 3.375x. After the 25% cost of delivery, every pound of spend returned £3.38 in contribution.
- Break-even ROAS: 1 / 0.75 = 1.33x. The campaign needed a 1.33x ROAS to break even. It comfortably cleared.
- LTV-adjusted ROAS: (45,000 × 3.0 × 0.75) / 10,000 = 10.125x. Across the three-year retention horizon, every pound of spend returned around £10.13 in contribution.
Four numbers, one campaign, four legitimate answers to four different questions. Which one the team reports up the chain depends on which question is being asked. The temptation is to lead with the 10.125x because it is the largest. The accurate report includes net ROAS alongside it, because the LTV figure only holds if retention holds.
What's a good ROAS?
The answer depends on gross margin. The 4:1 gross ROAS most-quoted in industry write-ups, including by BigCommerce, is a reasonable starting point at retail margin (around 25%), where break-even sits at 4.0x and “good” is comfortably above. It is the wrong benchmark for software, services, and most B2B, where margins are higher and the same 4.0x ratio represents a much healthier or much weaker position depending on the cost base.
Break-even ROAS at different margin levels:
| Gross margin | Break-even ROAS | “Comfortably above” |
|---|---|---|
| 20% | 5.0x | 7.0x and above |
| 25% | 4.0x | 6.0x and above |
| 50% | 2.0x | 3.0x and above |
| 75% | 1.33x | 2.0x and above |
| 85% | 1.18x | 1.8x and above |
The break-even column is arithmetic (1 ÷ margin). The “comfortably above” column is editorial guidance, not a sourced benchmark; read it as a rule-of-thumb for where a healthy campaign sits relative to its break-even line, not as a target your finance team will recognise.
The right “good ROAS” question is therefore: what is your gross margin, and how far above the resulting break-even line is the campaign sitting? At a 75% SaaS margin, a 3.0x ROAS is already healthy. At a 25% margin, the same 3.0x is below break-even and the campaign is destroying contribution on every pound spent.
The wider issue with the generic 4:1 benchmark is that it stayed in circulation long after the channel mix and margin mix it was calibrated against changed. eCommerce treats it as a near-universal target. B2B teams who imported it because it was the available number are reporting against a frame that was never meant for them. Replace it with your own break-even line and the conversation with finance changes shape.
What's different in B2B
Three things differ in B2B and change how the ROAS number should be read.
Revenue arrives later than the campaign. The eCommerce frame assumes revenue and spend show up in the same month. B2B buying cycles run several months on typical enterprise deals, often longer on complex ones. A campaign that ran in March can generate booked revenue in October.
The first reaction is to extend the reporting window; the second is to switch to a lagged comparison (March’s spend against October’s revenue). Both are real moves. The wrong move is to read June’s ROAS report as if it captured June’s spend. Most of June’s attributed revenue came from spend two or three quarters earlier.
The forecasting pain this causes shows up elsewhere too. The Salesforce State of Sales 7th Edition (2026) records the share of sales reps expecting to land in each quota-attainment band in 2025: 2% expect to hit 25% or less of quota, 12% expect 26 to 50%, 25% expect 51 to 75%, and 24% expect 76 to 99%. Sum the sub-100% bands and 63% of reps do not expect to make their number this year. A ROAS read against a missed quarter is reading against a moving target.
A short worked example shows the gap. A campaign runs in March at £20,000 spend. By June, attributed pipeline reads £180,000 and pipeline-weighted ROAS calculates to 9.0x. By October, closed-won revenue from the same campaign is £60,000 and closed-won ROAS calculates to 3.0x. The same campaign across two reporting windows produces two legitimate but different numbers, and reporting either at the wrong moment leads to the wrong decision.
The decision involves more than one person. B2B purchases run through a buying group rather than a single decision-maker, with each stakeholder bringing their own research and existing views to the decision. The campaign that converted the lead is rarely the only campaign the decision passed through, and the ROAS number rarely captures the share of the decision the campaign actually drove. The friction inside the buying group itself is real: 74% of B2B buying teams demonstrate unhealthy conflict during the decision process, according to a Gartner sales survey published in May 2025. That conflict adds noise the campaign-level ROAS reading does not see.
Pipeline is not revenue. Pipeline-weighted ROAS is useful for early reporting because it shows movement before closed-won revenue arrives. It is also dangerous because it does not survive contact with the close rate.
A campaign that generates £200,000 in opportunities at a 20% close rate has £40,000 in expected revenue, not £200,000. Pipeline-weighted ROAS shown to a CFO as if it were closed ROAS is the report that ends a marketing leader’s quarter.
The honest reading in B2B: name the variant being reported, name the window it covers, and name the close rate or LTV multiplier assumption baked into the number. The most useful thing the number does is tell you the lag and the noise are real. Reporting it as if they are not is what gets you the wrong decision.
ROAS vs other measurements
ROAS vs ROI. ROAS divides revenue by ad spend. ROI divides profit by total investment. The two answer different questions: ROAS measures whether a channel is doing its job, ROI measures whether the business is making money.
A campaign can run a 4.0x ROAS and a negative ROI if the cost base is heavy enough. A campaign can run a 1.5x ROAS and a positive ROI if the margin is high enough. Use both: ROAS at the campaign level, ROI at the programme or business level. The detail on the ROAS half sits in the formula walk-through.
ROAS vs CPA. ROAS divides revenue by spend (output over input). CPA (Cost Per Acquisition) divides spend by leads or customers (input over output, on a per-unit basis). CPA tells you what each lead or customer cost; ROAS tells you what each pound of spend returned.
They cover different parts of the same picture. A campaign with a low CPA but a low average deal value can still under-perform on ROAS. A campaign with a higher CPA but higher deal values can comfortably out-perform on ROAS. The ROAS calculation broken into components explains how the numerator and denominator interact under different cost-per-acquisition profiles.
ROAS vs MER. MER (Marketing Efficiency Ratio) divides total revenue by total marketing spend across the programme. ROAS is per-campaign or per-channel; MER is the rollup. MER is the right reading for the boardroom view; ROAS is the right reading for the channel-by-channel decision.
The two should agree on direction across a quarter. If they do not, the attribution layer needs work. The worked example earlier in this guide shows how the same campaign produces different ROAS readings across reporting windows, which is the most common practical reason MER and ROAS diverge inside a quarter.
The cluster: where to go from here
The ROAS calculator. The interactive page on this site that runs the four variants in one go for a given set of inputs (ad spend, attributed revenue, gross margin, optional LTV multiplier). The resulting URL is shareable, so a calculated answer can be sent to a finance counterpart without screenshots.
The formula, step by step. A longer-form walk-through of the calculation method, including edge cases and the maths shown explicitly for teams who want the working.
ROAS plus LTV. The long-cycle reading in detail, including how to derive a defensible LTV multiplier from retention data rather than guesswork.
ROAS benchmarks: how does your business compare? Sector-by-sector breakdowns for teams who need a sense of where their numbers sit against peers.
Three ways to improve your ROAS. Tactical moves for teams whose number is sitting below where it needs to be.
The bottom line
ROAS is one of several metrics a marketing leader needs to know cold. The four variants matter because each answers a different question. The break-even floor is the right answer to “what’s a good ROAS”; the 4:1 benchmark on its own is not. In B2B, the same formula reads honestly only when revenue lag, buying-group complexity, and the gap between pipeline and revenue are named in the report rather than absorbed into the number.
Frequently asked questions
ROAS, or Return on Ad Spend, is the revenue attributable to a campaign divided by the cost of that campaign, expressed as a ratio. Spend 10,000 and bring back 45,000, and your gross ROAS is 4.5x. The metric tells you whether a channel is generating revenue against its cost. It does not tell you whether you are making money; that is what net ROAS and break-even ROAS are for.
The answer depends on your gross margin, not on a generic benchmark. A 4:1 gross ROAS is the number most often quoted as “good”, but it is only good if your margin can support it. At a 25% margin, you need a 4.0x ROAS just to break even. At a 75% software margin, 1.33x breaks you even and 3.0x is already healthy. Compare against your own break-even ROAS, not against an industry rule of thumb.
Divide revenue attributable to the campaign by the cost of the campaign. The result is a ratio (4.5x) or a percentage (450%). The four variants are: gross ROAS (revenue divided by spend), net ROAS (gross multiplied by margin), break-even ROAS (one divided by margin), and LTV-adjusted ROAS (gross multiplied by margin multiplied by an LTV multiplier). See the calculator on this page for a working example.
No. ROAS divides revenue by ad spend. ROI divides profit by total investment. ROAS is a channel-efficiency metric used at the campaign level. ROI is a business-decision metric used at the programme level. Both are useful for different decisions. ROAS tells you whether a channel is doing its job. ROI tells you whether the business is making money.
The formula is the same, but three things differ in B2B and change how the number should be read. Revenue arrives months or quarters after the campaign rather than within days. The decision involves a buying group rather than a single decision-maker, and the share of that decision driven by any single campaign is rarely clean. And pipeline is not revenue: pipeline-weighted ROAS is useful for early reporting only, and should never be confused with closed-won ROAS.
Use a tracking setup that joins your ad-platform data, your CRM, and the revenue source of record your finance team trusts. The calculator on this page handles one-off scenarios. Ongoing tracking has to handle three harder problems: attribution (which ad gets credit for which revenue), identity resolution (matching the lead the platform saw to the customer the CRM holds), and the gap between marketing-attributed revenue and finance-recognised revenue. There is no shortcut.
