Quick answer: IROAS (Incremental Return on Ad Spend) is a marketing metric that measures the extra revenue an ad campaign actually causes, beyond sales that would have happened without the ads, for every rupee spent. It is calculated as incremental revenue divided by ad spend.
Regular ROAS gives a campaign credit for every sale it touched. iROAS only gives credit for the sales the campaign created. The difference is every customer who was going to buy anyway: people who already knew your brand, searched your name, or were one click from checkout when a retargeting ad caught them. To find iROAS, you compare a group that saw your ads against a similar group that did not. If the exposed group spends ₹12 lakh and the unexposed group spends ₹7 lakh, only ₹5 lakh is incremental, and that is the number your iROAS is built on.
Why iROAS matters
Every ad platform grades its own homework. Google, Meta and the rest report ROAS using their own attribution rules, and those rules count a conversion as long as someone interacted with an ad inside the attribution window, whether or not the ad changed the decision. That is how an account can show a healthy 5x ROAS while total business revenue barely moves when spend goes up.
The platforms themselves now acknowledge the gap. Google’s Conversion Lift reports include incremental ROAS as a standard metric, defined as incremental conversion value divided by total ad spend (Google Ads Help). Meta offers an incremental attribution setting that uses machine learning models to predict whether a conversion was caused by an ad (Social Media Today).
Even the platforms’ own incrementality estimates need checking. When Seer Interactive tested Meta’s incremental attribution on about US$1.05 million of ad spend in April 2025, Meta reported that 87% of conversions were incremental, while a GA4 cross-check put the figure at 67%, a 20-point gap (Seer Interactive). iROAS is the only return-on-ad-spend figure that answers the question a CFO actually asks: how much revenue would we lose if we switched this campaign off?
That is why Uginix treats platform ROAS as a starting estimate and iROAS as the number that decides budget. Scaling a campaign on a flattering ROAS that is mostly organic demand is the fastest way to spend more without earning more.
iROAS vs ROAS vs MER
These three numbers answer three different questions. Use all three, but know which one you are looking at.
| Attribute | ROAS | iROAS | MER |
|---|---|---|---|
| Full name | Return on Ad Spend | Incremental Return on Ad Spend | Marketing Efficiency Ratio |
| Formula | Attributed revenue / Ad spend | Incremental revenue / Ad spend | Total revenue / Total marketing spend |
| Question it answers | How much revenue did the platform credit to this campaign? | How much revenue did this campaign cause? | Is marketing as a whole paying its way? |
| Counts sales that would have happened anyway? | Yes | No | Yes (but only once, at business level) |
| How you get it | Read it in Ads Manager or Google Ads | Run a lift or holdout test | Divide revenue from your own books by total spend |
| Best for | Daily optimisation and bid decisions | Budget allocation between channels and campaigns | Monthly business health check |
In almost every account, iROAS is lower than platform-reported ROAS, because platform ROAS includes customers who would have bought without seeing the ad. The size of that gap varies a lot by campaign type, which is exactly why it is worth measuring rather than assuming.
How to calculate iROAS
The formula is simple. Getting a clean “incremental revenue” number is the hard part.
iROAS = (Revenue from exposed group – Revenue from control group) / Ad spend
The control group is a set of people (or regions) who were eligible to see your ads but were deliberately held back. If the two groups are different sizes, scale the control group’s revenue up to the size of the exposed group before subtracting.
Worked example: ecommerce
A Mumbai D2C brand runs a four-week Meta prospecting campaign with a holdout test.
- Ad spend: ₹2,00,000
- Revenue Meta attributes to the campaign: ₹10,00,000, so platform ROAS = 10,00,000 / 2,00,000 = 5x
- Revenue from the exposed group: ₹12,00,000
- Revenue from the control group (scaled to the same size): ₹7,00,000
- Incremental revenue = 12,00,000 – 7,00,000 = ₹5,00,000
- iROAS = 5,00,000 / 2,00,000 = 2.5x
Now check that against margin. Break-even ROAS is 1 divided by your gross margin. At a 35% margin, break-even is 1 / 0.35 = 2.86x. The dashboard says 5x and the campaign looks like a winner. The iROAS of 2.5x says it is below break-even: ₹5,00,000 of incremental revenue at 35% margin is ₹1,75,000 of gross profit against ₹2,00,000 of spend, a ₹25,000 loss.
Worked example: lead generation
For lead-gen businesses without a revenue number per lead, use iCPA (incremental cost per acquisition), which Google defines as total ad spend divided by incremental conversions.
- Ad spend: ₹1,50,000
- Leads the platform reports: 300, so reported CPL = ₹500
- Incremental leads from the lift test: 120
- iCPA = 1,50,000 / 120 = ₹1,250 per incremental lead
If a lead is worth ₹1,000 to you after close rate, this campaign loses money on every incremental lead, even though the ₹500 CPL looked fine.
How to measure iROAS in practice
There are three practical ways to get an incremental revenue number. Pick the one your budget and data can support.
- Platform lift studies. Google Ads Conversion Lift (user-based or geography-based) and Meta Conversion Lift split eligible audiences into test and control groups for you. Google’s Bayesian Conversion Lift can run from a minimum spend of US$5,000, and results can start appearing from the 4th reporting day, though conversion processing can take up to 10 days (Google Ads Help). This is the easiest starting point for most brands.
- Geo holdout tests. Switch a campaign off in a set of matched cities or states and keep it running everywhere else, then compare revenue using your own sales data. This works across any channel and does not depend on cookies or platform tracking. In India, match regions carefully: a metro like Mumbai should not be compared with a tier-3 town.
- Marketing mix modelling (MMM). A statistical model that estimates each channel’s contribution from historical spend and revenue. Open-source options such as Google’s Meridian have made MMM more accessible, but it needs a long, clean history of spend and sales data, and it works best when calibrated with lift tests.
A single iROAS number is a snapshot at one spend level, so retest whenever you change budget significantly. Returns usually diminish as you spend more, so the iROAS you measured at ₹2 lakh a month will not hold at ₹10 lakh.
Common mistakes
- Treating platform ROAS as iROAS. Platform ROAS includes sales that would have happened anyway. Budget decisions made on it tend to over-fund retargeting and branded search. Run at least one lift test per major channel before you trust the platform number.
- Assuming retargeting is your best campaign. Retargeting often shows the highest ROAS because it reaches people already close to buying. In Seer Interactive’s Meta test, hyper-narrow retargeting segments were among the least incremental audiences. Test it before scaling it.
- Ending the test too early. A test that is too short or too small produces noise that looks like a result. Let it run until the platform reports a clear certainty level, and give conversions time to process.
- Contaminating the control group. A sale, a WhatsApp blast or an influencer push that only reaches one group breaks the comparison. Keep everything else equal across test and control for the full test window.
- Reading iROAS without margin. iROAS is still a revenue ratio. Compare it against your break-even ROAS (1 / margin) to know whether the incremental sales were actually profitable.
Frequently asked questions
What is a good iROAS?
A good iROAS is any figure above your break-even ROAS, which equals 1 divided by your gross margin. At a 40% margin, break-even is 2.5x, so an iROAS of 3x is profitable and 2x is not. An iROAS below 1x means the ads brought in less new revenue than they cost.
What is the difference between ROAS and iROAS?
ROAS divides all revenue a platform attributes to your ads by ad spend, including sales that would have happened without the ads. iROAS divides only the incremental revenue, measured against a control group that did not see the ads, by ad spend. iROAS is usually lower and is the better guide for budget decisions.
How do I measure iROAS with a small budget?
Start with the lift tools inside the ad platforms. Google’s Bayesian Conversion Lift studies can run from a minimum spend of US$5,000. If a platform test is not available, a geo holdout test, where you pause ads in a few matched cities and compare sales, can give a usable estimate from your own revenue data.
Is Meta’s incremental attribution the same as iROAS?
Not exactly. Meta’s incremental attribution setting uses machine learning models to predict which conversions were caused by an ad, while a true iROAS comes from a controlled experiment with a holdout group. Treat the attribution setting as a useful directional signal and confirm it with a lift test before moving large budgets.
How often should I measure iROAS?
Measure iROAS for each major channel at least once a quarter, and again before any large budget increase. Incrementality changes as spend, audiences and seasonality change, so a result from six months ago is not a reliable guide to today’s campaigns.
How Uginix helps
Uginix sets up lift tests and geo holdouts for clients so budget follows the revenue your ads actually cause, not the revenue a platform claims, and if a campaign’s iROAS sits below break-even, we tell you and reallocate the spend. Our analytics and attribution work reconciles Google and Meta reporting against your real sales data, and our posts on Meta Advantage+ and AI Max and Performance Max cover where automated campaigns tend to over-claim credit. Not sure how much of your ad revenue is truly incremental? That is exactly what our free growth audit looks for.
Related terms
- ROAS (Return on Ad Spend): attributed revenue divided by ad spend; the platform-reported number iROAS corrects.
- iCPA (Incremental Cost per Acquisition): ad spend divided by incremental conversions; the lead-gen equivalent of iROAS.
- Conversion Lift Study: a controlled experiment that compares an exposed group with a held-out control group to measure the conversions ads caused.
- MER (Marketing Efficiency Ratio): total revenue divided by total marketing spend; the business-level view of marketing efficiency.
- Marketing Mix Modelling (MMM): a statistical method that estimates each channel’s contribution to sales from historical data.
Sources: Google Ads Help: Conversion Lift metrics, Google Ads Help: Bayesian methodology in Conversion Lift, Think with Google: Incrementality testing, Seer Interactive, Social Media Today, Google for Developers: Meridian.