iROAS (incremental return on ad spend) measures how much revenue your advertising generated from conversions it actually caused, excluding purchases that would have happened regardless of whether the ads ran.
That distinction matters more than most performance marketers realize.
Incremental ROAS reveals how much of your revenue advertising truly caused
ROAS vs. iROAS: The Core Difference
Standard ROAS is calculated as total attributed revenue divided by ad spend. If your platform reports 420 conversions at an average order value of $50 and you spent $4,200, your ROAS is 5x. Clean and simple.
The problem: not all 420 conversions were caused by the ads. Some of those customers were going to buy anyway, they just happened to see an ad before they did. Attribution models can't tell the difference. They assign credit to whatever touchpoint was most recent (or most observable), regardless of whether it caused the purchase.
iROAS strips out those organic conversions:
iROAS = incremental revenue / ad spend
Incremental revenue = conversions that would not have happened without the advertising.
A Concrete Example
You run a retargeting campaign. 100 conversions are attributed to it. Your average order value is $50 and you spent $1,000.
Attributed ROAS: 100 conversions × $50 / $1,000 = 5x
You run a holdout test. You suppress ads to 20% of your retargeting audience (the control group) and show ads to the remaining 80% (the test group). After four weeks, you find:
- Test group conversion rate: 4.2%
- Control group conversion rate: 3.5%
- Incremental conversion rate: 0.7 percentage points
Of those 100 attributed conversions, only about 30 were incremental, the rest would have happened without the ads.
iROAS: 30 conversions × $50 / $1,000 = 1.5x
The same campaign. The same time period. One number is 5x, the other is 1.5x. The platform reports the 5x version.
Why the Gap Exists
The gap between ROAS and iROAS is largest when your audience already has high purchase intent, which is exactly what defines most retargeting campaigns.
Retargeting targets people who visited your website. Someone who added a product to their cart and left is already seriously considering buying. They may well buy in the next few days with or without seeing your ad. When they do buy and click a retargeting ad on the way, the ad gets credit for a conversion it didn't create.
The same dynamic applies to brand search. Most people who type your brand name into Google were going to visit your site anyway. Bidding on your own brand name captures that traffic, but it doesn't generate it.
When iROAS and ROAS Are Close
For prospecting campaigns targeting cold audiences, people who have never heard of your brand, organic purchase intent is low. There is no counterfactual purchase that would have happened without the ad. In this case, most attributed conversions are genuinely incremental, and ROAS and iROAS will be relatively close.
This is one reason why prospecting campaigns often look worse in attribution than retargeting: they're reaching cold audiences with lower stated purchase intent. But their iROAS can be surprisingly competitive with retargeting when measured correctly.
What Counts as a Good iROAS
There is no universal answer, it depends on your margins, customer lifetime value, and cost structure. But a rough framework:
- iROAS below 1x: The campaign is losing money in incremental terms. You are paying more in ad spend than you are generating in incremental revenue.
- iROAS 1x–1.5x: Marginal. May be viable depending on gross margin, but worth scrutinizing.
- iROAS 1.5x–3x: Generally efficient for most business models.
- iROAS above 3x: Strong. Consider scaling carefully, high iROAS can sometimes reflect underspending rather than exceptional efficiency.
The threshold that matters most is your gross margin. If your product margin is 40%, you need at least a 2.5x iROAS to generate positive contribution after ad spend and cost of goods. Build your iROAS targets from your margin structure, not from industry benchmarks.