The sales your marketing actually caused, beyond what would have happened without it. It is measured by comparing people who saw the ads with a similar group who did not.
Formula
Incremental conversions = conversions in test group − conversions in control group (scaled to the same size). Incremental ROAS = incremental revenue ÷ ad spend
Example
Meera's Instagram retargeting ad set costs ₹20,000 of her ₹60,000 Instagram budget. The test group, who saw ads, makes 60 purchases; an equal-sized control group, who did not, makes 45. Only 15 purchases, ₹30,000 at ₹2,000 each, were caused by the ads, so incremental ROAS is 1.5x.
How you'll hear it
“Retargeting reports 15x, but the lift test says incremental ROAS is 3x, so it's still worth running but not worth tripling.”
What it helps you decide
Where an extra rupee of budget actually creates sales. It often moves money away from retargeting and brand search, which take credit for buyers who were coming anyway.
Where people misread it
People read platform ROAS as caused sales. Retargeting and brand search show people ads just before they were going to buy, so attributed ROAS is high and incremental ROAS can be much lower.
The story
Meera's retargeting ad set shows ads to people who visited her website. It takes ₹20,000 of her ₹60,000 Instagram budget. Meta reports 6x ROAS: ₹20,000 spent and 60 purchases worth ₹1,20,000. Her freelancer wants to triple it.
Before she does, she runs a lift test. Meta holds back an equal-sized group from the same audience, who see no ads, and compares purchases.
| Purchases, test group (saw ads) | 60 |
| Purchases, control group (no ads) | 45 |
| Purchases caused by the ads | 15 |
|---|
| Ad spend | ₹20,000 |
| Reported sales (60 × ₹2,000) | ₹1,20,000 |
| Reported ROAS | 6x |
| Incremental sales (15 × ₹2,000) | ₹30,000 |
| Incremental ROAS | 1.5x |
|---|
Three out of four buyers were coming back anyway. The ads caused a 33% lift, worth ₹30,000. At 1.5x, the campaign only just pays for its clothes and its ads; her simple breakeven is 1.43x. So she keeps it, at its current size.
But tripling it would mostly buy more credit for people already on their way. She moves the extra budget to prospecting and tests that too.
All figures are illustrative. A real lift test needs enough purchases in both groups to give a confident result.
Work out incremental conversions and incremental ROAS
The example's numbers are filled in. Replace them with yours.
Compare incremental ROAS, not reported ROAS, with your breakeven ROAS.
Don't confuse it with
- Attribution: who gets credit for a sale; incrementality asks whether the sale would have happened anyway.
- Brand lift: change in awareness or recall, not in sales.
Related terms
Want to use this at work? AI-led PPC Excellence is aFactor's 16-week program for paid ads.
Sources
- About Conversion Lift, Google Ads Help
- Facebook Lift Metrics Glossary, Meta Business Help Center