Are your ads driving incremental sales — or sales that would have happened anyway?
Holdout test inputs
Results
Incremental ROAS—
Reported ROAS
—
Incremental %
—
Incremental revenue
—
About this calculator
Incrementality is the truth check on attribution. Reported ROAS counts every conversion that touched an ad, but a meaningful percentage of those customers would have converted via direct, organic, or branded search regardless. Incrementality testing isolates how much revenue is actually caused by the ad spend.
This calculator does the math from a holdout test. Compare revenue-per-user in the exposed group (saw ads) vs control group (held out). Multiply the difference by exposed-group size to estimate incremental dollars. Divide by ad spend for true incremental ROAS — usually 30-60% of reported ROAS.
The discipline of running incrementality tests is hard. Each test takes 2-4 weeks of clean data, requires patience to not over-react to noise, and produces results most operators don\'t want to see. The first incrementality test usually reveals that one channel is producing far less incremental revenue than reported — which is uncomfortable but actionable.
Pair with the Tracking Gap Estimator (different problem — under-reporting), the Click vs View Attribution Gap, and the Attribution Model Comparison. Together these tools triangulate honest channel ROI vs platform-reported ROAS.
Frequently asked questions
What is incrementality?
Incremental sales are revenue that would NOT have happened without the ad. Reported attribution counts every conversion that touched an ad, but many of those customers would have converted via direct, organic, or branded search anyway. Incrementality measures the truly new revenue.
How do I run an incrementality test?
Geo or audience holdout test. Pick comparable groups (e.g., 80% of users see ads, 20% are held out). Compare revenue per user across groups. The difference attributable to ads = incremental lift. Most platforms (Meta, Google) offer holdout test features.
What's a typical incrementality rate?
Branded search: 10-30% incremental (most converts would happen via direct anyway). Cold paid social: 60-90% incremental (these customers wouldn't have found you otherwise). Retargeting: 20-50% incremental (many would have returned anyway). Numbers vary hugely by brand and channel.
Why does this matter for ROAS?
Reported ROAS counts non-incremental conversions, overstating true lift. A campaign showing 4× ROAS at 50% incrementality is actually delivering 2× true ROAS. Decisions made on raw ROAS systematically overspend on channels that take credit for sales they didn't cause.
What incrementality test should I run?
Start with branded search — usually has the lowest true incrementality (controversial). Then test retargeting incrementality. Then top-of-funnel. Each test takes 2-4 weeks of clean data and shows you which channels are actually driving new revenue.