Incrementality / Lift Test Calculator
Measure the true incremental lift of a marketing channel using geo holdout or audience holdout. Enter test + control group conversion volumes, get the lift percentage, statistical significance, and ROI implications.
Measurement Intelligence
Attribution over-reports. Incrementality tells you what drove the sale.
A customer who was going to buy anyway doesn't count as an incremental conversion. The only way to know true causal impact is to turn the channel off for a representative audience (geo or user holdout) and compare. This tool runs the statistics on that comparison.
Test design
All three designs use the same math. The right design depends on your channel and platform.
Measurements
Treatment group (saw ads)
Control group (held out)
Results
Incremental Lift
0%
vs control
Significance
–
p-value
Incremental Conv.
0
above control baseline
iCAC
$0
Cost per incremental conv.
iROAS
0.0x
Incremental ROAS
Verdict
Pending
Enter data above
Next steps
Incrementality numbers belong in attribution model comparison
iROAS from this tool should be your "true ROAS" benchmark against the various attribution models. Compare them side by side.
Frequently asked questions
What is incrementality?
The lift caused by an ad campaign that wouldn't have happened without it. The honest version of attribution.
How is incrementality measured?
A holdout test. Serve ads to a treatment group, hold out a control group, compare conversions. The difference is incrementality.
What is iROAS?
Incremental ROAS. (Incremental revenue) / (ad spend). The honest version of ROAS, excluding conversions that would have happened anyway.
When is incremental ROAS lower than attributed ROAS?
Almost always. Attribution counts organic conversions as caused by the ad. Incrementality removes them.
What is a geo holdout?
A test where some geographies get the campaign and matched geos are held out. Standard for channels that can target geo.
What is a user-level holdout?
Random users in the targeted audience don't get the ad. Measurable in Meta, Google Ads (Conversion Lift Studies), and tools like Haus.
How long should an incrementality test run?
4 to 8 weeks typically. Long enough to see conversion patterns. Short enough that market conditions don't drift.
How much lift is "statistically significant"?
Depends on sample size and baseline conversion. The calculator runs a two-proportion z-test: enter numbers to see the significance.
Can I test brand campaigns for incrementality?
Yes but requires patience. Brand effects compound over months, not weeks. Longer test windows (quarter plus) are needed.
Does Google Ads have a built-in incrementality test?
Conversion Lift studies for Display and YouTube, free. For Search, the only option is a geo or budget split test externally.
What is MMM?
Marketing Mix Modelling. A different approach using regression on long time-series data. Complements incrementality tests. Facebook's Robyn and Google's Meridian are open-source options.
Free · no account
SEORCE for Chrome, Edge and Firefox
Audits the page you are on: headings, meta tags, schema, indexability, and whether AI crawlers can read it. Export the result as a PDF. Free, with no account.
It audits one page at a time and does not track AI answers. About the extension
Traffic from search and AI
See which visits came from search and AI.
SEORCE Web Analytics runs from one script tag, with Google Search Console connected. It shows visits from every AI source, from search and from campaigns, and it is free on every plan.
3,000+ users · Free forever plan, no card needed