A/B READOUT ↗
Data Science · Analytics · Significance ↗

Is the lift real,
or just noise?

A two-proportion significance calculator for the holdout tests that validate every incrementality call — control vs. treatment, lift, confidence interval, and a clear verdict.

Open the tool ↓
The problem

A funded segment
needs proof.

An incrementality model estimates lift. Only a randomized holdout measures it. This tool runs the two-proportion test that confirms — or rejects — the assumption a budget decision was built on.

Two-proportion test
Enter control and treatment visitors and conversions — the tool computes conversion rates, absolute and relative lift, and a z-test p-value.
Confidence, not vibes
A 95% confidence interval on the lift, alongside the p-value, so "it worked" comes with a margin of error attached.
Pairs with incrementality
Built to confirm the treated-vs-baseline lift assumption behind an incrementality or targeting model before scaling spend.
The tool — A/B test readout

Did the funded segment actually lift — or just look like it did?

Control (holdout)
Visitors
Conversions
Treatment
Visitors
Conversions

Two-tailed two-proportion z-test. p pooled for the z-statistic; 95% CI on the difference uses the unpooled standard error. Significant at p < 0.05.
Seed values mirror the Segment 2 holdout referenced by the companion incrementality tool.