Enter the working inputs
Compare control and variant conversion rates with a two-sided significance test, confidence interval and practical-lift threshold.
Premium experiment analysis
Measure conversion lift without hiding the uncertainty.
Enter aggregate experiment counts to separate observed lift from statistical evidence and estimate the modelled value difference at a common traffic level.
1 → 3Premium runs per rolling 24 hours after free verification.
A/B Test Significance Analyzer workspace
Premium experiment analysis
A/B Test Significance Analyzer
Conversion and lift metrics, a two-variant table and a copy-ready experiment report.
Method and output
Plan with visible assumptions.
Review the method
The analyzer uses a pooled two-proportion z-test and a confidence interval for the absolute conversion-rate difference, then compares relative lift with the entered practical threshold.
Keep the useful output
Conversion and lift metrics, a two-variant table and a copy-ready experiment report. Copy or download the available export before leaving the page.
Check the boundary
Statistical significance is not the probability that a variant is better. Repeated peeking, poor randomisation, novelty, seasonality and multiple tests can invalidate the conclusion.
Before you run it
A/B Test Significance Analyzer FAQs.
Can I use this Premium tool for free?
Yes. A guest can use 1 Premium run per rolling 24 hours. A signed-in, verified free account receives 3 runs shared across all 20 Premium tools.
What does Premium change?
The £9 monthly Premium plan removes the 1-run guest and 3-run free-account allowances for normal interactive use and includes the available copy and file exports.
Does UtilHub save my Premium tool history?
Saved project history is not currently included. Premium inputs are sent to UtilHub for processing, the result is returned to this page, and you should download anything you want to keep.
How does the A/B Test Significance Analyzer work?
The analyzer uses a pooled two-proportion z-test and a confidence interval for the absolute conversion-rate difference, then compares relative lift with the entered practical threshold.
What should I check before using the result?
Statistical significance is not the probability that a variant is better. Repeated peeking, poor randomisation, novelty, seasonality and multiple tests can invalidate the conclusion.
Premium membership
Unlimited access to all 20 tools and exports for £9 per month.
Create an account to choose Premium. Creating an account does not by itself start a charge.