A/B Test Significance Calculator for Sample Size, Confidence and Uplift

    How to Calculate Sample Size and Uplift with the A/B Test Significance Calculator

    Calculate A/B test statistical significance, sample size, conversion uplift, confidence, and uncertainty before choosing an experiment winner.

    Formula

    Two-proportion test compares control and variant conversion rates with sampling uncertainty

    How to use the result

    Ship, stop, or continue only after checking effect size, uncertainty, sample adequacy, and guardrails.

    Statistical significance is not practical significance and repeated peeking inflates false positives. Inputs and calculations stay in this browser.

    VWO significance reference

    Determine statistical significance, calculate required sample sizes, and project the business impact of your experiments.

    Statistical Significance Calculator

    Determine if your A/B test results are statistically significant

    Control (A)

    Variant (B)

    Sample Size Calculator

    Determine how many visitors you need to detect a meaningful difference

    Test Duration Calculator

    Estimate how long your A/B test needs to run

    Revenue Uplift Calculator

    Project the business impact of your conversion rate improvement

    What is an A/B test significance calculator?

    An A/B test significance calculator helps you decide whether the performance difference between a control (A) and a variant (B) is likely caused by the change you made, rather than random chance. It estimates a p-value (or confidence) from your observed visitors and conversions and answers a practical question: “Is this result reliable enough to act on?”

    How to use this calculator

    Significance

    1. Enter Visitors and Conversions for Control (A).
    2. Enter the same fields for Variant (B).
    3. Select Calculate Significance to see conversion rates, uplift, z-score, and p-value.
    4. Use the result to decide whether to ship the variant or keep testing.

    Sample Size

    1. Set your Baseline Conversion Rate from historical data (not a tiny early sample).
    2. Choose a Minimum Detectable Effect (MDE) that is meaningful for your business.
    3. Pick Confidence and Power targets.
    4. Calculate the required sample per variant, then plan the experiment around that number.

    Duration

    1. Enter your Daily Visitors available to the experiment.
    2. Set the number of Variants (A/B is 2; multivariate tests are higher).
    3. Use the required sample per variant from the Sample Size tab.
    4. Calculate days/weeks needed to reach the total sample.

    Uplift

    1. Enter your Original and New conversion rates.
    2. Provide Monthly Visitors and an Average Order Value.
    3. Calculate additional conversions and the projected revenue impact.

    Practical tips for reliable experiments

    • Run tests for full business cycles (often at least 1–2 weeks) to avoid day-of-week bias.
    • Avoid peeking and stopping early; decide your sample size and decision rule upfront.
    • Track guardrail metrics (refund rate, bounce rate, lead quality) alongside conversions.
    • Validate tracking and randomization before interpreting any uplift.