New Trackly AI turns buyer intent into campaign action. Explore Trackly AI
Qualent Media Lead EstimatorSelect your ideal customer profile to see your estimated reach and potential CS, HQL, BANT, webinar, and MQL leads with Qualent Media.
AI Buyer Intent
Solutions
Services
IndustriesCase Studies
Resources
About Lead EstimatorSelect your ideal customer profile to see your estimated reach and potential CS, HQL, BANT, webinar, and MQL leads with Qualent Media.

Free B2B marketing tool

A/B Test Significance Calculator

Compare A/B conversion rates with a two-sided significance test and 95% difference interval. Estimate sample size at 80% power using your planned lift.

No sign-up. Clear method. Downloadable results.

Start with your inputs.

Enter your own figures, review the calculation and export the results. Defaults are illustrative examples.

Whole-number count for the same reporting period. Allowed: 1 to 1,000,000,000,000; whole numbers only.
Whole-number count for the same reporting period. Allowed: 0 to 1,000,000,000,000; whole numbers only.
Whole-number count for the same reporting period. Allowed: 1 to 1,000,000,000,000; whole numbers only.
Whole-number count for the same reporting period. Allowed: 0 to 1,000,000,000,000; whole numbers only.
Enter a percentage between 0 and 100. Allowed: 0.001 to 100.
Allowed: 0.001 to 1,000.

Illustrative example

Your results.

Complete valid inputs to see results.

Estimates depend on your assumptions. No inputs are sent or saved by this tool.

Method and definitions

Two-sided pooled two-proportion z test: pooled p = (conversions A + B) ÷ (visitors A + B); SE = sqrt(p × (1 − p) × (1/nA + 1/nB)); z = (rate B − rate A) ÷ SE. The difference interval uses unpooled standard error and 1.96. Sample planning uses 95% confidence, 80% power and equal allocation.

Use independent random assignment, unique binary conversion counts and a preplanned fixed horizon. This normal approximation requires at least 5 observed successes and failures in each arm; otherwise inference is unavailable. Do not repeatedly peek or test many variants without correction. Statistical significance does not measure practical value. Sample size assumes a two-sided test, equal allocation and the chosen baseline and relative lift.

Reference: NIST two-proportion test.

Worked example

300 conversions from 10,000 A visitors versus 400 from 10,000 B visitors gives 3% versus 4%, a 1-point absolute increase and approximately p = 0.00012.

Use this tool in your workflow

  1. Replace the example inputs with your own figures or facts.
  2. Use matching definitions, units and reporting periods.
  3. Review the method and scope before applying the output.
  4. Copy or download the results for your next planning discussion.

For help turning a plan into a campaign, explore Qualent Media’s B2B campaign strategy services.

A clear basis for your next step

Questions, answered.

Understand the method and the boundaries of the result.

How does this a/b test significance calculator work?

Two-sided pooled two-proportion z test: pooled p = (conversions A + B) ÷ (visitors A + B); SE = sqrt(p × (1 − p) × (1/nA + 1/nB)); z = (rate B − rate A) ÷ SE. The difference interval uses unpooled standard error and 1.96. Sample planning uses 95% confidence, 80% power and equal allocation.

How should I interpret the result?

Use independent random assignment, unique binary conversion counts and a preplanned fixed horizon. This normal approximation requires at least 5 observed successes and failures in each arm; otherwise inference is unavailable. Do not repeatedly peek or test many variants without correction. Statistical significance does not measure practical value. Sample size assumes a two-sided test, equal allocation and the chosen baseline and relative lift.

Can I download my results?

Yes. CSV export includes the inputs, unrounded numeric results where applicable, method and scope. Copy results is also available; if clipboard access is unavailable, a text download is provided. This tool runs in your browser without sending or storing your inputs.

Build a clearer marketing plan.

Discuss your audience, campaign priorities and measurement with Qualent Media.

Discuss your campaign