Formula
Approximate sample size uses baseline rate, desired lift, confidence level, and statistical power.
A 5% baseline rate with a 20% relative lift often needs thousands of visitors per variant.
Scenario / Breakdown
How to use this calculator
Enter the current campaign or growth numbers, then compare the main result, secondary metrics, score, and recommendation. The score is a simplified signal designed to help prioritize optimization work.
Tips to improve results
- Run tests long enough to cover weekly behavior patterns.
- Avoid calling winners too early.
- Use larger samples for small expected lifts.
- Prioritize high-impact tests when traffic is limited.
What the result means
A strong result usually means the current stage is efficient enough to scale. A weak result usually means the campaign has leakage, poor economics, weak conversion, or insufficient sample quality. Use the interpretation with your actual cost, revenue, and tracking data.
FAQ
What is A/B test sample size?
This calculator provides a practical estimate and interpretation for a/b test sample size calculator. Use it as a planning tool, then compare results with your actual marketing data.
What is minimum detectable effect?
This calculator provides a practical estimate and interpretation for a/b test sample size calculator. Use it as a planning tool, then compare results with your actual marketing data.
What is statistical power?
This calculator provides a practical estimate and interpretation for a/b test sample size calculator. Use it as a planning tool, then compare results with your actual marketing data.
How long should an A/B test run?
This calculator provides a practical estimate and interpretation for a/b test sample size calculator. Use it as a planning tool, then compare results with your actual marketing data.
Why do small lifts need larger samples?
This calculator provides a practical estimate and interpretation for a/b test sample size calculator. Use it as a planning tool, then compare results with your actual marketing data.