Experiment Sample Size Calculator

Plan your A/B tests with statistical rigor. Calculate the required audience size to detect meaningful changes.

1. Define Your Parameters

Before calculating, understand the levers of your experiment. Adjusting these values trades off between certainty, speed, and risk.

📊

Baseline Rate

The current conversion rate of your control group.

Why it matters: Events closer to 50% variance require smaller samples than rare events (e.g., 1%).
🎯

MDE

Minimum Detectable Effect. The smallest lift you care to find.

Trade-off: Detecting tiny changes requires massive sample sizes.
🛡️

Significance (α)

The risk of a "False Positive" (Type I error). Usually 5%.

Standard: 95% Confidence Level means α = 0.05.
🔋

Power (1 - β)

The probability of detecting a real effect (Type II error avoidance).

Standard: 80% is the industry standard balance of effort vs. risk.

2. Sample Size Calculator

Adjust the inputs to see how sample size requirements change dynamically.

Target Rate: 22.00%

Testing if B is different from A (better or worse).

Sample Per Variant 1,203 visitors
Total Required 2,406 visitors (Control + Variant)

Hypothesis Visualization (Standardized Z-Score)

Visualizing the overlap between the Null Hypothesis (A) and Alternative Hypothesis (B). As Sample Size increases (calculated above), these curves separate, reducing overlap errors.

3. The Cost of Certainty

How does changing the Minimum Detectable Effect impact the required sample size? Notice the exponential growth as you try to detect smaller improvements.