A/B Sample Size (Conversion Rate)

Inputs

Assumes independent Bernoulli trials (visitors), equal-duration exposure, and a fixed baseline conversion rate.

Baseline conversion rate (control), %

Effect size mode

Minimum detectable effect (MDE)

Significance (alpha)

Power (1 - beta)

Test

Allocation ratio (treatment:control)

Number of variants (A/B/n)

Multiple-comparisons alpha adjust

Optional: total daily visitors eligible (both groups)

Notes: this is a planning estimate. Sequential peeking, multiple variants, bot traffic, and non-stationary baselines will break assumptions.

Result

Enter inputs and click Compute.

Method (formula)

Let p1 be baseline control rate, p2 treatment rate, \Delta = |p2 - p1|.

We use a common approximation for two proportions with allocation ratio r = n2/n1:

zα = Φ⁻¹(1-α/2)  (two-sided)
zα = Φ⁻¹(1-α)    (one-sided)
zβ = Φ⁻¹(power)

n1 ≈ ( zα*√(p̄(1-p̄)(1+1/r)) + zβ*√(p1(1-p1) + (p2(1-p2))/r) )² / Δ²
n2 = r*n1
where p̄ = (p1 + r*p2) / (1+r)

This matches the style of many online calculators; it’s not a substitute for a proper experiment plan.