TrafficLoopback
Sign in / Sign up

A/B Test Sample Size Calculator

How many visitors do you need before you can call a winner?

Inputs

5%

Your current conversion rate for the control.

10% relative lift

Smallest lift you care about detecting. Smaller MDE = much bigger test.

Results

Sample size per variant
31,232
visitors each
Total sample size
62,464
2 variants combined
Estimated test duration
32 days (4.5 weeks)
Detecting a lift from 5.00% to 5.50% (absolute change of 0.50 pp).
Long test
Consider increasing traffic, relaxing your MDE, or running a bandit instead so you stop wasting impressions on losing variants.
Bandit alternative (estimate)

A multi-armed bandit could reach a winner with roughly 30-50% fewer impressions because it stops sampling losing variants early. Below is a conservative estimate using a 40% reduction.

Bandit total
37,478
~40% fewer impressions
Bandit duration
19 days (2.7 weeks)

Sample size vs MDE

At your baseline of 5%, smaller lifts require dramatically more traffic. Marker shows your current MDE of 10%.

1k10k100k1M5%10%20%30%40%50%31,232 / variantMinimum detectable effect (relative lift)Sample size / variant (log scale)

Skip the math. Let a bandit run your tests for you.

TrafficLoopback runs Thompson-sampling bandits on your owned traffic, so you find winning ad creatives without manual significance tests. Free during beta.

Start free in beta
More free tools
Bandit Simulator
Watch a multi-armed bandit beat a 50/50 split — live.
Significance Calculator
Is your A/B test result real, or just noise?
Break-Even Calculator
What CPC can you actually afford to pay?