Question 1

How do I calculate sample size for an A/B test?

Accepted Answer

Pick your metric type, set your current rate/value, the smallest change you care about, power (80–90%), and significance (usually 5%). The tool tells you how many people you need in each group and in total.

Question 2

What is Minimum Detectable Effect (MDE)?

Accepted Answer

It’s the smallest lift you want to be able to detect. Smaller MDE → more people needed. Bigger MDE → fewer people needed.

Question 3

When should I use proportion vs. continuous?

Accepted Answer

Proportion: yes/no outcomes (e.g., conversion). Continuous: numbers (e.g., revenue, time). For continuous, provide mean and standard deviation for each group.

Question 4

Can I use unequal traffic splits (e.g., 80/20)?

Accepted Answer

Yes. Unequal splits are OK. They often need more total traffic but can reduce exposure to a risky variant.

Sample Size Calculator

Frequently asked questions