Minimum Reliably Detectable Effect (MRDE) Calculator
Calculate the Minimum Reliably Detectable Effect (MRDE / MDE)—the smallest relative and absolute metric uplift your experiment can detect with 80% or 90% statistical power at a 5% significance level.
Visitors, impressions, or users allocated to each experimental group
Current conversion, click-through, or engagement rate
Probability of detecting a true effect when one exists
Smallest percentage increase over baseline your test can reliably prove
How to Use This Calculator
- Enter your sample size per test group.
- Enter your current baseline conversion rate percentage.
- Select statistical power (80% standard).
- Review the smallest relative and absolute effect your experiment can detect.
Formula & Calculation Methodology
Derived by inverting the standard two-sample z-test power equation. As sample size n increases, the detectable effect threshold shrinks proportional to 1/√n.
Worked Example
20,000 users per branch at 4.0% baseline with 80% power
100,000 users per branch at 5.0% baseline with 80% power
Frequently Asked Questions
What happens if a true effect is smaller than the MRDE?↓
If the true effect is smaller than the MRDE, your test will likely finish inconclusive (p > 0.05), failing to reach statistical significance even if the variant is genuinely superior.
Why is MDE inversely proportional to the square root of sample size?↓
Because standard error scales with 1/√n. To cut your detectable effect in half (e.g. from 10% MDE down to 5% MDE), you must quadruple your sample size.
Can MRDE be calculated before launching an experiment?↓
Yes. Calculating MRDE during experiment planning is standard practice to prevent running "underpowered" tests that waste engineering time on undetectable goals.
What is a limitation of MRDE calculations?↓
It assumes binomial conversion variance. Continuous revenue metrics with heavy skew (whales) require higher sample sizes or log-transformation.
Related Calculators
Sequential Testing Duration and Boundary Calculator
Open calculator →Sample Size Calculator
Open calculator →False Positive Risk Calculator for A/B Test Peeking
Open calculator →Dunnett's Correction A/B/n Sample Size Estimator
Open calculator →Sample Ratio Mismatch (SRM) Diagnostic Analyzer
Open calculator →Confidence Interval Calculator
Open calculator →Source Citation & Disclaimer
Informational tool only; not licensed tax, legal, or financial advice.