Method
FDR correction
Benjamini-Hochberg control of the false discovery rate.
7 cases from 7 companies apply FDR correction. 1 was checked in an online A/B test; 1 report results in production.
Typical tasks: Avoid false wins in A/B tests, Get experiment results faster, Predict long-term effect from short tests, Evaluate models and 2 more. Companies: Spotify, Pinterest, Zillow, Lyft, Etsy and 2 more. Often combined with: Bayesian A/B test, Power analysis / MDE, A/B test, t-test, LTV prediction and 20 more.
| Why Spotify Is Not Using Bayesian A/B Testing | media-streaming | 2026-09 | |
| From Eight Hours to Eight Minutes: Automating A/B Test Analysis on Databricks | gaming | 2026-06 | |
| Measure Less to Learn More: Using Fewer, Higher-quality Metrics to Capture What Matters | social | 2026-04 | |
| LLM-Powered Relevance Assessment for Pinterest Search | social | 2025-12 | |
| The Data Infra Behind Zillow's 3x Growth in Experiment Volume | marketplace | 2023-03 | |
| Challenges in Experimentation | mobility | 2022-04 | |
| Imbalance detection for healthier experimentation | marketplace | 2022-03 |