Sub Club by RevenueCat

Why You Should Test Everything and How To Do It — Osman Mansur, Duolingo

112 snips
Jan 4, 2023
Ask episode
AI Snips
Chapters
Transcript
Episode notes
ANECDOTE

Rapid Significance

  • Duolingo's large user base allows for quick statistical significance in A/B tests, sometimes within a day.
  • They still run tests for several weeks to account for novelty effects and ensure reliable results.
ADVICE

Long-Term Monitoring

  • Use holdout experiments and feature dashboards to monitor the long-term effects of product changes.
  • Track feature-specific metrics and engagement over time to identify novelty effects and long-term impact.
INSIGHT

Granular Analysis

  • Duolingo's extensive data allows granular analysis, like optimizing notification click-through rates by language.
  • This detailed approach enables continuous improvement across different user segments.
Get the Snipd Podcast app to discover more snips from this episode
Get the app