A/B testing with React Native
summary
Practical guide to running A/B tests in React Native: fetch experiment JSON at splash, use the Optimizely SDK activate/track flow or — better — return variants from a server-side decision to handle language/platform targeting. Native release cycles delay iterations (~3 weeks) so prefer CodePush for faster rolls. Watch for UX regressions (example: bumping images 150→500px is neutral then 300px wins after rerun), expect many weak hypotheses (author cites ~90% failures) and timebox a splash-screen timeout.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:21Defining A/B Testing
- 2:15Why use A/B testing?
- 5:18When to avoid testing
- 7:27The A/B testing process
- 7:51Forming hypotheses
- 11:16Real-world experiment case
- 13:21Analyzing experiment results
- 14:03Iterating experiments
- 14:29Web vs native testing
- 15:13Implementing mobile tests
- 18:03Personal story and break
- 18:27Remote config settings
- 18:40Using Optimizely
- 18:54Technical implementation
- 22:33Server-side decisions
- 23:08Network resilience
- 23:44Common pitfalls
- 24:38Reproducibility and culture
- 26:57Sharing knowledge
- 32:31Closing remarks
related talks
08mReact Native and Badoo story of a massive experiment
06mSuccessful rapid prototyping with React Native
27mBuilding a Product with React Native
11mTips for Building Successful React Native Apps
24mThe Performance Myths of React Native apps
30mMaking React Native Fast & Lean
31mThe performance myths of React Native
29mA practical approach to building truly native cross-platform applications at scale