Definition #
ML systems running multivariate tests and interpreting results.
Key Characteristics #
- Sample size calculation
- False discovery control
- Auto-winner implementation
- Cross-channel sync
Why It Matters #
Identifies optimal variants 5x faster than manual testing (Optimizely).
Common Use Cases #
- Landing page optimization
- Email subject line testing
- Pricing page layouts
Examples #
- Google Optimize
- VWO
- AB Tasty
FAQs #
Q: Statistical significance?
A: Auto-stops tests at 95% confidence level.
Q: Mobile app testing?
A: SDKs for iOS/Android with real-time feature flags.