Best Practices
- Using Split Dependencies to Isolate Experiments
- Interpreting Inconclusive Results
- Successfully Experiment with Anonymous and Logged-in Users
- Querying Data in Split
- Running an A/A Test
- Using the metric details and trends data
- Choosing your degradation threshold for alerting
- Naming Event Types
- Metric Filtering
- Metric Capping
- Reviewing Metrics during an experiment
- Sample Ratio Mismatch Check
- Best Practices for Full Stack Experimentation