How to Properly Split-test Your WordPress Website
summary
Step-by-step checklist for valid WordPress split tests: set goals and KPIs, pick metrics and behavior/qualitative tools (Google Analytics Content Experiments, Optimizely, Visual Website Optimizer, Nelio, heatmaps and session recording), calculate sample size and honor the two-sided t-test requirement (example: 1,500 visitors per variation over 44 days), account for seasonality and business cycles, avoid caching and browser/device mismatches, prefer client-side jQuery/JS variations or separate URLs, expect ~30–40% win rate and treat “losing” tests as learning.
AI-written summary · about
chapters (27)
- 0:00Introduction and welcome
- 0:45What is split testing
- 3:18Multivariate vs split testing
- 5:12Why split test designs
- 6:19Usability testing vs split testing
- 7:05Stepwise vs continuous redesign
- 8:30Continuous testing benefits
- 9:38Setting up measurements
- 12:07Defining sample size
- 13:55Seasonality in testing
- 14:39Aligning tests with business goals
- 15:48Creating testing roadmaps
- 16:34Business cycle considerations
- 17:40Technical checks before testing
- 19:28Tools and resources list
- 20:12What to test first
- 22:19Testing on WordPress
- 23:46Testing this presentation
- 24:51General testing pitfalls
- 27:03Misconceptions about failure
- 28:08Success rates in testing
- 28:30WordPress caching issues
- 30:16Plugin overload problems
- 31:42Clean testing solutions
- 32:02Recommended resources
- 32:26About the speaker
- 33:07Q&A and closing
