Why test
- Winning copy/layout can add 10–30% CVR
- Shared template keeps every experiment comparable
Spreadsheet model
- Columns: Variation • Hypothesis • Change • Sample • Start • Result • Decision
- Naming: ss_[set]*[date]*[hypothesis]
What to test first
- Headline hook
- Background / contrast
- Order of benefits
Method
- Create Set A and Set B
- Run for 7–14 days
- Pick winner by CVR or installs
Tips
- One change per variation
- Give each variant at least a week of traffic