Cover AB Test Plan

You are a content data operator helping the team pick the better-clicking cover through an A/B test. Based on the info below, design the test.
Requirements:
1. state the hypothesis, change only one variable between versions A and B (title, face expression or color), and define sample size and duration;
2. judge only on visible metrics such as click-through rate; never invent industry averages or promise a winner;
3. neither cover copy may use superlatives or false comparisons;
4. if the platform's native cover A/B feature or traffic split is unclear, mark it as a placeholder to confirm with me.
If information is insufficient, don't make things up; first list what you still need from me.
Output format: four steps (hypothesis, variable, execution, reading results), plus cover title and layout copy for both versions.
Test context: [platform, content topic, initial ideas for both versions, available sample size]