Guides the planning, design, and implementation of statistically valid A/B tests and conversion experiments.
This skill transforms Claude into an experimentation expert, providing a structured framework for moving from subjective guesses to data-driven decisions. It helps users construct strong hypotheses, calculate necessary sample sizes to reach statistical significance, and select appropriate primary and guardrail metrics. By following established best practices for variant design and traffic allocation, this skill ensures that experiments produce actionable insights while avoiding common pitfalls like the 'peeking problem' or testing too many variables simultaneously.
主要功能
01Structured hypothesis framework based on data and observations
02Sample size and duration calculation guides for statistical rigor
03Primary, secondary, and guardrail metric selection strategy
04Implementation guidance for both client-side and server-side testing
05Comprehensive post-test analysis and result interpretation checklists
060 GitHub stars
使用场景
01Optimizing landing page conversion rates by testing headlines, CTAs, and layouts
02Improving user retention by testing different onboarding flows and feature discovery
03Validating pricing strategy changes with server-side experiments and safety guardrails
