What problem does it solve? Teams often run experiments without clear hypotheses, stop tests too early, or misread statistical significance, leading to false conclusions and wasted traffic. This Skill provides a rigorous framework for planning, running, and analyzing A/B tests and building a continuous experimentation program. ## Core Features & Use Cases - Hypothesis & Test Design: Structures hypotheses with the observation-belief-outcome-metric framework and selects the right test type (A/B, A/B/n, MVT, split URL) with traffic allocation guidance. - Sample Size & Duration Planning: Provides quick-reference sample size tables by baseline conversion rate and minimum detectable effect, plus duration formulas and sequential testing guidance via the references guide. - Experiment Program Management: Covers ICE prioritization, experiment velocity metrics, playbook documentation templates, and weekly-to-quarterly review cadences. - Use Case: A growth marketer wants to test a new pricing page headline. The Skill calculates that at a 3% baseline conversion rate they need roughly 31,000 visitors per variant to detect a 20% lift, defines primary and guardrail metrics, and produces a pre-launch checklist and results documentation template. ## Quick Start Ask the assistant to help design an A/B test for a specific page or change, providing your current conversion rate and traffic volume.
