What problem does it solve?
Product teams struggle to turn vague ideas into testable experiments, estimate the correct sample sizes, and interpret A/B results with practical statistical rigor; this Skill brings structure to hypothesis design, metric definition, and result-driven decisions.
Core Features & Use Cases
- Hypothesis formulation: Write clear If/Then/Because hypotheses that specify intervention, metric, and causal rationale.
- Metric selection and guardrails: Define a single primary decision metric plus guardrail and secondary diagnostics to avoid misleading wins.
- Sample size calculation: Estimate per-variant and total sample requirements from baseline rates/means, MDE, alpha, and power.
- Prioritization and launch rules: Score experiments with ICE (Impact, Confidence, Ease), set stopping rules, and avoid common pitfalls like peeking or instrumentation drift.
- Result interpretation: Translate p-values and confidence intervals into business-relevant decisions and investigate heterogeneity and novelty effects.
Quick Start
Use the experiment-designer skill to write an If/Then/Because hypothesis, compute the required sample size with baseline and MDE, and rank the test using ICE for prioritization.
