A/B Test Designer

Design a rigorous A/B test for: Hypothesis: [what we're testing]. Metric: [primary KPI]. Current conversion rate: [baseline]. Minimum detectable effect: [X%]. Confidence level: [95%]. Power: [80%]. Calculate: (1) Required sample size per variant. (2) Expected test duration. (3) Randomization unit (user/session/device). (4) Guardrail metrics to monitor. (5) Segmentation strategy. (6) How to handle novelty effect. (7) Statistical test to use (z-test/t-test/chi-squared/etc.). (8) Decision criteria before starting.