You’re launching a new credit-card acquisition flow with a revised APR disclosure and signup bonus. Design an end-to-end A/B test: define the unit of randomization, eligibility/exclusions (e.g., existing customers, fraud), primary and guardrail metrics (e.g., approved accounts, activation rate, delinquency/default), minimal detectable effect, statistical power, sample size, and expected test duration. Address selection bias (pre-approval, underwriting), cross-channel interference, peeking/early stopping, and regulatory constraints (e.g., fair lending). How would you analyze heterogeneous effects by segment while controlling false discovery?