Diagnose and optimize shared workspace marketplace conversion

You manage a two-sided marketplace for shared workspaces. Daily sessions are ~200,000. Booking conversion fell from 3.2% (2025-08-01 to 2025-08-14) to 2.4% (2025-08-15 to 2025-09-01) after a ranking change shipped on 2025-08-15; simultaneously, partner supply in downtown dropped ~10% due to local events, and paid traffic mix shifted from search to social. a) Build a metric tree from conversion down to component rates (search->view->contact->booking). Which slice-and-dice checks and counter-metrics would you run first to isolate cause vs correlation (supply elasticity, position bias, latency, partner acceptance, cancellations)? b) Propose a log-based analysis to attribute impact to the ranking change versus supply/mix confounders (e.g., diff-in-diff across unaffected suburbs, CUPED, synthetic controls). c) Design an A/B test for a candidate fix to ranking: define primary metric, guardrails (cancellations, partner rejection rate, search latency p95), and pre-registration. d) Powering: with baseline CVR 2.4%, target relative lift 6%, 50/50 split, 14-day runtime, 200k sessions/day, alpha=0.05, power=0.8—are we sufficiently powered? If not, show the knobs you would adjust (MDE, traffic allocation, duration), and justify risk controls for marketplace interference and SUTVA violations. e) Outline what you would ship if experiment results are heterogeneous across city, device, and traffic channel.