Ford Data Analyst Case Interview: Manufacturing, Warranty, and Connected-Vehicle Quality Analytics

This case mirrors real Ford interviews that emphasize pragmatic, operations-focused analytics and clear storytelling for cross‑functional stakeholders. You’ll play a Data Analyst supporting a vehicle launch quality team. Prompt: over the last four build weeks, warranty claims and connected‑vehicle alerts have spiked for a high‑volume Ford SUV built in a specific North American plant. Using messy, multi‑source data, quantify the issue, identify likely drivers, and propose actions a plant, supplier, and service network can take within the next two build cycles. What you receive: (1) warranty_claims.csv (claim_id, vin, repair_date, labor_op, part_number, symptom_code, dealer_code, total_cost); (2) production_builds.csv (vin, plant, line, build_week, model, trim, engine, transmission, supplier_id); (3) telematics_events.parquet (vin, event_ts, dtc_code, odometer_km, ambient_temp_C); (4) dealer_service.csv (vin, ro_open_date, wait_days, region, csat_score). Expectations (typical at Ford): • Clarify business framing (impact on customer, safety, throughput, and cost). • Sanity‑check data lineage and join keys (VIN, build_week) and note privacy safeguards (PII handling). • Compute and interpret Ford‑relevant KPIs: defects per 1,000 vehicles (DPV), warranty cost per vehicle (WCPV), first‑pass yield (FPY proxy from rework/claims), dealer cycle time, time‑to‑first‑failure, alert rate per 1,000 active vehicles. • Segment by plant/line, model/trim, supplier lot, build week, weather band, and dealer region; visualize with control charts and cohort plots to detect special‑cause variation. • Form hypotheses (e.g., supplier lot/shift effect, calibration version, environmental drivers), test with simple statistical checks, and quantify savings/risk (e.g., expected claim avoidance, throughput gain). • Produce a brief storyboard (exec‑ready) recommending actions such as a supplier containment, targeted OTA/TSB, or temporary process check at the station, including an implementation timeline and monitoring plan in Power BI. Evaluation criteria reflect Ford’s style: structured problem solving under time pressure, manufacturing and service domain fluency, ability to query/reshape data (SQL/Python), crisp visuals and narrative for plant/quality leaders, and a bias for practical, near‑term impact aligned to safety, quality, and customer satisfaction.