Quality control in manufacturing industries cannot be separated from the interaction between workers, machines, and work procedures, since most production failures stem not only from technical malfunctions but also from the dynamics of the socio-technical system as a whole. PT. XYZ, a paint manufacturing company in Bekasi, West Java, continues to face persistent defective product problems. Based on production data from January to October 2025, the number of defective products reached 625 units, representing 4.84% of total production of 12,977 units, which far exceeds the company's maximum tolerance standard of 2% per year. The two most dominant types of defects identified are color inconsistency with the established standard and the formation of bubbles on the paint product surface. This study aims to identify the types and causes of defective products, determine the most dominant risk factors, and formulate systematic improvement proposals. The Failure Mode and Effect Analysis (FMEA) method was applied to prioritize risks based on the Risk Priority Number (RPN), supported by the Pareto Diagram, Fishbone Diagram, and 5W+1H analysis. The results show that three causes were classified as critical, all exceeding the average RPN of 199.5: human error (RPN = 240), non-standard raw materials (RPN = 216), and poor product viscosity (RPN = 210). Improvement recommendations include routine operator training, consistent SOP enforcement, tighter raw material inspection, regular machine calibration, and strengthening inter-divisional communication between Warehouse, Quality Control, and Production divisions to reduce the defect rate below the 2% company standard.