Manufacturing sector companies are typically categorized based on size or product type to aid in policy formulation and long-term planning. The existing classification systems, however, overlook economic performance, which makes it difficult to plan policies precisely and allocate resources. This study aims to close this gap by identifying economic factors that set manufacturing companies apart using data mining and machine learning. The main objective is to uncover similarity patterns among manufacturing companies and to describe each segment’s characteristics addressed in this study by utilizing historical data from the Manufacturing Economic Census Survey. The Cross-Industry Standard Process for Data Mining (CRISP-DM) framework guides the extraction of meaningful patterns from the economic census survey data. The study attempts to efficiently group manufacturing organizations according to their economic performance by utilizing unsupervised learning algorithms such as fuzzy C-means clustering and K-means clustering. The findings deviate from the existing categorizing method and yield six groups that are defined by financial metrics such as net book value, staff qualities, and spending distribution. Each segment is described in detail, including key characteristics such as business structure, primary activities, workforce qualifications, and financial metrics. These insights are instrumental in making informed decisions regarding future policy changes and financial aid initiatives, ultimately fostering a more resilient and competitive manufacturing sector. This research underscores the significance of data-driven approaches in elucidating industry dynamics and guiding strategic interventions to propel economic growth and enhance the well-being of Malaysians.