Decoding community enterprise sales performance drivers: visual analytics with heatmaps and decision trees

Community enterprises in Thailand’s secondary provinces struggle to identify which marketing factors drive sales performance. This study fills that gap by applying visual analytics and explainable machine learning to community enterprise operators in Chai Nat province. Using a sequential mixed-methods design, we combined semi-structured interviews with 21 key informants and a structured survey of 383 enterprise operators. The quantitative analytical stage—consisting of Pearson correlation, stepwise multiple regression, correlation heatmaps, and Gini-based decision trees—identified the hierarchical drivers of sales variance. Distribution channel effectiveness and value-based pricing emerged as the two dominant predictors, jointly explaining 31.5% of the variance. Product quality and promotion functioned as standardized baseline prerequisites rather than competitive differentiators. Decision tree analysis identified the point of transaction, product visibility, frictionless digital ordering, and secure payment as the critical catalysts for sales performance. For practitioners and policymakers, sustainable growth requires investing in customer-centric digital logistics and transactional convenience. Methodologically, this study shows how visual analytics and interpretable machine learning enhance transparency in applied business analytics, offering a replicable framework for similar resource-constrained regional markets.