<![CDATA[This study aims to analyze the influence of burnout levels, uncertainty of working hours, and income uncertainty on the productivity of online motorcycle taxi drivers in Mataram City. The study used a quantitative approach with a survey method and purposive sampling, involving 110 active drivers with a minimum of six months of experience. Primary data were obtained through a 1–5 Likert scale questionnaire, while secondary data were obtained from literature and previous research to strengthen the theoretical framework. The burnout variable was measured based on three dimensions of Maslach and Jackson (1981): uncertainty of working hours through daily hour fluctuations and workday inconsistencies, uncertainty of income through income stability and incentives, and productivity through work effectiveness and income balance. The analysis was conducted using Random Forest via Python according to the method of Liaw and Wiener (2002) to model non-linear effects and assess the contribution of each variable to productivity. The results showed that burnout was the most dominant factor in determining productivity, followed by uncertainty of income, while uncertainty of working hours had a limited and contextual influence. Drivers with low to moderate burnout and stable income showed higher productivity, while high burnout was correlated with low productivity. These findings confirm that work flexibility in the gig economy does not automatically increase productivity; Burnout management and income stability are prerequisites for optimal performance (Amrudin, 2019; Ginting et al., 2025). This research provides empirical understanding of the dynamics of gig economy work in mid-sized cities and forms the basis for employment policies that are more responsive to the conditions of digital platform workers.]]>
