Work design is being constantly transformed by technology and requirements for polyvalent workers piloting the work system instead of operating it. Performance no longer depends exclusively on working hours, but in multiple aspects that balance the vision of production systems results. The purpose of this paper is to investigate self-management teams’ implementation and its effects on performance. A Brazilian cosmetics company case study of team development was used as a guidance for this investigation and company performance reports were the data source for descriptive statistics and multivariate analysis. Boxplot performance analysis was applied over team building stages along with One-way Variance Analysis to test average performance differences among stages’ transitions. Tukey test was sequentially applied to identify its statistic differences in pairs. Results revealed team development reached the expected performance over team stages. Forming to storming, though, was the only transition with no performance average gains. Storming to norming was the highest improvement, which meets literature principles. The present research is limited to a single sociotechnical environment and performance measurement is based on average data. Therefore, specific conclusions imply further research to extend findings to other contexts. However, the present study identified empowerment best practices to achieve superior performance. The originality of this paper consists on providing consistent connection between team development theory and practice, exploring sociotechnical approach benefits through practices into performance.