Developmental or evaluative? Understanding the impact of algorithmic evaluation on gig workers' thriving at work

Algorithmic evaluation plays a pivotal role in regulating gig workers' job performance; however, extant studies have reported inconsistent findings regarding its effects. Although prior research has examined the consequences of algorithmic evaluation, limited attention has been paid to how it influences gig workers' proactive behaviors such as job crafting, and subsequent thriving at work. To address this gap, this study identifies two types of algorithmic evaluation, namely developmental and evaluative algorithmic evaluation, and draws on approach-avoidance motivation theory to investigate their differential effects on thriving at work. Using survey data from 435 valid responses and employing an SEM-ANN-NCA approach, we find that evaluative algorithmic evaluation induces avoidance job crafting, which subsequently undermines thriving at work. In contrast, developmental algorithmic evaluation fosters approach job crafting, thereby enhancing thriving at work. Furthermore, time pressure strengthens the positive relationship between evaluative algorithmic evaluation and avoidance job crafting. This study contributes to the algorithmic management literature by revealing the dual nature of algorithmic evaluation and demonstrating how different evaluation approaches trigger distinct proactive behavioral responses and influence gig workers' thriving at work.