In this paper, we introduce two contingency factors --task type and network characteristics—that examine how individuals learn from experience. We hypothesize that task specialization and variation have positive impacts on IS professionals’ learning from experience. We further hypothesize that this performance effect of learning is contingent upon task type and characteristics of domain-specific knowledge networks. In particular, specialized experience will be more beneficial to learning when a task is a locating task-type or when network centrality is high. In contrast, varied experience will be more beneficial when a task is a diagnosing task-type or when network betweenness is high. The research model will be validated in the context of postimplementation enterprise system support. The study incorporates a social network perspective to study learning by experience, and contributes to the knowledge management field. Findings will provide practical insights on managing IT human capital and improving IS support services.