Making up performance: the construction of “performance” in venture capital firms’ portfolios

Purpose - The purpose of this paper is to investigate performance measurement practices in venture capital firms. Specifically, the author examines how two organizations make complexity and uncertainty manageable by mobilizing performance measurement. This study draws on the framework of pragmatic constructivism (PC) (Nørreklitet al., 2016; Nørreklitet al., 2006, 2010), focusing on the integration between the four dimensions of PC, namely, facts, values, possibilities and communication. Design/methodology/approach - The paper adopts a comparative case study methodology. Findings - The findings show that performance measurement practices are strongly influenced by values playing out via integration in actor-world relations, whereas Venture A mobilizes an actor-centric approach, leading to an open, holistic performance measurement system (PMS) which is based on non-financials, a close involvement in operational matters and actors’ judgement using accounting as “learning machines” (Burchellet al., 1980); and Venture B draws on an analytical approach emphasizing on the role of financial indicators and control enacting accounting as “answer machines” and as “tool for computation” (Burchellet al., 1980). These different approaches to PMS, actor-centric vs analytical, are guided by different values about actor-world relation(s). Originality/value - The paper provides a context-sensitive account on the relationship between uncertainty and performance measurement practices. First, this paper contributes by providing evidence on how actors use accounting to manage uncertainty and complexity by differently integrating actor-world relation(s) (Nørreklit,et al., 2016; Nørreklitet al., 2006). Second, this study resonates with recent calls for more industry-specific and context-sensitive investigations (Messner, 2016). Finally, the author contributes to the literature asking for more research on the role accounting plays in managing uncertain conditions (Chenhall and Moers, 2015).