LAUSR

Abstract This study addresses the following question: For a given managerial, firm, and industry setting, which individual metrics are effective for making marketing-mix decisions that improve perceived performance outcomes? We articulate the key managerial takeaways based on testing a multi-stage behavioral framework that links decision context, metrics selection, and performance outcomes. Our statistical model adjusts for potential endogeneity bias in estimating metric effectiveness due to selection effects and differs from past literature in that managers can strategically choose metrics based on their ex-ante expected effectiveness. The key findings of our analysis of 439 managers making 1287 decisions are that customer-mindset marketing metrics such as awareness and willingness to recommend are the most effective metrics for managers to employ while financial metrics such as target volume and net present value are the least effective. However, relative to financial metrics, managers are more uncertain about the ex-ante effectiveness of customer-mindset marketing metrics, which attenuates their use. A second study on 142 managers helps provide detailed underlying rationale for these key results. The implications of metric effectiveness for dashboards and automated decision systems based on machine learning systems are discussed.