Performance-based incentives are widely used in B2B sales; however, sales managers find it challenging to customize such incentives. In response, we propose a new methodology to assess salespeople's capability using performance history data. To evaluate relative performance, we filter out the impact of systematic factors of job characteristics (e.g., channel type, territory, and market size). Therefore, by providing appropriate benchmarks, our approach captures the idiosyncratic component of performance, allowing a direct comparison of a salesperson with other counterparts in the firm. We also account for prospective customer characteristics that let us distinguish between competent salespeople and those fortunate enough to have a customer mix with a high predisposition to buy. Using a data set of automobile insurance sales to small business owners in the U.S., we model salesperson performance as consisting of three stages: generating leads, converting leads to customers and upselling. Our output, a “grade card” for each salesperson, contains metrics that can aid managers in setting individual level performance goals, in deciding who should be rewarded or let go, and in diagnosing salespeople's strengths and weaknesses. Our easy to implement methodology is useful for a wide variety of applications, wherever relative performance needs to be judged.

基于绩效的激励措施被广泛用于B2B销售中。但是，销售经理发现定制此类激励措施具有挑战性。作为回应，我们提出了一种新的方法来使用绩效历史数据来评估销售人员的能力。评估亲戚绩效，我们会滤除工作特征的系统性因素（例如渠道类型，地区和市场规模）的影响。因此，通过提供适当的基准，我们的方法可以捕获绩效的特有成分，从而可以将销售人员与公司中的其他业务人员直接进行比较。我们还考虑了潜在的客户特征，这些特征使我们能够区分称职的销售人员和足够幸运的销售人员，这些客户具有很高的购买倾向。使用向美国小企业主销售的汽车保险数据集，我们将销售人员的绩效建模为三个阶段：生成销售线索，将销售线索转换为客户以及向上销售。我们的输出是每个销售人员的“成绩卡”，包含的度量标准可以帮助经理制定个人水平的绩效目标，确定应该奖励或放手的对象以及诊断销售人员的优缺点。我们的易于实现的方法论对于需要评估相对性能的各种应用都是有用的。