Mystery shopping (MS) is a widely used tool to monitor the quality of service and personal selling. In consultative retail settings, assessments of mystery shoppers are supposed to capture the most relevant aspects of salespeople’s service and sales behavior. Given the important conclusions drawn by managers from MS results, the standard assumption seems to be that assessments of mystery shoppers are strongly related to customer satisfaction and sales performance. However, surprisingly scant empirical evidence supports this assumption. We test the relationship between MS assessments and customer evaluations and sales performance with large-scale data from three service retail chains. Surprisingly, we do not find a substantial correlation. The results show that mystery shoppers are not good proxies for real customers. While MS assessments are not related to sales, our findings confirm the established correlation between customer satisfaction measurements and sales results.