Using Regression Analysis to Evaluate KPI Implementation at Volvo Penta North America

Sammanfattning: Most companies want to measure the performance of various areas of their operations. By doing so, it is easier to identify weaknesses or problems and take action to improve the performance in those areas. This study is conducted in collaboration with Volvo Penta North America and seeks to evaluate the possibilities of implementing a performance indicator for their dealers. The aim of this thesis is to investigate if there is a correlation between Volvo Penta’s evaluation system for their dealers, their Dealer Operating Standard score (DOS-Score) and their respective Sales Revenue, as well as the individual segments of the DOS and the Sales Revenue. In other words, if the evaluation system can be used as a performance indicator for how good the financial performance of a dealer is. The analysis is based on first-party data from Penta regarding the operation of Penta’s dealers. By using Linear Regression, it was found that the Adjusted R-Squared of the model with Aggregated DOS against Sales Revenue was 0.1403 and the Adjusted R-Squared for the model with the Segmented DOS against Sales Revenue was 0.1983. Thus, there is no significant correlation between the Aggregated DOS and Sales Revenue. However the results from the Segmented DOS-score against Sales Revenue indicates that it is possible to improve on the current DOS algorithm. Further research with more confounders considered is required to improve the model.