壽險業務員績效與業績目標關係之分析研究

Abstract

大部分的人壽保險公司都會實施獎勵制度來刺激員工的表現，然而要計算這些獎勵制度帶來的實際影響並不容易，也很難透過這些制度來預估每年實際的收入，因此保險公司時常面臨預估的目標與最後實際的收入有所差異，保險公司希望能夠了解每個獎勵制度所能帶來的收益有多少。 在本篇論文中，我們先透過資料視覺化 (Data Visualization) 的方式來了解公司的整體營運狀況，接下來我們利用分群演算法 (Clustering Algorithm) 來將保單以及產品的貢獻度進行分類，在特定的保險類型之產品對公司的整體收益有比較高的貢獻度。除了分析產品對公司的貢獻之外，我們也使用分群演算法來將每個業務員每年的貢獻進行分類，不同組的業務員的貢獻有不小的差距，因此，我們觀察到業務員在每年的表現中是有存在不同的差異。 利用上述的方法對公司有基本的了解之後，我們分析長期在公司工作的業務員，可以透過這些業務員來了解他們長期的發展跟公司之間的關係，經過分群演算法的計算後，大致上可以將這些業務員分成兩類，貢獻度較高的業務員人數大概佔所有業務員人數的百分之二十，而這些業務員的業績大概佔整間公司的業績之百分之八十，當我們將這相同的模型套用到每一年所有的業務員上，我們找到了相同的結論。 除了有八十二十法則存在於這間公司，我們也觀察到這間公司的產品與獎勵制度間的關聯非常重要，如果公司發行的產品符合市場需求並且有提出吸引業務員的獎勵制度，他們的收入可以在短時間內提高許多，如果在未來公司能夠符合上述的兩個條件，可以幫助他們的業績在短時間內迅速成長。

Implementing the incentive program is very common in the life insurance company. However, the real influence of these incentive systems cannot be easily calculated, and the organization suffers that it is unable to predict the actual income annually. The company usually faces great gaps between its target and the real operation with the incentive programs. The company wants to know the contribution of each incentive program when they implemented. In our thesis, we use the visualization method to understand the overall trend of the company. After that, we use clustering methods to separate the performance of insurance policies and products. We discover that some of the insurance types of products have a higher contribution. We also use clustering methods to separate the performance of each agent and we discover that each group of performance have gaps between them. We realize that the contribution between agents has great diversity. Afterward, we analysis the agents who have worked in the company for a long time to discover the relationship between agents and the company. We observe that the agents can generally be divided into two groups and those who in the top group is around twenty percent of agents in the company and their contribution to the company is near eighty percent. When we fit the same model back to all the agents in every year, we have the same conclusion. In addition to discovering the rule of 80-20 exists in the company, we observe that the company can improve their income if they launch the products meet the market requirement and provide the incentive programs that attract agents. In the future, the company can spur its performance growth rapidly when they meet the two factors that we mentioned above.