نوع مقاله : پژوهشی
نویسندگان
1 گروه مدیریت صنعتی، پردیس بین المللی کیش، دانشگاه تهران، تهران، ایران
2 گروه مدیریت، دانشکده مدیریت و علوم مالی، دانشگاه خاتم، تهران، ایران
3 گروه مدیریت تکنولوژی و نوآوری، دانشکده مدیریت صنعتی و فناوری، دانشگاه تهران، تهران، ایران
4 گروه مدیریت صنعتی ، دانشکده مدیریت و علوم اجتماعی ، دانشگاه آزاد اسلامی واحد تهران شمال ، تهران ، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
The clustering of customers and analysis of their lifetime value is one of the basic strategies of policy-making in production and marketing. Since the advertising strategies of many businesses are implemented regardless of the financial behavior clusters of customers and the expected income resulting from the dynamic financial behavior of customers, the advertising policy based on the customers’ lifetime value in the dynamic clusters of their financial behavior can save a lot of advertising and marketing costs to manage imposed costs on businesses. This is even though the studies conducted in the field of customer lifetime value analysis have not paid attention to advertising strategies based on lifetime value. The existing methods for determining the lifetime value of customers rely on static clustering of customers' status. In contrast, customer clustering based on their financial behavior, it is a dynamic and time-varying phenomenon. Therefore, in this research, the lifetime value analysis of customers and the optimization of advertising strategy in the dynamic clusters of their financial behavior using a self-organizing neural network and dynamic stochastic programming have been discussed. For this purpose, based on the status of each customer belonging to a cluster during consecutive weeks, a Markov chain is formed from the clusters containing customers. The lifetime value of customers is based on the transition probability matrices of customers' status from one cluster to another. It is also estimated based on the expected income of customers in each cluster. The results showed that customer clustering with the self-organizing neural network method can explain at least 91% of the changes in the data. Also, the findings showed that the optimal advertising strategy for each cluster of customers with a certain expected income level will lead to different lifetime values for customers, which shows the importance of dynamic clustering of customers and determining the optimal level of advertising cost in each cluster to gain more customers lifetime value.
کلیدواژهها [English]