Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 55630 (01)
- University: Sharif University of Technology
- Department: Industrial Engineering
- Advisor(s): Kianfar, Farhad
- Abstract:
- Businesses need to narrow down their segmentation and divide customers into smaller groups because some low-loyalty customers have the potential to become loyal customers. Therefore, we have focused on non-loyal customers. This study proposes a framework in which customers with low loyalty value are examined in four six-month time periods and an overall two-year time period and divided into smaller periods. First, a clustering was performed with K-Means algorithm and RFM model, then the migration pattern of customers from one cluster to another was identified using the sequential pattern technique. The transition of customers and their potentiality was determined and according to these criteria, we can determine whether our non-loyal customer has the potential to be spend on or not. The proposed framework was implemented using transaction data of two years of a retail store. Customer clustering was done in both two years data and four time periods data. The migration pattern was done and gave an understanding of the dominant patterns of customers. Potential customers were identified and the level of customer transition between clusters was determined. The focus of the thesis is on customers with low loyalty, and according to that, cluster one and two (with low loyalty) were divided into four categories for more detailed analysis: potential, lost, low value, and worthless. Each of these categories has different marketing policy to make them more loyal.
- Keywords:
- Customer Loyalty ; Customer Relationship Management ; Data Mining ; Marketing ; Customer Segmentation
Digital Object List
-
محتواي کتاب
- view
Bookmark