Name
Abstract | ||
|---|---|---|
Purpose - The purpose of this paper is to present application of recency, frequency and monetary value (RFM) approach to predict customer insolvency using telecommunication data corresponding to RFM of late payments. The study tackles a serious problem that telecommunication companies often face and shows the ways to deal with it. Design/methodology/approach - Based on a real telecom customer data, RFM approach was tested against decision trees and logistic regression models. Proposed models were evaluated with lift measure, area under the receiver operating characteristic and the ability to detect significant amount of money owed by insolvent customers. Findings - The main findings from the research are twofold: RFM approach offers a viable alternative for customer insolvency classification. The proposed models perform well and all of them can capture significant amount of money owed by insolvent customers what is of high importance for the revenue assurance. Originality/value - In comparison to previous studies proposed research presents novelty in the following areas. First, it deals with RFM applied to insolvency data (previous studies dealt with direct marketing data). Second, with these three variables it is possible to act as an early warning system for predicting the risk level and probable anomalies as quickly as it is possible (data retrieval and computational time is reduced). Third, RFM approach was tested against decision trees and logistic regression and the quality of the models was also assessed three months after the estimation. |
Year | DOI | Venue |
|---|---|---|
2016 | 10.1108/K-04-2015-0113 | KYBERNETES |
Keywords | Field | DocType |
Economics,Decision making,Risk management,Knowledge management,Modelling,Telecommunications | Decision tree,Telecommunications,Computer science,Insolvency,Originality,Risk management,Revenue assurance,Logistic regression,Payment | Journal |
Volume | Issue | ISSN |
45 | 5 | 0368-492X |
Citations | PageRank | References |
2 | 0.39 | 10 |
Authors | ||
1 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Tomasz Zabkowski | 1 | 32 | 11.28 |