department of informatics

Customer Performance Measurement - Analysis of the Benefit of a Fuzzy Classification Approach in CRM

Printer-friendly versionSend by emailPDF version
Details of Authors
Authors: 
Darius Zumstein
eMail: 
dzumstein@gmx.ch
About
Thesis Type: 
Master
Submission Date: 
1. March 2007
Abstract: 

Customers are the most valuable asset of a company. As a result, customers have to be clas- sified, analysed, evaluated, segmented and managed according to their value for the company using appropriate tools and methods of Customer Relationship Management (CRM). This master thesis proposes fuzzy classification as a multidimensional data analysis and man- agement method suitable for realising these CRM processes and for establishing profitable customer relationships. In contrast to other data mining and statistical methods, fuzzy classifi- cation and fCQL (fuzzy Classification Query Language) allow the classification of customers into more than one class at the same time.
The application of the fuzzy classification approach to widely used management tools like the SWOT, portfolio and ABC analysis and to scoring models enables a better and fairer classifi- cation, segmentation and management of customers. So far, these methods have mostly been applied uncritically with sharp classes, although sharp segmentation can obviously be very arbitrary, imprecise, unfair and discriminatory and may have negative effects.
The application of the fuzzy portfolio analysis within the scope of performance measurement is especially suited to classifying, analysing, evaluating and improving important monetary cus- tomer performance indicators, like turnover, contribution margins, profit and customer equity, and non-monetary indicators, such as customer value, satisfaction, loyalty and retention. Surprisingly, little research has been done on Customer Performance Measurement (CPM) and customer performance indicators despite the increasing theoretical and practical impor- tance of CRM. This work discusses a holistic customer performance measurement framework with 170+ Customer Performance Indicators (CPIs) and relevant Key Customer Performance Indicators (KCPIs). To avoid misclassifications, to improve the quality of customer evaluations and to exploit customer potential, it is suggested to classify all indicators fuzzily.
Customer performance indicators for revenue and profitability, and customer investment, rela- tionship, recommendation, information and cooperation indicators allow to segment customers precisely, to optimise fuzzy classified customer portfolios, to drive the proposed CRM success chain and to define customer strategies in order to increase corporate profits and growth.

Keywords: 
Fuzzy classification, fuzzy Classification Query Language (fCQL), Customer Relationship Management (CRM), analytical CRM (aCRM), customer performance measurement, customer performance indicators, fuzzy customer segmentation, management tools, fuzzy portfolio analysis, fuzzy credit rating.
Documents