|Rosella Machine Intelligence & Data Mining|
| Home | DataMining & Machine Learning | Products & Downloads | Site Map | Contact |
<!- menu column -->
What is Predictive Customer Lifecycle Management (CLM)?
The purpose of Customer Life-cycle Management (CLM) is to maximize both customer retention and profit. CLM differs from CRM in the sense it emphasizes time-varying nature of customer management. CLM process starts with marketing research analysis and customer acquisition followed by conversions, up-selling, re-selling, cross-selling, and more importantly churn management. It attempts to maximize utilization of customer potentials which is the most important resources in business.
How Predictive Customer Relationship Management (CRM) differs?
In essence, predictive CRM and predictive CLM covers the same customer care activities with the exception that CLM has more emphasis on time-varying customer patterns. Both predictive CRM and predictive CLM covers activities described in the subsequent sections.
Customer Acquisition and Market Research
Acquiring new customers is a very costly and time consuming process. Efficient acquisition is thus very important. Customer understanding is the critical factor for developing efficient and effective marketing strategies and plans. Customer understanding can be obtained from existing customers. Predictive methods such as customer profiling and relevancy analysis can be used. If customer information is not available or scant, marketing survey may be used to collect relevant information.
Many customers start as trial basis. Early customer experience can have determining effect. Analyzing customer usage patterns and preferences, conversions to long term customers can be identified and intensive support is provided to those who can become long term loyal customers. Predictive rule inference based on predictive modeling can automatically detect customers in need of care, and thus improving customer conversions. The following figure shows customer dashboard indicating customer's predictive lifecycle prospects. This will greatly improve customer visibility;
Customer Attrition and Churn Management
Customers do stop buying for various reasons. Some may no longer need services. Others may seek for better alternatives. Common reasons include lower-cost products and better suited products. Predictive methods can identify churn potentials proactively and preventive measures can be taken. The figure above also contains a churn risk indicator. The customer dashboard contains very useful information determining what actions should follow for the customer concerned. For more, please read customer retention.
Debt Collection Management
Timely collection of debts occurring from business is very important for business. While soft methods may prolong collection periods or lead to bad debts, aggressive collection methods may upset loyal customers. Customer-specific appropriate collection strategies are needed. Collection management with rule-based predictive business knowledge is the way to go.
Customer transactions often carry risk for bad debts. Predictive analytics can be used to identify potentially risky transactions and mitigate risk. The figure at the section "Customer Conversions" also shows a delinquency risk indicator, which is very important lifecycle management information. For more, please read credit risk analysis and insurance risk analysis.
Fraud is a commonly occurring problem in business. Sophisticated detection methods based on predictive rule-based business knowledge can be used to audit potentially fraudulent transactions and claims. Real-time auditing of transactions can mitigate the risk of fraudulent transactions processed.
As more and more business transactions are processed without human involvement, it is important to provide real-time monitoring and auditing systems. For more, please read Trandaction Audit and Monitoring
Good customer life cycle management also involves selling other related products. Based on past purchase records, recommendations for other products can be generated using predictive methods, thereby increasing the probability of cross and up selling. For more, please read Basket Analysis.
Up-selling is to sell higher-value products to customers of existing services or products. Customers who bought products in the past may be offered with higher-value products after certain periods. Note that in case of service products, customers may be offered immediately since there are no physical products that should be phased out.
Re-selling generally applies to consumable goods. When customers consume purchased products, customers are supposed to re-order. Sending reminder mails in a timely manner can not only boost sales but also prevent churns. Ideally catalogs are sent to customers so that products are arrived one or two week earlier than the consumption of previous purchases. Customer segment specific catalogs may be developed and sent based on individual customerís purchasing patterns. For more, please read RFM marketing.
Customer Loyalty Programs
Many businesses offer various customer loyalty programs. Predictive analytics can be used to segment customers into management groups.
Customer trends change constantly. Understanding of trend changes and directions is very important in business. Predictive trend analysis provides business visibility. Time-series sales trend analysis may be used to monitor and detect changes in customer trends. For more, please read sales trend analysis.
Call Center Analytics
Predictive CRM/CLM is essential for call-center based customer care since it can better utilize call-center resources, which is call center operators. Prioritizing customer calls with predictive analytics, greater efficiency of call centers can be achieved.