Enterprise Background

The telephone marketing system of an insurance company has experienced several years of business development, with the number of customers marketing more than 30 million, the number of insured customers more than 30, 000, and the success rate of 0.1184%. How to transform open marketing into refined marketing and how to use the accumulated historical information to better locate the valuable and demanding customers has become an insurance company 's current problem to be solved.

Enterprise Objectives

1、Automation and process of decision-making to reduce human error and shorten decision-making period.

2、Business personnel need to obtain intuitive decision-making assistance information.

3、Make the data available to provide value for the enterprise more effectively.

Solution

The telephone marketing system has a loop process from data creation to customer marketing to data recovery and reuse. How Telephone Sales Representative (TSR) utilizes collection of historical information to combine data mining technology with business processes to improve efficiency has become increasingly important.

How to refine

Documentation: Initially screen the customers who have been documented, narrow the scope of customer marketing, and lock in the customers who are most likely to respond.

TSR customer acquisition link: Convert random customer selection into systematic customer classification.

Marketing: targeted marketing of insurance products to target customers.

Model development steps——take "Product Recommendation-Insurance Model" as an example:

Based on the data mining platform of SPSS Modeler, the customer product recommendation scheme is established by using data mining technologies such as cluster, association rules and sequence association. The existing system needs TSR to judge the products recommended to the customers according to the customer underwriting history information and business guidelines, and the automatic product recommendation scheme can automatically recommend the products recommended by the insured customers according to the customer purchase product sequence correlation.

1. Customer Clustering

Data: customer information, such as gender, age and occupation.

Establish customer clustering model: customers in the same class have greater similarity in underwriting requirements, different classes in the maximum difference in requirements.

2. Analyze the rules for insured products for different customer groups

Data: information about previous underwriting by the customer.

Analyze the sequence rules of customer insured products in different groups.

3. Establish automated product recommendations

Regularly and automatically update according to the latest information of customer underwriting. New rules can be found and utilized in a timely manner. Combine product recommendation with electronic marketing system to realize automation.

System Process

Application Effect

Reduce the workload of TSR manual judgment, and reduce the risk of artificial judgment.

The sequence rules can be updated automatically according to the latest information of customer underwriting, and the new rules can be used to recommend products to customers in real time, thus improving the underwriting rate of customers.

Relevant Cases:

1. Apparel Brand Retailer – Creating a Mobile and Real-Time Information Platform
2. Shopping Basket Analysis or Recommendation Engine for Retail Industry
3. Call-Out Precision Marketing
4. Insurance Company Builds the Telemarketing Response Prediction Model Based on SPSS Modeler