Case Background

XX is a leading home shopping brand, which provides the high-quality and carefully-selected family life products, such as daily necessities, home appliances, digital electronics, beauty and skin care, apparel, nutrition & health care, cultural gifts, and etc.

With multiple channels of service system- TV, network, telephone, XX Company could meet the shopping needs of consumers. With the development of multimedia commerce, more and more customers could accept the new shopping methods. XX Company is faced with bigger market and more challenges. It becomes the urgent issues to be resolve by XX Company about how to increase the market share, maintain the profitability, and take advantage of the complementary effect between multiple channels.

Business Objectives

1、Take advantage of the proactive marketing advantages and capabilities of the Call-Out Center.

2、Improve the marketing accuracy under the existing configurations.

3、Improve the product arrangement better.

4、Seize the dynamics of customers proactively.

Solution

Model Development

The company selects IBM SPSS Modeler as the tool for precision marketing project. After completely learning the marketing model and development direction of XX’s Call-Out, it adds five major models to the existing business model to achieve precision marketing, including: potential customer rating model, customer segmentation model, product recommendation model, sales volume prediction model and customer churn model.

Potential customer rating model: The potential customer rating model ranks customers and identifies the characteristics of potential customer groups with high purchase intention. Under limited resources, it will firstly conduct marketing for high-potential customers.

Customer segmentation model: The segmentation model based on customer consumption characteristics not only outlines the characteristics of the customers who make high contribution, but also lays a foundation for further product marketing. It ensures the similarity of the same type of consumption habits.

Product recommendation model: It establishes the cross-sales rules between products based on the customer's previous purchase experience. For different customers, it provides limited recommendation order of related products for different customers.

Sales volume prediction model:It predicts the sales volume of each product category, product subcategory and single item based on the sales history, with the prediction period reaching 12 months.

Customer churn model: It judges the customer churn/dormancy situation in the future based on the actual purchase situation, so as to find out the key factors affecting the churn. Moreover, it rates the loss/dormant prediction, so as to give a clear warning of the possibility of churn, and to prompt the customer service personnel to call back to that customer.

System Deployment

To use the results of model analysis in a fast, more convenient and efficient way, the above model is directly connected with the XX Company's new generation of callback system. The model imports the analysis results in the business system based on a daily/weekly/monthly or real-time basis, so as to maximize its guidance effect.

Relevant Cases:

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