Dataiku Case Studies Churn Prevention

Edit This Case Study Record

	Churn Prevention Dataiku

Churn Prevention

Dataiku

Technology Category	Analytics & Modeling - Predictive Analytics
Applicable Industries	E-Commerce
Applicable Functions	Sales & Marketing
Use Cases	Predictive Replenishment
Services	Data Science Services
Challenge	Showroomprive, a leading e-commerce player in Europe, was facing a challenge with customer churn. The company was using static rules to trigger marketing actions, which were common to all customers and did not take into account the individual value of each client. This approach was not effective in preventing churn and improving customer loyalty. Showroomprive wanted to refine its client qualification process to anticipate, prevent, and reduce churn rates. The company aimed to detect clients with a high potential of no longer buying from the website based on individual purchase rates and refine the targeting of marketing campaigns for each potential churner to improve customer loyalty. Read More
About Customer	Showroomprive.com is a leading e-commerce player in Europe. Founded in 2006, the company has over 20 million members and hosts about 15 flash sales per day, attracting over 2 million visitors. In 2014, Showroomprive generated €480M in business volume, representing a 40% growth compared to 2013. The company was facing a challenge with customer churn and wanted to refine its client qualification process to anticipate, prevent, and reduce churn rates. Read More
Solution	Showroomprive used DSS to develop a solution that predicts whether or not a buyer will return to the website to make a purchase. The solution automated the integration and enrichment of a variety of data sources, including customer data, order and delivery data, and web logs. It created more than 690 features derived from this data depending on variables such as clicks on sales, orders, litigation, customers, and more. The solution also tested multiple machine learning algorithms to achieve the best predictive model. This approach allowed Showroomprive to internalize all of the work revolving around this solution, from R&D to production. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Predictive Analytics

Applicable Industries

E-Commerce

Applicable Functions

Sales & Marketing

Use Cases

Predictive Replenishment

Services

Data Science Services

Challenge

Showroomprive, a leading e-commerce player in Europe, was facing a challenge with customer churn. The company was using static rules to trigger marketing actions, which were common to all customers and did not take into account the individual value of each client. This approach was not effective in preventing churn and improving customer loyalty. Showroomprive wanted to refine its client qualification process to anticipate, prevent, and reduce churn rates. The company aimed to detect clients with a high potential of no longer buying from the website based on individual purchase rates and refine the targeting of marketing campaigns for each potential churner to improve customer loyalty.

About Customer

Showroomprive.com is a leading e-commerce player in Europe. Founded in 2006, the company has over 20 million members and hosts about 15 flash sales per day, attracting over 2 million visitors. In 2014, Showroomprive generated €480M in business volume, representing a 40% growth compared to 2013. The company was facing a challenge with customer churn and wanted to refine its client qualification process to anticipate, prevent, and reduce churn rates.

Solution

Showroomprive used DSS to develop a solution that predicts whether or not a buyer will return to the website to make a purchase. The solution automated the integration and enrichment of a variety of data sources, including customer data, order and delivery data, and web logs. It created more than 690 features derived from this data depending on variables such as clicks on sales, orders, litigation, customers, and more. The solution also tested multiple machine learning algorithms to achieve the best predictive model. This approach allowed Showroomprive to internalize all of the work revolving around this solution, from R&D to production.

Impact #1	Showroomprive was able to detect potential churners amongst mono-buyers with an AUC of 0.819.
Impact #2	The company was able to internalize all of the work revolving around the solution, from R&D to production.
Impact #3	The solution allowed Showroomprive to automate the integration and enrichment of a variety of data sources, create more than 690 features derived from this data, and test multiple machine learning algorithms to achieve the best predictive model.

Impact #1

Showroomprive was able to detect potential churners amongst mono-buyers with an AUC of 0.819.

Impact #2

The company was able to internalize all of the work revolving around the solution, from R&D to production.

Impact #3

The solution allowed Showroomprive to automate the integration and enrichment of a variety of data sources, create more than 690 features derived from this data, and test multiple machine learning algorithms to achieve the best predictive model.

Benefit #1	Showroomprive was able to detect potential churners with 77% accuracy.
Benefit #2	The company was able to create more than 690 features derived from a variety of data sources.
Benefit #3	Showroomprive achieved an AUC of 0.819 in detecting potential churners amongst mono-buyers.

Benefit #1

Showroomprive was able to detect potential churners with 77% accuracy.

Benefit #2

The company was able to create more than 690 features derived from a variety of data sources.

Benefit #3

Showroomprive achieved an AUC of 0.819 in detecting potential churners amongst mono-buyers.

Download PDF Version

Overview

Churn Prevention

Operational Impact

Quantitative Benefit