Repeat Customer Behavior Analysis for an Online Retail and Distribution Client

Our clients is one of a largest ecommerce company Its is one of the first comprehensive virtual retail business, selling credible brands to customers nationally, operating in a multimedia environment

The client objective is to:

Understand the customer’s store preference and their revisit behaviours  and evaluate the performance and ROI outcome of campaigns which are running at the same time at different malls.

Transorg’s solution

• Developed a Random Forest trained on 80% of dataset
• Variables such as demographics, transactional, product category, CRM (complaints, queries…), campaign type TV/ Web/Both

Key Impacts

• Predict customers with high propensity to purchase in next month with an accuracy of 87%
• Customers with vintage > 6 months and made last purchase within 52 days are highly likely (64%) to repurchase in the next month
• Customers with only 1 transaction in past 3 months have significantly lower (12%) repeat purchase than those with multiple transactions (36%)
• Females tend to repurchase more than males