Leveraging RFM Analysis for Customer Segmentation to Boost Customer Engagement within Olist Brazilian E-Commerce Platform

Background

Under the period of February 2017 to August 2018, 96.648% of Olist customers didn’t make a second purchase. Olist Brazilian E-Commerce businesses must modernize their marketing and customer strategies by personalizing them to improve their services, retain customers, and attract new ones.

Methodology

1. Define the research questions.
2. Collect the data.
3. Clean the data.
4. Perform Exploratory Data Analysis (EDA).
5. Perform Feature Preprocessing to create recency, frequency, and monetary value columns.
6. Create RFM segments using quantiles.
7. Define RFM segments.
8. Group RFM segments.
9. Analyzing RFM segments.
10. Build a Tableau dashboard.

Conclusion

• An analysis of 75,389 Olist customers revealed that 35.99% are "lost customers," characterized by long purchase intervals and low order volume. Additionally, 23.5% are classified as "high risk to churn" due to their recent purchase inactivity. In total, a concerning 59.49% of our customers exhibit low engagement with the platform. Given the substantial size of this segment, regaining their interest is crucial.
• Conversely, 12.40% of customers are identified as "potential customers," signifying recent purchases but minimal spending.
• The remaining customers are distributed across categories including "new customers," "potential loyalists," "very loyal," and "VIPs," with respective percentages of 9.19%, 8.98%, 6.80%, and 3.12%.
• While attention should be paid to all segments, prioritizing strategies to retain "lost customers" and those at "high risk to churn" is most pressing.