The basic definition of customer segmentation is the process of organizing customers into specific groups based on shared characteristics, behaviors, or preferences.
Basically, there are two methods for observing business problems such as customer retention, customer loss analysis (churn), customer finding and acqusition efforts based on KPIs metrics in data science and business world, these are RFM analysis and CLTV (customer lifetime value) prediction. We'll use RFM analysis in this project with Olist dataset that has information of 100k orders from 2016 to 2018 made at multiple marketplaces in Brazil.
RFM is the determination of some metrics based on the shopping behaviours of customers in the past, and the creation of scores based on these metrics. Here is that metrics I emphasized:
- Recency: The freshness of the customers activity, time since last order
- Frequency: The frequency of customers transactions or visits, total number of transaction
- Monetary: The money of customers spends on purchase, total or average transaction value
