Building an RFM Analysis Model for a Wine Company

RFM (Recency, Frequency, Monetary) Analysis is a method of customer segmentation that marketing teams can use to better understand and market to their existing customers. Wine companies have customer segmentation patterns that makes this type of analysis especially valuable for identifying up-sell opportunities, targeting marketing campaigns and promotions to the right customers, and improving churn rates.

I worked with a wine company to develop and deploy an RFM analysis model in Google BigQuery and created dashboards in Qlik that allowed their marketing and sales teams to derive insights and make marketing decisions from the model.

RFM analysis typically involves scoring each customer based on their relative ranking for:

• Recency - how recently the customer made a purchase
• Frequency - how often the customer makes purchases
• Monetary - how much customers spend per purchase

After customers get a score, some segmentation method is manually applied - generally categories like VIP customers, loyal customers, etc. are determined by looking at the data and determining thresholds.

The model I created used percentiles to do this type of scoring, but then also used a k-means clustering method to automate segmentation across the 3 areas, leading to a more organic segmentation of customers. The marketing team could then look at the segments and determine targeted actions to make for each group.