Customer Segmentation: RFM Scoring, Personas & k-Means by Hand
RFM from a raw order log, quintile scores with PERCENTILE and PERCENTRANK, named segments, segment value charts, z-scores, a k-means iteration by hand, the elbow method and campaign economics: 8 graded labs on one fictional homeware shop.
The first lesson is free — no account needed.
About this course
A customer segmentation course that builds an RFM analysis and a small k-means clustering by hand on one fictional homeware shop, for marketers, CRM analysts and data students. You turn 24 raw orders into recency, frequency and monetary value per customer, score 20 customers from 1 to 5 with data-driven quintile cut-offs from PERCENTILE and PERCENTRANK, map the scores to segments such as Champions and At Risk and size a win-back campaign, compare each segment's share of customers with its share of revenue in a chart, standardise the variables with z-scores so that no variable dominates the distance, run one k-means iteration with distances, nearest centroids and AVERAGEIF, choose k with the within-cluster sum of squares and an elbow chart and decide which segments get the autumn mailer from their expected net profit. The first lab is free.
Lessons
- Recency, Frequency and Monetary from a Raw Order LogFree
- Quintile Scores: PERCENTILE Cut-offs and PERCENTRANK Pro
- From Scores to Segments: Champions, At Risk & Co. Pro
- Segment Size vs Revenue Share Pro
- Standardising Variables with z-Scores Pro
- k-Means on RFM: One Iteration by Hand Pro
- Choosing k: Within-Cluster Sum of Squares and the Elbow Pro
- Segment Profiles and Targeted Campaign Economics Pro