online_retail_II.xlsx
Customer Segmentation
Dynamic RFM behavioural intelligence using four-cluster K-Means segmentation.
Customers analysed
4,312
Valid identified purchasers
Selected clusters
4
K-Means K value
Silhouette score
0.331
Current cluster cohesion
Davies-Bouldin index
1.011
Current cluster separation
High-Value Loyal Customers
778
18.04% of customers
Recent, frequent and high-spending customers.
At-Risk Regular Customers
1,211
28.08% of customers
Previously active customers requiring retention attention.
Recent Low-Value Customers
945
21.92% of customers
Recent customers with lower purchasing value.
Dormant Customers
1,378
31.96% of customers
Customers with the longest time since their latest purchase.
Customer segment distribution
Distribution rebuilt from the active dataset.
High-Value Loyal18.04%
At-Risk Regular28.08%
Recent Low-Value21.92%
Dormant31.96%
RFM cluster profile
| Segment | Customers | Avg recency | Avg frequency | Avg monetary |
|---|---|---|---|---|
| High-Value Loyal | 778 | 13.83 days | 13.56 | £7,381.58 |
| At-Risk Regular | 1,211 | 80.21 days | 4.14 | £1,759.99 |
| Recent Low-Value | 945 | 23.95 days | 1.97 | £532.08 |
| Dormant | 1,378 | 190.57 days | 1.30 | £305.66 |