8.11 Cluster characteristics

Once we form the clusters, our next task is to find the characteristics of the customers in those clusters. First, we will see the average values pertaining to recency, frequency, and monetary value.

cluster_data %>%
  group_by(km_cluster) %>%
  summarise(mean_recency = mean(recency),
            mean_frequency = mean(frequency),
            mean_mv = mean(monetary_value),
            members = n()) %>% 
  mutate_all(round, 3)

Table 8.3: Cluster Characteristics

km_cluster	mean_recency	mean_frequency	mean_mv	num_members
1	1.514	3.931	688.823	72
2	3.298	7.567	1859.503	104
3	0.191	11.702	3368.381	141
4	9.350	2.503	513.232	183

Table 8.3 shows that customers in cluster 1 spent $3,368 in the last 12 quarters and have low recency and high frequency.⁴⁸ Cluster 2 customers seem to be the worst lot because they spent only $513 in the last 12 quarters with high recency and low frequency. It’s possible that the customers in the segment are no longer planning to return and buy anything. Thus, the firm may have lost them.

An important exercise at this point is to name these customer segments. 
For example, cluster 2 can be named "Lost Cause" because they are probably 
not returning. Can you think of names for other customer segments?

---

48. Low recency means that these customers made purchases very recently.↩