Factor evolution explorer

The Factor Explorer panel helps to understand the evolution of churn, cross-sell, up-sell, etc. factors and to navigate into the root-causes.

The Factor explorer is organized with 3 panels :

  • Set of factors - top panel : reminder about the related and the selected filters, the cluster or product view (see Deep Diving into Cluster) and 3 sets of factors as 3 different tabs
    • Factors In cluster : set of factors defined by the cluster's definition and their co-correlated factors and linked factors (see below)
    • Factors In Product Family : main factors included into all the clusters of the product family and their co-correlated factors and linked factors
    • Factors In All categories / all factors from the Data Reference : all factors and their co-correlated factors and linked factors
  • List factors of the selected set of factors - left side : list of the factors of the selected sets and their co-correlated factors and linked factors, regrouped eg. Revenue OOB is linked to Missed Payment, etc. and co-correlated with OOB/ MRC ratio (see Deep Diving into Cluster for more details)
  • Factor evolution chart - right side : evolution over the last 12 months of the comparison between churners (cross-sellers, ...) and the benchmark population.

The co-corellation factors are calculated by the application as the co-correlation between the 2 data (see Deep Diving into Cluster).

The links are defined into the Data Reference (see Data Reference)

Factors In cluster

List of churn factors of rank 1 and 2

Rank 1: churn factors explaining the churn and shown on the left side

Rank 2: churn factors which do not define the cluster per se, but which have a strong a strong correlation of churn factors of rank 1

This approach allows users to further dig and have a broader view of potential actionable levers

Factors In Product Family

This is the list of all key churn factors explaining churn for the cluster and for the Marketing Product Family at large

If a churn factor appears here but not in the cluster, it means that this factor is probably important in defining another cluster of the same Marketing Product Family

Factors In All categories

Here are list ALL the churn factors

This long list allows marketing teams to deep dive and test any possible churn factor drawn from the data model