Determining the most profitable households and segments for a positive impact
A major bank in the United States
Relationship management programs are meant to develop and sustain customer relationships. These programs are a crucial element in banks retaining customers and cross-selling. It is important to execute programs that rationalize costs and reap strong returns. Targeting the right customers can significantly elevate the performance of such programs.
The client partnered with EXL to elevate their relationship management program’s performance. Knowing that the set of customers chosen for directing such a program decides its fate, the bank wanted to target customers with the highest propensity for cross-selling. They knew they couldn’t rely on gut instincts or simple metrics to choose the most profitable target segment, making a detailed customer data analysis necessary.
A two-stage approach was determined.
- A comprehensive performance analysis was done for households that were previously part of some prominent campaign
- Selection of households was based on performance analysis insights Performance Analysis Objectives
- Understand cross-sell performance of households
- Evaluate selection strategies using segmentation approaches
- Drive recommendations for building next year’s portfolio
It is important to choose the right target segment by selecting the correct metrics and plugging them into a well-devised strategy for maximum impact. Using the right blend of metrics with guidance from analytics experts, banks can obtain their desired results from relationship management programs and eventually increase their cross-sell rate.
- 58% high profitability households selected
- 86 basis points of incremental cross-sell rate generated
- $1.17M of annual forecasted contribution due to cross-sell in four months
The dollar profitability band in which a household falls – high, medium, low – affects the cross-sell rate; the higher the profitability, the higher the purchase rate
The more often the bank contacted a household, the better its performance
More branch-engaged households show higher cross-sell rates Using these variables, EXL built a strategy to select households for the program. The purchase probability of households was also computed.
Target Selection Process
To incorporate the findings of previous year’s portfolio performance analysis, the following measures were taken:
- Customers with higher profitability were preferentially allocated to the portfolio
- Least contacted households across campaigns were excluded
- Customers with the highest number of branch transactions were preferred.
Households within each group were ranked using a three-step process:
Step 1: Grouping within profitability bands
Step 2: Ranking based on propensity and branch transaction ranking. Households were ranked within different profitability groups separately
- Propensity rank: The higher the purchase probability, the better the rank
- Branch transaction rank: The higher the number of transactions, the better the rank.
- Composite ranking: Sum of propensity rank and branch transaction rank
- Final ranking: After obtaining composite ranking, the groups were ranked as per their profitability
Step 3: Assigning households to the portfolio based on their final rankings After selecting households for the portfolio, a control group was formed by extracting a sample from the selected group. The control group had similar composition as the selected group and was isolated from any sort of contact with relationship bankers.