Understand the 'Branch ­Importance' Index

When analyzing member behaviors and branch activity, there are many factors to consider beyond branch assignment. A holistic approach requires looking at many things, including member behaviors, channel migration, profitability by channel, market growth potential and new member and new account trends.

CU Rise Analytics (now a part of Trellance) developed a proprietary statistical model, the Branch Importance Index, to make a holistic assessment of the role branches play within the credit union's delivery channels. It analyzes which channels members use to meet their multitude of financial services needs, when and how.

For example, do members prefer to open accounts in-branch, but then migrate primarily to digital channels? Or, is a weak digital presence inhibiting members, pushing them to other channels for transactions that could be more easily and conveniently completed online? Do even highly digitally-engaged members prefer to conduct more complex transactions in-branch, such as home loans?

The Branch Importance Index provides credit union leaders with a more detailed, quantifiable understanding of the way members use branches in the larger context. It's more likely, when looking at a single branch in isolation, that decision-makers could draw the wrong conclusions. They should instead be equipped with a complete understanding of how members interact with their branches in order to anticipate how each decision affects the larger whole.

Model Attrition Risk

One of the biggest threats when deciding to close a branch is the risk that members will leave the credit union. It's a source of stress for many leaders, and the answer isn't obvious without more deeply understanding members' behaviors and patterns.

Part of assessing branch importance should include studying and scoring the engagement behaviors of each member. This considers how many products they use and which ones, the frequency of engagement and how high digital usage is. Then, members can be segmented based on important shared characteristics.

Statistical models can use engagement information, coupled with historical attrition data, to more accurately predict the best, worst and most likely scenarios when it comes to possible attrition. Then, a credit union can work to achieve the best-case scenario by implementing a customized communication strategy tailored to each engagement segment.

Project Profitability Scenarios

With a greater appreciation of how members use branches along with possible attrition and retention scenarios, credit unions can employ models that estimate potential profitability outcomes.

Anticipating the impact on profitability involves modeling various scenarios that account for:

• Members relocating to a different branch; • Member adoption of different channels; • Various attrition/retention levels; and • Changes in member engagement score.

There are many factors to consider in a branch decision, and whenever possible, quantification can help credit union leaders grasp the facts and have better clarity. But even data scientists like us know that numbers aren't the entire story.

The fate of employees, history and community presence have a strong emotional component that must be weighed alongside the data. But considering them together creates a fuller picture and equips a credit union to make a balanced, informed decision.

Even when making difficult branch decisions, data insights make it more possible to provide exceptional service and ensure members feel valued. A member relationship can transcend location – and even deepen – when credit unions truly understand members' needs, find innovative and personalized ways to meet them, and most importantly, show they care.

Karan Bhalla

Floyd Salamino

Karan Bhalla (left), SVP and Floyd Salamino, Vice President of Analytics Trellance Tampa, Fla.