By David Ross
I once received an escalation call from a member whose card had been declined for fraud prevention while checking into a Paris hotel. She argued that we should have known it was her using the card because she had purchased tickets on Air France two months prior using that same card. We should have been able to predict the airline and hotel transactions were related. That was her expectation in 2008, and expectations have only continued to grow in the decade since.
After years of partnering with credit unions across the country, my team and I have found the best approach to knowing and serving members and their expectations is through the use of Predictive Analytics.
Predictive Analytics can help credit unions determine future probability. This is different from traditional reporting – or Descriptive Analytics – which focuses on what has already transpired. To be clear, a credit union needs solid Descriptive Analytics before moving into the predictive space. It is crucial to understand and regularly review portfolio and transaction reports and dashboards to evaluate KPIs, trends and other key metrics.
Digesting these items is important not just to measure how well the portfolio has been managed but also to identify potential areas of opportunity. While such analysis is valuable, it is not predictive in nature and cannot assist in projections.
Predictive Analytics Can Forecast Outcomes
Predictive Analytics is built on statistical data modeling, and the best-known predictive model is probably the credit score. Just as the credit score can assess the probability that a loan will be repaid, other predictive models are built for a variety of use cases.
For instance, predictive models can be built to identify members who are likely to attrite or become detractors, allowing the credit union to intercede before there is any impact to retention rate or net promoter score (NPS). These are two powerful examples that leverage data to know the member better, which is imperative in order to meet today’s expectations.
Cost and Complexity Are Obstacles
Transitioning to Predictive Analytics can be challenging. To begin, detailed granular-level information is required. Even if a credit union has captured this data, it still needs to deploy a statistical modeling platform with enough capacity and computational power to work with said data. Then, a human resource familiar with the techniques required for predictive analysis is needed. This could mean hiring a data scientist – a position that currently commands a six-figure salary and is in high demand and short supply. Clearly, significant resources and costs are involved, making the path to Predictive Analytics somewhat difficult to traverse.
Partnering with an Expert Can Be Efficient
A credit union could seek out a vendor or credit union service organization (CUSO) to assist with the transition. This is a very practical approach but requires credit unions to carefully evaluate potential partners. Important aspects for consideration include how the partner will get the data necessary to build the models, how frequently data needs to be transmitted if it must be sent outside the credit union, and whether PCI or PII compliance is in place.
One must also understand the actual capabilities the partner can achieve – are they providing reports (Descriptive) or are they building models (Predictive)? You must also determine what happens with the final output. Would the credit union simply receive a list of members likely to close their accounts? Would it be a scored list? What are next steps? Taking this proactive approach will help ensure the selected partner is best aligned to meet the credit union’s needs.
Once a credit union fully utilizes Predictive Analytics, it will have the ability to provide a level of service that exceeds even today’s high expectations. Members want to feel as if their credit union truly knows them at the individual level and with Predictive Analytics, credit unions can accomplish just that.
David Ross is VP, Advisors Plus Predictive Analytics at PSCU and has spent the past 19 years working in a wide variety of data science and analytic roles. Most recently, he was responsible for global fraud analytics strategy and execution at Citi, a position that required working with countries around the world to implement best-in-class analytics tools. David is currently leading the initiative to develop models for Predictive Analytics at PSCU.