The next big idea for the credit union industry

with: Peter Keers, PMP and Engagement Manager at OnApproach

The credit union industry is on the cusp of significant challenges with the potential to disrupt the financial services landscape as we know it. Big Data and Analytics is driving a new breed of competitor into what has been a very traditional marketplace. The industry will need to envision and build out the “Next Big Idea” for credit unions to stay competitive and successfully navigate the next 10 years.

Today, credit unions are seriously lagging other industries (e.g. – retail) when it comes to the use of Big Data and Analytics to better serve members. The industry urgently needs to find a means of becoming proficient in Big Data/Analytics or face the threat of being disrupted in the same way that Borders Book Stores, Blockbuster, and most recently, the taxi industry (currently under attack by UBER) have all been disrupted.

One reason for the lag has been the lack of an industry standard solution capable of integrating transaction level data from the many disparate data sources used daily by credit unions. Enterprise-level data integration is a prerequisite for establishing a viable Big Data/Analytics capability.

In an attempt to address the data integration issue, credit unions have devised a number of tactical approaches:

  1. The number one “data integration” tool for over 95% of credit unions is the Excel spreadsheet. While convenient, this alternative is labor intensive, error-prone, and by no means an enterprise-level tool. In using Excel, it is common that credit union managers spend 80% of their time on report preparation and only 20% on the actual analysis needed for decision-making.
  1. Often, a credit union will purchase multiple additional modules from their core vendor in the hope of achieving some measure of data integration. This strategy yields limited results since most credit unions use software from many vendors.
  1. There are many vendors with “analytic” solutions focused on specific areas such as loan portfolio analysis or teller efficiency. These tools usually do little, if anything, to integrate all of the data a credit union needs to generate analytics that are broadly meaningful and relevant. These vendor products are simply not designed to be enterprise-level data integration solutions.

If these approaches are insufficient, what is the alternative for credit unions interested in developing a Big Data/Analytics capability?

Filene research (www.filene.org) recently conducted a study focusing on the increasing importance of Big Data when dealing with credit union members.

“Companies as varied as Amazon, Google, Walmart, and Wells Fargo are turning to “big data” for customer insights that will help them serve clients and capture market share. Big data is the analysis of huge data sets, and while individual credit unions may not have the resources of a corporate giant, advances in data storage and software tools mean that credit unions can start using similar tools and deriving similar value.”

“A prerequisite for developing these (predictive) and other models is a well-maintained database with as much transactional detail as possible. The credit unions that can capture transaction types and locations will come out ahead, because transaction origin correlates highly with credit scores and helps to predict future financial products.”

A few credit unions have tried to build this capability themselves but have found it to be much more complex, expensive, and time consuming than originally thought. The five year cost of building and maintaining a data warehouse and analytics infrastructure is about $2.0 million dollars. Most credit unions lack both the expertise and the budgetary appetite required to build and maintain a “home grown” Big Data/Analytics solution.

Yet, the seemingly overwhelming task to build a Big Data/Analytics capability is not beyond the reach of the average credit union. The “Next Big Idea” will leverage the special collaborative culture of the credit union movement to manage the huge investment of time and cost.

Analytic Data Model

Imagine an industry standard “Analytic Data Model” (ADM) connecting to any vendor system and integrating its data at the transaction level. In effect the ADM becomes a universal translator for systems from multiple vendors covering multiple subject areas.

The ADM accomplishes this by defining a common data standard for credit union Big Data/Analytics. For example, every loan system has an ID field uniquely identifying the loan instrument. Some systems call this Account ID while some call it Loan Number. The ADM maps both into a common standard data definition called “Loan ID”.

The ADM could be installed “in-house” or “hosted” depending on the preference of the credit union. With this industry standard ADM in place, credit unions would be able to achieve significant economies of scale in several ways:

  1. A Single Source of Truth – The ADM provides credit unions with a single source of truth through the automated integration of transaction level data from disparate systems. This avoids the problem of the same metrics reporting different values because the data was sourced from separate, “siloed” systems.
  1. Analytic (Business Intelligence) Tool Agnostic – Analysis is greatly complicated if the underlying data is scattered among various sources with propriety names and formats. Very expensive analytics tools themselves sometimes have data integration functionality. However, it is not unusual that each new analysis requires a unique data integration. For tools without data integration functionality, the analyst must either manually integrate the various sources or skip integration entirely. The ADM provides a complete, consistent, and easy-to-access platform that allows the analyst to focus on performing analysis.
  1. Interchangeable Applications – In today’s world of reporting and analytics report developers spend many hours creating hundreds of reports that overlap with those created by other report developers in the same industry (e.g. – Loan Application Analysis).With the ADM in place, the collaborative culture of credit unions would lead to the creation of applications, reports, and dashboards that can be accessible by others in the industry. Furthermore, industry vendors could also be invited to write applications that connect to the ADM for any credit union could use. This would also allow the creation of an “Application Market” in which applications could be made available for sale across the industry.
  1. Predictive Analytics – Predictive analytics are “forward-looking” analyses and reports that can be used for predicting next best product, basket analysis, personalized marketing, more competitively priced loans, and much more.Unfortunately, no single credit union (with the exception of Navy) has enough data to effectively execute predictive analytics. Furthermore, the cost of a data scientist and the time and cost required to build predictive models is beyond the reach of most credit unions.The ADM enables collaborative data pools in which credit unions can contribute enough data to increase the accuracy of the models and get back information that can be used to improve their internal business processes.

The creation of the Analytic Data Model will likely be the work of another collaborative innovation of the credit union movement, Credit Union Service Organizations (CUSOs). Wikipedia describes a CUSO as organizations that, “Provide avenues for innovation and creativity that would not typically occur within the confines of a credit union”.

The CUSO model is an ideal way for the development of the ADM to occur. By bringing together the best collaborative energies of the credit union movement, the critical mass necessary to drive Big Data/Analytics initiatives can be achieved. Luckily, Big Data/Analytics CUSOs are already starting to form and are well on their way to transforming credit unions into Analytic Competitors positioned to win in the next 10 years.

Paul Ablack

Paul Ablack

Paul started OnApproach to specialize in leveraging data to support better decision making and drive profitability at various organizations. Prior to starting OnApproach, Paul was Vice President/General Manager of ... Web: www.onapproach.com Details