Temenos Talks logo

Beyond Black and White Decisioning

By Kris Frantzen 7 Nov 2016

If you follow professional sports, you've probably noticed the steadily growing influence of analytics in the way teams make roster-building, and even on-field, strategic decisions. Starting with the success of the "Moneyball" Oakland A's of the early 2000's, organizations across all sports have uncovered new ways to use data. And, the positions of power and influence in front offices are being increasingly filled by folks who know how to harness the data to set and execute strategy - as opposed to those who played, coached, or scouted the game.

In a similar manner, the influence of data has become prevalent in many businesses today. But it was the story of how the sports world has been ushered into the new age that came to mind as I attended sessions at the recent CUNA Lending Council Meeting in Las Vegas. In sessions regarding optimizing automation, consumer lending roundtables, and numerous exhibit hall and hallway conversations, CEOs and Chief Lending Officers shared their initiatives and challenges to employ the data available to them to make better and more efficient loan decisions. The plethora of data available on the applicant and past performance of similar loans can outline a valuable story - we just need to know how to read it.

Previously, whether or not a baseball player would succeed in the major leagues was exclusively determined by the trained eyes of scouts and a few rudimentary data points. Today, data analytics can reveal the relative value of a player - offensively and defensively - in virtually any scenario that may be encountered during a game. This sophistication allows for valuable nuance in decisions regarding which players to add to the team and which to utilize in certain in-game scenarios.

Consider those advances when thinking about the use of automated decisioning in the loan origination process. A loan decision in the past may have been determined by nothing more than a simple credit score and a few ratios. And, that decision would be a simple "pass/fail". You qualified or you did not. There was no examination of the factors contributing to a score, good or bad, or analysis of other points of data that could reveal more of the story. No nuance.

This week, I listened to financial institutions discuss how they've hired data scientists and carved out entire departments dedicated to analyzing data to optimize credit policy in order to efficiently add quality new loan volume to their portfolios and increase account holder satisfaction. We're moving beyond black-and-white decision-making, and utilizing the data to help us navigate into the shades of gray.

So, the relatively simple decision trees that have been employed by financial institutions are being replaced by much more complex combinations of loan eligibility and pricing rules. In turn, financial institutions are realizing the more rudimentary decisioning and pricing capabilities of legacy loan origination systems are not equipped to support their more sophisticated requirements going forward. In evaluating the best technology options, the term Decisioning as a Service (DaaS) has become more prevalent in conversation. The utilization of a decision engine separate from the origination system is aimed at ensuring sufficient support for increasingly complex guidelines. There are some key considerations and potential challenges in this approach.

  1. Attributes: To start, there are key product attributes that are pertinent to decisioning and pricing, but also to other components of the loan origination, such as application validation, document generation rules, and disbursement.
  2. Disparate Systems: Maintaining a decision engine integrated to (but separate from) the rest of the origination process often means needing to configure and maintain those product attributes in two systems.
  3. Implementing changes: The financial institution should also determine how easily it can implement change within the DaaS solution. Does the financial institution need to make a development request of the decision engine vendor for each required change?
  4. Reporting: Finally, the financial institution should determine if the separate components allow for the type of consolidated reporting it requires across its entire origination cycle.

A better alternative to a separate DaaS solution may be to utilize a loan origination system offering a powerful business rules engine capable of supporting a sufficient number of complex decisioning and pricing criteria, with functionality to allow business users to create and maintain the rules. Such a solution would allow for one system-of-record for maintenance of all product data and all origination data for data analysis and reporting. An open system with a robust application programming interface will allow the decisioning functionality, or another component of the origination system, to be integrated within a custom home banking or other front-end application. Your requirements are going to change and your market evolves - but they will only get more sophisticated and complex. Make sure your technology supports the complexity you'll need in order to keep ahead of your revenue and account holder service goals.

Our experts are on hand to help you

Request a call

Call me back