The energy from Day 1 continued as the Conference attendees were treated to autumn cornucopia of Big Data and Analytics information.
Bill Goedken, CEO of Idea5, presented on “Mining Gold: New Trends in Big Data”. Bill started by saying “seat of the pants” management no longer will be acceptable. In fact, regulators, due diligence, and competition will force financial institutions to change this.
Bill advised credit unions to start with smaller projects. In this way, he said, small “wins” can be the start of even bigger successes. Bill also gave several examples of how credit unions can use Big Data: fraud detection, assessing trends in delivery channels, making comparisons to peer groups, and predictive modeling.
The second session was “Personalized Marketing to Engage your Members”. The presenters were Chris Giles, CEO of CU Wireless; Ray Goodson, CEO of The Landmark Image; and Rahul Nawab, President and Founder, IQR Consulting, Inc.
Rahul talked a lot of using data to measure results of various credit union initiatives. He gave an example of data mining with the aim of increasing incremental credit card spend. He got the audience laughing at the end of his talk with the quip, “If you torture data long enough it will confess.”
Chris spoke about the CU Wireless program called Buy Local. He said the credit unions of 10 years ago would have amazed to know that in 2015 there would be serious competitors (e.g. - Apple, Moven, and Simple) that have no branches. Credit unions need to respond to stay relevant, increase value, and use technology. He encouraged the attendees to use data analytics to “maximize our newest asset”. This includes aggregating and managing data to deliver value.
Ray Goodson talked about “touchpoints” for marketing to members that build brand, improve member retention, and reinforce the relationship with the member. Ray showed a variety of examples for printing projects and noted that customer experience was the key differentiator.
The third session was “Predictive Analytics for Retail Lending” presented by Dr. Joe Breeden, CEO of Prescient Models, LLC. Dr. Breeden, an expert in predictive analytics for financial services. Dr. Breeden noted some important points about data analysis in retail lending:
- Input data is highly regulated with the same inputs available to all lenders.
- Artificial Intelligence-based results would never pass examiner review or validation.
- Any model output eventually rolls up to a finance report involving even more layers of regulation and review.
Dr. Breeden said he believed Apple and Google ultimately won’t compete with credit unions due to the regulatory burden. In fact, it is that very regulatory burden that will in part drive credit unions to adopt Big Data and Analytics tools. His research shows that Big Data and Analytics spends today are already focusing on regulatory compliance. He noted that the Fed, OCC, FDIC, and NCUA have all put forecasting and stress testing (and thereby, predictive analytics) on center stage.
Dr. Breeden said pricing is where stress-testing will go first. With regard to pricing, he said:
- Pricing is often meet-the-market or moving average. In 2005, mortgage lenders were pricing based on 2003 originations. The pricing model most likely to “break the bank” is “meet the market”, but any approach can be done badly.
- Pricing often ignores loss timing. New loans always low risk, but don’t stay that way.
- Pricing models rarely consider the future environment or even the current or average environment.
- Financial institution failures are pricing failures first.
In closing, Dr. Breeden advised:
- Build predictive models first for pricing. These will be a fraction of the cost of regulatory compliance models.
- Pooled repositories across CUs can bridge the data gap.
- The first institution with a forward looking pricing model was a credit union.
Session 3 had two breakout meetings. One was “Self-Service Analytics for Credit Unions”. The panelists were Keith Bennett, SVP of Information Technology, Denali Alaska Federal Credit Union; Jerry Shaul, BI Architect from SEFCU in Albany, New York; Owen Mibus, Senior Data Architect, OnApproach; and Lorrie West, Engagement Manager for OnApproach.
Keith and Jerry both gave the stories of their journeys toward self-service. They both agreed their senior managers were bullish on analytics but creating a system where users could do self-service is still a work in progress.
Lorrie and Owen then demonstrated OnApproach’s QuickConnect – Lending. This is a “semantic layer” which, as a part of OnApproach M360 product, allows many visualization tools to portray the data in the data warehouse.
The other breakout meeting was “How will Big Data & Analytics Impact Retail Lending” The panelists were Courtney Collier, Product Manager, CU Direct; Dale Fosselman, SVP Corporate Development, Denali Alaska Federal Credit Union; and Thomas Kammer, President, ValuCheck.
Tom started off with a discussion about the importance of data quality in analytics. Credit unions need to pay attention to the quality of their data because it is the very foundation for the accuracy of analytics. Tom gave the example of the accuracy of home addresses and how they impact the accuracy of collateral values.
Courtney Collier talked about disruption in the lending industry and how competition is making the process of getting a loan much easier. Unfortunately, many credit unions are not enabling auto decisioning on loans which results in lost opportunity. Courtney felt that if more attention were paid to the decision criteria for automated decisioning, it would be more loans for the credit unions and a much improved experience for the borrower.
Dale Fosselman talked about using Big Data X Big Math to create a new way to forecast Allowance for Loan Loss (ALLL) that is forward looking vs backward looking (which is the state of the art today). Denali FCU has been testing a forward looking ALLL forecasting model that has the ability to predict the probability of loss at the loan level, using a number of economic factors which include, employment rate, house pricing index and loan vintage. This is cutting edge technology that is also allowing them to look differently at how they score the credit risk of a member, where the FICO score becomes an input not the sole determinant of the risk score. There was consensus among the presenters that the lending market is changing rapidly and those changes are being driven by the use of analytics, the need for simplicity, speed, accuracy and convenience. Credit unions have an advantage but it will be short lived if they are not able to up their game!
The fourth session was “Innovative Analytic Applications for Credit Unions”, a showcase of products that use data from OnApproach M360. The panelists were Ken Burns, EVP Sales and Development at Intuvo; Matt Davis, Founder of GameFI; and Keith Kelly, CEO and Co-Founder of Rate Reset.
Ken Burns described Intuvo, a software package that comes pre-loaded with marketing campaigns that can be customized by the credit union. The product can be deployed in 30-60 days, generate a 20-70% boost in close ratios and a 20 point boost in Net Promoter Scores. It features simplified compliance, and improved LO productivity.
Matt David of GameFI described how using games can increase member and user engagement. He mentioned that employee engagement was a sweet spot. GameFI leverages game mechanics to encourage deeper employee engagement. One tool was Leader Boards which show how all performers doing. Also there are challenges employees can issue to their co-workers to spur performance.
Keith Kelly spoke about how Rate Reset uses existing loan data to offer members to extend loan terms and thereby increase interest yield. In programs at actual credit unions, member monthly payment decreased 30% and member satisfaction increased. Keith noted that there is a risk for early payoff and losing not only interest income but also offer the member an opportunity to create a relationship with another lender. Rate Reset lets members feel they are in control when in fact the CU is in control by limiting the members to whom the offer is extended.
The final session of the day the Credit Union Analytics Best Practices Competition. The finalists for the top prize were:
Ideal Credit Union (and the winner) - VIP Member Rewards Program - Ideal was looking for ways to increase into “share of wallet” by rewarding members for transaction and product growth across a variety of areas known as the 4 “C”s – car, credit card, casa (house), and checking. It uses 64 different business rules are applied across three separate subject areas within a data warehouse. Ideal has paid more than $1.5 million in member rewards as a result of this program.
Affinity Plus Federal Credit Union - Channel Analytics – This report identifies the channel in which a transaction occurred. The report helped APFCU reallocate $3M from physical branch expansion into improved Remote Deposit Capture capability.
Dupaco Credit Union – Greenback Impact – The purpose of this program is to quantify member cost savings as the result of moving their business to the credit union. The program quantified over $3M in interest saved 2 years in a row for members refinancing loans.
TopLine Credit Union – DailyPulse Report – This is a branch reporting provide easy access to production data reflecting recent, current and goal amounts. It replaces existing manual input spreadsheets and features nightly updates, self-service capability, and drill down.
For pricing information to budget for the 2016 conference, please reference the pricing table and the “more information” section on the 2015 webpage here, or contact Austin Wentzlaff at email@example.com or (320) 444-0291.