How To Drive Results From Your Credit Union Data Analytics

Data analytics is a critical aspect in managing an organization and making better decisions. The current technology has opened up a wide array of possibilities, but how should a company actually use data analytics? In a recent podcast, John Best addresses some of the practical uses for data analytics and gives real life examples of how you can use the technology to enhance your organization from the board room to the call centers.

Foundation of Data Analytics

The topic of optimizing your data to make better decisions starts with the assumption that you have already built a solid foundation. An insufficient or cumbersome data system will cripple your efforts to build a data driven culture with successful analytics strategies. This section addresses a few key aspects of a solid data analytics system. However, if you find that your current system is lacking in some of these areas, keep reading! There is still plenty of valuable information for you to consider in terms of using data analytics well.

The first pillar of a well-built data analytics system is corralling the data. If you have different systems for your mortgage data, credit card data, accounting data, etc, then those have to be moved onto a platform that allows you to access the data all in one spot. Ideally, you have a data warehouse for this and as data is continuously moved into the warehouse, it is systematically organized. That means that the data is identified, cleaned, sorted, and normalized. For example, if you have one system that uses the term “payment date” and another system that uses the term “due date”, those terms have to be normalized so that the analytics system recognizes those terms as synonymous in collecting data and building reports.

Second, you must have access to tools that allow to you sort and visualize the data. And it must be tools that someone in your organization actually knows how to use. The most powerful data gathering systems are useless without software that helps companies understand and track the data. The number one tool that comes to mind when I think of business intelligence software is Tableau. It facilitates your ability to ask questions and answer them with data. The software is intuitive, powerful, and creates beautiful charts and graphs to help with visualization.

The next pillar is data governance. This is important because you don’t want to get behind as people add new systems. A case in point is our FIVE system (Financial Interactive Voice Exchange). It is currently able to link with the Amazon ECHO so that users can interact with their financial institutions. It produces all kinds of data points, such as what kinds of requests people are making, what time of day they are making the request, etc. That is the perfect system to run through data governance. In an ideal world, it would be run through data governance before being released in order to make sure that the data it collects is normalized with the main data warehouse. That way, the new data can be collected and added to the overall system smoothly. In other words, use a data governance system that accommodates new data without cluttering up the whole system. Make sure that the room you just cleaned up stays clean.

Finally, you must trust your data. This is the most important part of a data analytics foundation. There is absolutely no point in analyzing data that you do not trust. Not only is trust the most critical aspect of the foundation, it is also the most delicate. If something is off and someone notices, not only will they distrust your report, they will distrust the entire data analytics system that produced the report. If you do not have a trustworthy data foundation, you will simply waste your time trying to build a data driven culture. It is essential.

Strategy for Data

Once you have a platform for managing data, it’s time to build a strategy. What are you going to do with it? How are you going to collect it? What are your key growth strategies? If your key growth strategy is to grow organically by putting up billboards, what are the indicators for the success or failure of that strategy? How are you going to measure your growth? Who is measuring growth? What is your plan if the indicators show an increase or decrease?

The answers will be unique to various businesses and organizations. The key is to ask—and to build a team of people that asks—questions that help clarify goals and methods based on actual data.

You also need to consider what kind of data you are willing to collect. You need to ask what data is in bounds. Protect your data and the data of your clients. A credit union is loaded with data that would be fascinating to analysts looking for purchase trends. Take care to ensure that you collect only legitimate data and then store it properly and securely.

Challenge Conventional Wisdom

Take a cue from Mark Cuban, investor on the television series Shark Tank, who quipped, “My businesses are usually built around challenging conventional wisdom, so I tend to gain by taking the other side. It’s been very profitable and entertaining for me”

What does it mean to challenge conventional wisdom? In terms of data analytics, it means having a healthy cynicism for any claim that is not backed by data. Perhaps this is best summarized by the email signature of an Amazon data employee, “In God I trust, everyone else bring data.”

Here are several examples of conventional wisdom that may be inconsistent with data. Consider these examples and learn to spot the ways in which false assumptions have crept into your decision making models.

The first assumption is the capture business model. It is very common in the industry to hear that bill pay is a sticky feature. The thought is that if someone takes time to set up 10 or 15 payees in a bill pay system, they will stay with you simply because they don’t want to go through the trouble of switching systems and re-entering data. Even if another company offers a better service, they will stay with your system to avoid the hassle of transfer. That assertion needs to be challenged. The world has changed and there are tools that export data to a new system with ease. It is no longer enough to assume that you have captured the business of customers who have invested the time to enter payees in your system.

Another assumption is that digital transactions are cheaper than humans. Digital systems are hailed as the best option for handling nitpicky transactions like password requests that come into call centers. After all, it might cost $2.00 for a person to handle those transactions and $0.03 to run a digital answering system. But have all of the factors been considered? What about compliance to the current regulatory environment, or the work it takes to connect to all those systems? Is there a negative impact from removing human interaction? A person might notice that the client has almost paid off their car and proceed to ask the client about plans for getting a new one; a digital system will simply give the password. These are questions that must be asked of your data analytics before rushing to convert to digital.

What about the common belief that all mobile members are all millennials? Frankly, this is one that has cursed us. Your mom’s favorite feature might be taking a picture of a check and depositing it because she doesn’t want to have to go to the bank. What happens if you apply that logic to the Amazon ECHO? Don’t compartmentalize by demographics.

How about the common estimate that only 25% of members use branches. Really? Check your data. It’s different strokes for different folks and you may find out that your branches have significantly different numbers.

These days, seeing someone paying for groceries with a checkbook is like seeing a unicorn. It’s clear that the use of checks is diminishing, but what is the actual reduction? How has PayPal hurt your ATM transactions? Is it even feasible for a credit union to become a client’s primary financial institution? Ask your data, ask your data, ask your data.

Build a Data Driven Culture

Foster a business culture where no one accepts a raw answer. A data driven culture is one where everyone understands that knowing the exact numbers is crucial to improving accountability and scenario planning. Grow a team that is hungry for data. It is natural to work in a reactive way, simply responding to circumstances as they arise. Placing a high value on data allows companies to be proactive.

Use data, not a finger in the air, to determine deadlines. Ask how long the last 10 projects took and don’t naively assume that the eleventh will take half the time. Learning to ask the right questions and elevating the role of data will take time. Build from the top down and continue to foster this kind of culture until one day, a question will come up in a board meeting and your team members will be content to wait for the hard numbers. Or better yet, one of them will pull out their tablet and use Tableau right there.

Research from the Aberdeen group shows that data driven organizations saw a year over year increase in revenue of 27%, whereas other organizations increased by only 7%. Get the real answers and build a data driven culture.

Prescriptive Analytics and its Uses

There are three kinds of data analytics: historical, descriptive, and predictive. Historical looks at how many people went to the branch on what day at what time. Predictive asks how many people we think will go. Prescriptive considers how we could influence the number of people that go to the branch.

One of the big considerations for credit unions is the current expected credit loss or CECL. In 2020, you are expected to go from the descriptive model of measuring incurred loss, to a predictive model of reporting expected loss. Many companies are hunting around for a one trick pony that will fulfill this new model. However, any data analytics system worth its salt should be able to predict expected loss, and it’s absurd to apply this kind of predictive analytics to only one area. Why not use these features to predict other areas like expected credit gain? It’s valuable to look at where regulatory folks are driving us and make sure that there aren’t other uses for these mandated technologies.

Uses For Predictive and Prescriptive Analytics

Detecting fraud is one of the most beneficial uses for predictive analytics. Years ago, some consultants for a company came across a number of Netflix payments for $40. Now, the premium subscription for Netflix is only $13.99/month, so $40 is clearly an outlier. It turns out that someone was trying to hide transactions by disguising them as Netflix payments, but the numbers didn’t match. Using a system that is able to predict the normal pattern of transactions and then monitor for outliers is a huge asset in detecting fraud.

As data analytics systems become more and more powerful, you can them to detect red flags. If more than 15% of your population is being subjected to multifactor authentication, something’s wrong and you should have a data platform that recognizes that. Other red flags include online ACH debits that are more than twice as much as the credits, a digital active users number that is less than a third of your total customers, password failure rates, or a sudden high enrollment online.

You can also use machine learning algorithms to make sure that people are properly adhering to a separation of duties and that employee transactions within the credit union are not purchases for their families and vice versa.

These red flags don’t necessarily indicate fraud. It may be a problem with the website, or a feature that has been overlooked. Either way, using predictive analytics gives your organization another set of watchful eyes.

Regulatory compliance presents another opportunity for predictive data. Years ago, our company set up recurring payments online so that members could see, make changes, or cancel payments. Later on during an audit, it was discovered that some loans had an agreement that if the auto pay was shut off, the rate went up. We had no idea that some of the loan rates were tied to the payment method, and the problem had to be fixed manually. Programming your analytics platform to keep track of things like that will save you a lot of trouble.

Commercial lending is also full of data analytics possibilities, particularly with regard to evaluating businesses and traffic. You can also use data to make decisions regarding fair lending, or mortgage compliance with Freddy Mac and Fanny Mae. The bottom line is that predictive analytics is the future of data analysis and your business will do well to implement it broadly.

Influencer Model

Let’s shift gears and talk about another aspect of using data—driving revenue by tracking influencers. This is where data analytics enters the prescriptive realm of influencing behavior. Alpharank is a company founded by Brian Ley as a solution for influencing demographics. Ley noticed a social phenomenon that went like this: if you call a friend and ask if they want to join you at a club, they say they’ll be there, but only if Martin comes. As you continue to dial phone numbers, you keep getting the same answer until it becomes clear that all you really needed to do was invite Martin and everyone else would show up without even being asked. Martin is the influencer, and the same kind of strategy works for credit unions. Alpharank is a company that offers a method of using data analytics to track your influencers so that they can be leveraged to spread positive activity. If you have a way to track your influencers, you can show them your latest product, offer them special deals, or simply give them a heads up about a new feature. In turn, they will influence others, and this is a great way to reduce money in advertising.

Collaboration is the Best Defense

The value of a data report is directly tied to the amount of data that it represents. Unfortunately, as credit unions, all of our data from a scale perspective is small. Think about the data of a credit union compared to the data available to Bank of America. There is no way to compete with the amount of data that they can analyze based on their hundreds of accounts. From a data standpoint, it would be easy for credit unions to be marginalized. The best defense against this is to collaborate and share data. Now of course, protecting sensitive client information is paramount, but there are ways to share anonymized data that allows for more information and a better report without compromising security.

Use Data Analytics to Respond in an Innovative Way

If a member of your credit union who used to purchase all of their airline travel on your credit card suddenly stopped, would anyone call them? Probably not. This scenario illustrates the opportunity to use powerful data analytics to interact with members in an innovative and personalized way. If a call came in about a 10 day payoff on a 0% car with a 36 month loan, you would be all over it, afraid of losing that loan; however, the revenue lost by the airline customer is probably much more significant than the lost loan.

Here’s another example: if a member’s direct deposit suddenly stops, what does that mean? Perhaps they changed jobs. In that case, a simple phone call might be all that is needed to get the deposit back or to offer them help with any financial issues. Perhaps they are fed up and wanting to leave. Again, a phone call would be an important step in finding out why they are leaving. What if the deposit stopped because of a processing error on the part of their employer? Wouldn’t it be great to offer the kind of customer service that would call that member to report the anomaly and offer to check with their employer? If the halted deposits are due to a lost job, a phone call can be the first step in helping to reorganize their loans and get them back on their feet. As someone once pointed out, perhaps the deposits stopped because they are deceased. Does your credit union have a policy for reaching out to the families of deceased members? If the answer is no, why not? These are the kind of innovative considerations that help companies excel in caring for their members and minimizing loss.

Summary

So what is your data strategy? Is it a strategy to drive sales? One to enhance a knowledge base? A strategy to determine the best place for your next branch? Data analytics is an essential tool to develop and monitor these strategies.

Challenge conventional wisdom and build a culture that prioritizes facts and exact numbers. Otherwise, you are simply judging by the wind.

Collaborate with other data sources to build better reports and recognize the future of predictive analytics so that you can begin implementing its features in multiple aspects of your organization.

And finally, it’s not enough to have the data. Let the possibilities of the new technology reshape how you make decisions and interact with your members.

Article written by John Best – CEO, Best Innovation Group

On our recent podcast Episode #231 Branded Currency: John talks his new book Breaking Digital Gridlock available on Amazon

A version of this article first appeared on Breaking Banks.

About the Author

Brett King is a widely recognised top 5 FinTech influencer. He is a futurist, an Amazon bestselling author, an award winning speaker, hosts a globally recognized radio show (Breaking Banks), is the CEO of Moven, and in his spare time enjoys flying as an IFR pilot, scuba diving, motor racing, gaming (mostly FPS) and Sci-Fi. He advised the Obama administration on the Future of Banking, and has spoken on the future in 50 countries in the last 3 years.

Breaking Banks, #1 show on VoiceAmerica Business, is the leading global fintech podcast with more than 5.5 million listens from 172 countries. Breaking Banks broadcasts, are live every Thursday at 3pm EST in NYC on 1160AM WVNJ Radio and globally via VoiceAmerica’s Business Channel. 

His latest book Bank 4.0: Banking everywhere, never at a bank will be shortly released on Amazon.