Data is a key differentiating factor for all organizations. Knowing where your data is unique, and where there is sufficient data to draw distinction is critical to moving data from dead weight to working asset.
PrecisionLender Chief Analytics Officer George Neal discusses his approach to data and how it can be utilized depending on the area of emphasis. Additionally, he shares how a new mindset around and treatment of data can lead to a change in organizational behavior.
Maria Abbe: Hi. Welcome to The Purposeful Banker podcast, the podcast brought to you by PrecisionLender, where we discuss the big topics on the minds of today’s best bankers. I’m your host, Maria Abbe, Content Manager at PrecisionLender, and today I’m joined by George Neal. He’s our Chief Analytics Officer at PrecisionLender. Thank you all for joining us.
Today we’re talking about data. It runs through every bank institution serving as a key differentiating factor, which means it’s pretty important. It’s time we give it the attention it deserves. That’s why we brought in George to help us tackle the topic. Before we get started, George, do you mind introducing yourself and then telling us a little bit about your role here at PrecisionLender?
George Neal: Certainly, and thanks for the brief and nice introduction there, Maria.
Maria Abbe: Of course.
George Neal: My name is George Neal. I am the Chief Analytics Officer here at PrecisionLender. I’m also a 20-year banker. My role here is actually quite simple. The task that I’ve been given and the task that my team has is to derive value from the data here at PrecisionLender.
Maria Abbe: Now, George, when you say derive value from data, what exactly do you mean, because that seems like something every organization does, just as a course of doing business.
George Neal: It’s a good question, Maria, because while every organization believes they get value from their data, in many cases, all they get is after the fact information. When we’re talking about things like reports, dashboards, etc., they present after the fact information, but they don’t effectively change behavior. Here at PrecisionLender, when we talk about deriving value from data, we measure that in terms of changing behavior. We view it as valuable information, and valuable data changes the behavior at the customer interaction. It changes the behavior in the back of the bank. It changes behavior wherever it is found, and that’s the definition of value.
Maria Abbe: Okay. How can we think about and treat our data so that it actually does change behavior?
George Neal: I think there’s a couple of tricks to that, the first of which is you have to know where your data is unique and valuable, whether it’s the fact that you have unique insights into your market and unique experience. In the case of a bank I previously worked with, we had a very unique market presence. We had access to customers that other people did not have access to, so we were able to understand their behavior. Here at PrecisionLender, we actually see transactions happening across a wide variety of clients so that we can see how different approaches have different results, and we can make recommendations as a subsequent result of that.
In order to do so here at PrecisionLender, we’ve adopted the jobs-to-be-done framework. Oftentimes, it’s easier to ask the question of yourself, “What do I want my data to do?” To think about your data as an actual living team member and say, “If my data were a team member, what would be their job? Is my data even equipped and able to do that?”
Maria Abbe: If an organization finds that the data isn’t equipped to do that job, what does that mean, and then what needs to be done?
George Neal: Well, in truth, it’s hard to fire your data, and you can’t easily send your data to a training class, so in many cases, the analogy breaks down a little bit there. What you do need to do is treat the data as a core asset. We have to treat the processes that maintain and support our data as every bit as important as the processes which create them. In reality, if you’ve got a situation where you’re treating your data as a second class citizen, you’re going to have poor data quality. You’re going to have a poor team member executing that role, whatever you’ve assigned to your data. You’ve got to spend some effort and some time in making certain that it’s up to quality and up to the standards that you need for it to do the job you’ve assigned it.
Maria Abbe: Do you have an example that you might be able to share with us?
George Neal: Here at PrecisionLender, we have an artificial intelligence that we’ve named Andi. Andi gets in a lot of information. She takes in information from the application, from the utterances of people asking her questions, and from all of the data that we’ve seen across the relationships that we have insight into, thanks to our client base. All the information that Andi has isn’t always correct. When she’s actually making a decision, we have built-in processes that monitor what goes in, what comes out, and the decisions that she’s made. When we see that something isn’t correct, we can immediately take action to improve the data quality behind it, to improve the response, or any other thing that’s required to make that data value function the best it can possibly be.
Maria Abbe: Wow. To the average bystander like myself, she sounds awesome, but it sounds like a lot of work.
George Neal: It sounds that way, but it doesn’t have to be. The trick here is that it needs to be integrated into the right places. What I’ve seen in a number of organizations, is that when you detach the support, maintenance, and quality control of your data from the people that are actually benefiting from the use of the data, it becomes much harder.
Maria Abbe: Interesting. What’s the payoff for that?
George Neal: It depends a lot on your organization, and data payoffs differ depending on where, of course, you put the effort. You could find new market segments. You could enhance markets. You could reduce risk. There’s any of a number of things that could be done. We have a couple that we use here through Andi that make good examples.
We have a skill that predicts utilization on line of credit. When you go in to price a line of credit, Andi can predict what that utilization is going to be. She’s only able to do that because we have so many observations of lines of credit that we’re able to do a reasonable job of predicting based on this customer with this type of relationship, this type of line of credit, this is the likely utilization. That allows for better capital planning. It allows for better assessment of profitability, and overall, it improves the position of the bank. That’s just one example, and that’s here at PrecisionLender through our AI. I’ve seen numerous examples in marketing, operations, etc. where again, you see huge value, but it’s driven by the ability to change behavior.
Maria Abbe: Interesting. Now, you’ve mentioned Andi a couple of times. Just for the rest of us to have an understanding, how much data goes into Andi?
George Neal: The short answer would be a lot. It depends greatly on which aspect of Andi you’re talking about. In this case, Andi has access to our entire support website for looking for answers to questions that people pose. Her various skills have various levels of data that go into them. I mentioned the line of credit utilization skill, so there’s roughly 100,000 lines of credit that went into constructing that skill. Again, quite a bit.
Maria Abbe: Yeah. Going back to that question I asked earlier, it seems like a lot of work, but does an organization really need that much data in order to make an impact?
George Neal: No, because not everything is a machine learning or an AI problem. It just so happens that that’s where my passion sits, and it’s what I get the pleasure of working on here at PrecisionLender. If you really look at it in terms of what is the behavior we’re trying to change, that’s not necessarily a data quantity issue. A great example is would you like your front-line staff to wish someone a happy birthday if they come in on their birthday? Well, I can imagine that your frontline staff don’t have the birthdays of every customer memorized, but your computer systems certainly do.
Making that data available at that point of interaction to change that behavior is something that every organization is capable of doing. It’s not difficult to say, when someone types in George Neal, trying to access my account, if it pops up and highlights that it’s my birthday, and someone can say, “Hey, George. Happy birthday.” That changed behavior, it created a different customer experience for me. It added value to the relationship by making me feel like I was acknowledged as a person. It’s purely beneficial, and it doesn’t require a huge sum of data to do.
Maria Abbe: That’s really cool. It really helps us as humans be better humans, if you think about it, and do what we do best in building relationships.
George Neal: Exactly. One of the concepts we have behind Andi and really behind all of PrecisionLender is let the computers do what computers do best so that humans can do what they do best. Computers are great at memorizing details. Computers are great at doing math. Computers are infinitely patient in looking for anomalies. They don’t mind sorting 100,000 items to find the one that’s different. As people, we could do all of those things, but that’s really not an effective use of our time. What we can do that computers can’t, there’s no way for a computer to effectively smile at someone, say, “Hey, Maria. We’re glad you came in. Happy Valentine’s Day. Have a piece of candy.” The computer doesn’t do that as well as we do.
Maria Abbe: That is really cool. Well, George, thank you so much for sharing all of your knowledge with us. I know that I learned a lot, and I hope everyone else did, too. That will do it for us today. Thank you everybody for listening. You can always find more information about today’s episode at precisionlender.com/podcast, and if you like what you’ve been hearing, make sure to subscribe to the feed in iTunes, SoundCloud, or Stitcher. We would love to get ratings and feedback on any of those platforms. Thanks for listening. Until next time, this has been Maria Abbe and George Neal, and this is The Purposeful Banker podcast.