Collections Consistency Problem | AI Journey Ep. 1

Adam Parks (00:00)
Hello, everybody. Adam Parks here. Welcoming you to our first episode of the AI Journey, getting beyond the demos and talking about technology and its actual impact on the debt collection industry. And today I have one of my really good friends here to join me for our first episode talking about the same accounts with different results. Why collection consistency is key to success and Charlie Bonner joining me here today with NCB Management Services has a storied background working for major banks, helping fintechs to build their collection operations, and now working with a debt buyer. So I feel like Charlie's perspective is something that I really wanted to bring to the table today for this discussion. So Charlie, thank you so much for joining me. I really do appreciate you coming on and having this chat.

Charlie Bonner (00:51)
Sure. Yeah, thanks Adam. Appreciate it.

Adam Parks (00:52)
So, absolutely. So Charlie, for anyone who has not been as lucky as me to get to know you through the years, can you tell everyone a little about yourself and how you got to the seat that you're in today?

Charlie Bonner (01:01)
Yeah, sure. No problem at all. Be brief. I spent roughly 30 years on the banking side with two of the major banks, ⁓ Citi and also with Chase. And then recently in the last, before I joined NCB, I was with a startup called Mercury Financial, which is now part of Atlanticus. And during that timeframe over the 30 years, I've been in finance on the collection side and also have worked on the recovery side of the business.

So from running dialers to app approval to looking at numbers on a P&L, it's been a fun run for 30 years. And then I decided four years ago to come join one of the industry leading companies in NCB Management Services. And I've been here just over four years. And the pleasure is I've known you for a good long time since you started the industry.

Adam Parks (01:43)
I mean, quite literally since I was very, very new to the industry ⁓ all the way back in my CreditMax days. Charlie, We've talked a lot about the application of technology to the debt collection industry, all the different things that are happening. But as someone who has managed vast portfolios that included debt sales, that included not just debt sales, but agency and litigation services, right? All three legs of the stool of debt collection. Consistency is hard and we can take the same accounts and we can apply these different treatments and we can see some vastly different results. And it's not just high level strategy. I think it's the actual orchestration of these strategies and how we can see that differentiation. What has your experience been through the years as you've taken the same portfolio or a very similar portfolio month to month and tried different strategies? What have you seen in your experience?

Charlie Bonner (02:39)
can tell you the key thing is setting it up correctly. It doesn't even matter now with AI to where it was back in the day with manual. A couple key things from our previous life when we were selling debt in a sense, we were doing at some point 30 to 31 debt sales a month, right? But here's how we, yeah, it was crazy back in the day. But here's how we set it up to save our FTE times. We didn't have AI then. So what we did was we did automation through an IVR, right? If Adam called me and he was at ABC debt purchasing company. We never saw that customer. It was automatic IVR that went back to you. Same with the web, everything. So was kind of like the first time to say, hey, here's a little customer treatment. I'm not going to become the clearinghouse for your direct payments, right? I don't own the account anymore. It's now Adams. I want to get you the money as quick as possible. So those are the things we did, honestly, it was through trial and error. So what we did was we had some buyers at that point wanted some we did not. Right. That's in the old days. That's several years ago. So going on to a FinTech. how we set up and have a little bit more tools than you would have at a bigger bank, right? We were able to set up the automation, self-service, and this is even before all the AI boom that's happening right now, Adam. So we really saw a big difference from, you know, leaving the bank, the big banks to the fintech. It was almost like night and day. We were able to get payments in the house quicker. We were able to satisfy the customers, the solution, right? We want to get that customer back on track. That was the key for us at that time, is to help, say, Adam, get a resolution, get him back on track, get him rehabilitated in a sense, and possibly down the road, start lending money again to him, right? That was the key for us. And really, it worked through trial and error and setting it up the right way. Fast forward to today's world, a little different, but we'll go to that in a minute.

Adam Parks (04:16)
Well, think right now we're looking at this increase in accounts. And I know that you were working through the 2008 crisis as we saw this massive increase in the volume of accounts. We saw a decrease in the liquidity of those accounts. And I know we saw that again, like kind of dragged on for many years, 2008 kind of being the pinnacle. And from the TransUnion Debt Collection Industry Report in 2025, in the survey, we saw that 75% of companies are expecting volumes to increase. 48% are expecting a significant double

Charlie Bonner (04:24)
is that

Adam Parks (04:45)
digit increase and that these mid-sized firms were expecting to see the largest growth. Now, in order for everyone to do more with less, in my opinion, the standardization of certain things becomes critical. But at the same time that we want to standardize things, we can't lose the personal touch that ultimately makes us the debt collection industry and allows us to operate in, lets call it the shadow of the banks themselves, right? Because we're dealing with a small subset of accounts that we're no longer performing. So how do we start to look at that balance between standardizing the way that we're doing things for stronger consistency and also balancing that with the need for personalization and that attention that helps us bring those consumers to resolution?

Charlie Bonner (05:39)
I tell you, this is a really good question and you and I have talked in the past. We always need that human intervention in a sense, Some customer will talk to you no matter what. You can have the chatbot, you can have everything else that you want out there, but that customer wants to talk with you. And honestly, Adam, it can go to a 12-minute call where they're going through their personal life, their willingness, ability, why they were going to pay now, and they tell you a story. Your job is to show empathy and listen, right? Make that person feel like you are really there and you really are there for them, correct? That's always going to be present there. We're not saying AI is going to replace everything in a sense, but the key thing here is that I've always said and you and I've talked about in the past is the data going in, right? There's going to be that customer, honest to goodness, there's going to be that customer who doesn't want to talk to you or me, but that person is going to be up at three o'clock in the morning worried about their bills. How are they going to pay him? They're going to go into a portal. They receive the text that day or they get an offer and they go on the portal, guess what? All the AI behind it, the data is there, the data is clean, the data is showing behavioral, right? It's saying, here's what I can offer you today. This person has done X, Y, and Z, they've done a great job and now they're going from, you know, maximize now and making this up at a 70% settlement, right? This customer has that. They want to get their credit period set up, but you know what's funny? We've never talked to him or her. That person never once called us and it's all handled in the middle of the night, batched up.

That person finds the resolution they're looking for different than the person who wants to speak with you. They wait till 8 a.m. to call or at 8:30 at night, East Coast time to say, hey, we want to resolve this, please. Can you help me? But I think the big thing for the AI side of it is the clean data going in. If it is not clean, we're not validating it, that all that machine learning behind it and the algorithms that work it, correct, that I've learned from the CIO here where I work, it's not going to be good. It is going to spit out what you put in, correct? It's not going to do. If you have bad data, it's not been validated and it's not clean. That's what you get, right? So that's a validation part that comes in with this. So that learning is key to helping that customer out.

Adam Parks (07:41)
Garbage in and garbage out. And if we, can make up a thousand phone calls to a phone number, but if it's not the right phone number, we're never getting through, we're never going to be able to execute on it and we're going to have these problems. So it sounds like as we look to solve the consistency problem and how we're working accounts or the results that we're seeing on the accounts that we're working, the big chunk of that is related to the cleanliness and structured data that we're able to feed the machine because machine learning is going to learn based on the data that it's given. And I had a really interesting conversation recently with Manny Plasencia from TransUnion, who has operated on the creditor side for many years prior to joining to you. And that was one of his big things was garbage in garbage out. You can have all the best tools in the world, but if we're not feeding it and we're not connecting with the right people, then we're never really going to get there. But then becomes the question of let's call it digital trust.

And whether we're making a phone call, a text message, an email, we have to separate ourselves from those scams. And I think one of the things that we don't really talk about enough is like domain authority as an example. So I'll use the email channel. And as we talk about domain authority, if I've got three agencies that are sending emails, I'm going to get different results. There's going to be a consistency difference there. Some of that may be orchestration, some of that may be data. And I also think some of that is going to come down to whether or not they're able to develop digital trust, both with the consumer and with the networks, Gmail, Hotmail, et cetera, that are receiving those emails and the filters that they're using to stop things from hitting the inbox that they consider to be spam. How have you looked at, let's say, managing an agency network? and looking at those differences in consistency across those types of channels.

Charlie Bonner (09:36)
We have previous life here. When I've had the agency network and talking with my peers here, the key is one thing we do first, Adam, is we look at all the email templates, right? That's automatic. We want to see what they're sending. We got to, it's our brand, right? We have to protect it in a sense. We also, we want to be, really work as a good partner with them to see what's happening. And then what we try to do the best then is where sometimes there's failures. They're sending all these

Adam Parks (09:51)
Yeah, of course.

Charlie Bonner (10:03)
Where is the success? Are we getting anything from it? Now, we're going to work with our partners, right? They're spending the money to send those emails out and so forth. Where is the click through? Where is the payment? What is happening here? Why isn't this happening? You know, in a sense that we're getting payments resolved as customers. or why are you sending emails to a customer that doesn't even read it? You've just sent seven, eight, nine emails over time, right? You're getting nothing. Is there a different approach? Right now, do we look at the tech side as some of all this stuff we are very acclimated with? But the key comes down to is the data output from that. Is it really successful? What are you clicking through? Are they going to the website? Are they hitting the link and so forth? But like you said, they got to trust it first. They're going to look at that. They'll probably call and say, hey, Adam, is this really you? Is this you over here at XYZ Company? You know, but that is the key, honestly, the key. And if we can get them to click through, hit that portal, make that payment without talking to an FT, that's a win for all of us, right? We resolve it, we save the time, and then our FT are working on other higher priority or higher risk accounts.

Adam Parks (11:05)
I think as the consumers are getting younger, there's an interesting dynamic that's starting to change. I believe, and I've got some pretty good data to back it up, that consumers are most likely to want to service their account through a similar channel to the way that they originated it. When's the last time you talked to somebody who went into the bank, got a credit card application, and physically filled it out on paper and submitted it to the bank?

Charlie Bonner (11:22)
Yeah, I don't know when.

Adam Parks (11:31)
I can't even think of the last time that I did that, maybe before the year 2000. So the consumers now, or if they're going to a FinTech and they're originating this loan online, the expectation of a digital engagement, whether that's self-service portal, an AI bot, chat bot, whatever the case may be, is significantly higher because that was the expectation and potentially part of the decision criteria they used when they selected that lender to begin with, because they said, I want to borrow from a fintech. I don't want to go into a bank and fill out an application. I want to borrow from my phone or I want to book my iPad, my laptop, whatever, whatever their flavor is. But with that type of origination and what we've seen in the, the federal reserve, Fred data is that there's an increase in originations that's happening.

Charlie Bonner (11:59)
Correct. Okay.

Adam Parks (12:21)
pretty much uniformly across product types. But what happens when this volume increase continues, but we can't keep our workflows up to date? We can't hire fast enough. 88% of companies are having trouble hiring, 81% are having trouble retaining the people that they are able to hire. Collection training classes are getting smaller, and I think it's improving on some level over the last, let's call it, year or so as compared to 23 and 24 where it was extremely difficult. But what happens as these volumes increase? How do we as creditors look at the service providers and say, how do we evaluate the workflows and are they modernizing themselves at a pace which is going to support the future expected volumes that we're thinking we're gonna see from 25, 26, and probably 27.

Charlie Bonner (13:16)
I think it goes down right to what we talked about earlier, that data point where we could use automation. I think you nailed it with the hiring part. It is very hard to hire, seriously, and to find good folks who will stay with you, who are good at collections and so forth, even the back office side, but I really believe that data side to where we can bring it in through the AI component and have that data where we're using that automation.

Adam Parks (13:21)
extremely hard.

Charlie Bonner (13:41)
It goes back to your processes, right? What is streamlined? What is going on here? What is so manual? Why are we doing it? Speed to market, right? I want to get with Adam. Like you said, folks went to FinTechs. They got invited to go there, by the way. They could have had a bankruptcy on their credit bureau seven years ago. They're getting rehabilitated. We're going to invite Adam back. We want you as a customer, right? We want to help you out because you know what? you become bigger and a good share of our wallet, we want to be there for you. But you know what? They're going to to automation.

The AI side, behavioral. They're doing all that. They're not sitting there making the, you know, pounding the phone and only doing the reg F side of the seven, seven. You see them saying it's just, it's gotta be that part where they have to use the automation. It is crucial in today's world. There should be no reason why a company couldn't ingest three, 400, all the accounts with all their other clients, right? And use that automation. They know what to hit either through the omnichannel, the digital side of it. and what they need to put out to the collectors, right? Like we always said, there's always going be a collector. There always will be. And I'll tell you something, I was just listening to this the other day and reading through what Jamie Dimon said just here yesterday. He said, it's time to embrace the AI side of it. He goes, in our next two generations, your grandkids and their kids, right? They're going to probably live in a world where they're going to average age to be 100 years old, right? And probably most of the diseases are going to be cured, not all. And guess how they're doing it? Rapid. Data, they're learning, they're on the spot doing it all. It's a great observation from that gentleman. I mean, he's done those things since taking over Bank One many, many years ago. And I think he has a valid point. And from him being at the largest bank and making a statement, probably woke people up little bit like, all right, maybe we need to listen to this. He knows. He's invested in it, right? So it just kind of stuck with me over the last day or so. He has a good point. he talked about automation, data. clean it, make it good. So that's kind of my feelings on it. Honestly, Adam, where I think we could truly use the AI driven automation to get those customers taken care of. And the ones we need to talk to, we'll know. We'll go out and we'll know who to talk to.

Adam Parks (15:42)
So what I'm hearing is that we need to work smarter, not harder. And that the biggest bang for the buck is potentially hiding in the prioritization and workflow treatment strategies, not necessarily just how many calls can I outbound. And I think that the use of the AI voice agents is going to continue to improve and

Adam Parks (16:03)
I look at AI right now, similar to the way that we look at, and I forget the name of the semiconductor growth rule, how many circuits are going to be on a semiconductor and how it was a double or triple every year. And then we look at data growth and what's happened. We've seen exponential data growth year over year since the invention of the iPhone, because people are taking more pictures, right? It's just happening. And with those little computers in our pockets all of the time.

Charlie Bonner (16:18)
year over year.

Adam Parks (16:28)
we're increasing the amount of data that's necessary to be stored. I mean, look, back in 2010, would you ever think that you'd have terabytes of storage on a telephone? Like that was not a realistic comprehension at that point. It would have been tens or $20,000 for a phone and it would have been, you know, the size of the 1980s Gordon Gekko brick phone. It was just kind of a different time, but we're seeing that same kind of exponential growth potentially even.

Charlie Bonner (16:52)
Yeah.

Adam Parks (16:57)
faster right now and what's starting to happen from a technology standpoint. But I go back to a challenge that I think we have here, which is really why does similar portfolios perform differently as the volume increases? it the people problem and we just can't put enough people on the phones to make that or are we just at the beginning of learning how we can truly orchestrate our communication strategies to meet this growing need of accounts? I want to say it was last year I was hired to write a research paper by a pretty large financial institution and they asked me to predict the volume of charge-offs both in the number of accounts and dollar volume going out to 2030, which let's be honest, is kind of a ridiculous task on some levels, right? I mean, how am I supposed to predict that? So I use simple linear regression analysis looking at, let's call it 23, 24, and into the beginning of 25. And the growth pattern is exponential over that time period. How do we start to create more consistency? And I know you're going to say, we're going to have to use these automated workflow and orchestrations, but any tips or tricks for the audience in terms of what they should be looking at and looking for to evaluate the organizations that are running these collections.

Charlie Bonner (18:22)
I want to get back just a little bit back to what you were saying about delinquencies, right? What we're seeing, I look at it every day, right? We're seeing right now on the debt purchase side, it's got to be almost 20 to 25% increase quarter-over-quarter. Year-over-year, 26 could be honestly, Adam, it could be one of the biggest years ever for seeing debt that's available direct, not resales, these are directs. These are direct from banks, fintechs, auto.

Give me commercial so for this is not a reseller. This is direct from the lender and it is astronomical what we're seeing right now, which parlays right into what we're talking about. How are you going to handle that right? The biggest thing is change management, right? People need to embrace it, right? There's two changes add the AI and so forth, but the three piece somebody taught me a long time ago. People, processes and performance, right? You don't have the right people. You don't have the right processes. You know what happens to performance.

Again, it goes back to putting those three things in play. It is crucial to have your processes set up correctly through testing, controls, and seeing how everything can be always enhanced, right? Big portfolios, I've seen them myself from, you know, $10 million to $800 million, right? How are you going to do that through a different strategy and workflow? You got to have that workflow. X days we do this, this, this, and this.

Where's our digital side of it? Where's this? You really have to map that out to a T. If you don't have that, you'll just choke on the volume. Or you're passive, right? You just outsource everything. See any other ways you can do it.

Adam Parks (19:53)
But even if you're outsourcing everything, you still have to look at the vendors that are operating for you. And from a creditor's perspective, most creditors are not handling everything internally. It's been my experience that most creditors, if they're operating any collection operations internally, it's a, it's as a champion challenger against their third party solutions. And this was an interesting comment that somebody had made from a FinTech at a private event I was at recently where they basically said, you know, look, the creditors are also investing in technology and artificial intelligence and they're making these investments as well. But the more of these investments that they make, the more difficult our lives become. And the smarter we're gonna have to be because the low hanging fruit is gonna get picked up by their self-service technology, by their AI technology, by their workflow improvements and they've got a better visibility than we do into the consumer situation because they can see line by line and they're ingesting in real time how those consumers are paying and what that ultimately looks like. So there's that whole piece too, but one of the things that I'm curious about is we're getting better connecting with the consumer. We're connecting potentially faster, but in connecting faster, are we causing any issues here where the consumer has not had enough time for that life changing event that pays off that $15,000 balance. And I feel like, are we gonna see consistency through this volume growth as we connect with the consumer earlier in the life cycle? Are they gonna be at a point to actually, they're feeling the financial pressure, but are they gonna be at a point to take an action to resolve the account?

Charlie Bonner (21:25)
Yeah. financially sure.

Adam Parks (21:33)
at that point in the life cycle.

Charlie Bonner (21:35)
That's a point. Is something that is kind of scary, to be honest with you, right? Do we want to keep on pressure? Look at the market, right? Even though we have unemployment where it is and so many different outside noises, right? The sense of effectiveness, they could be, it could really run them into an impact where like, look, I just can't, I just can't pay it. They're just not going to happen. What can you do to help me then? Right? What are we going to do to say, hey, Adam, I'm giving you my share of the wallet, right? It's going to be that key

Charlie Bonner (22:02)
point where we have that data behind the data science, right? To say, look, this person, they do have a willingness, right? They do want to pay. We got to make them feel comfortable to pay us. But just pounded on them, we got to follow the rules, right? We all do within the reg F and so forth. of course we all, mean, that's evident, but I think there is going to come a tipping point here. You're starting to see it more on the auto side, right? Deficiencies are, I'm worried. There's a little bit of 08 in me, right?

Adam Parks (22:06)
Thanks watching. was just the beginning.

Charlie Bonner (22:27)
Yeah, it's kind of scary. It's a little scary. Is it the new mortgage thing from 08 right? We'll see what's happening. You see folks going out of business, they're not lending anymore. It's getting scary, but I really do believe we could see that become a big problem where we're just hammering them too much and saying, look, I don't have it. I really don't have it. So that's going to be our listing skills if we're talking.

Adam Parks (22:47)
Do know what the current average LTV is for a used car across the United States?

Charlie Bonner (22:52)
No, I haven't looked at in a while. What is that? I'm interested.

Adam Parks (22:54)
127% and it spiked significantly from around 100% at the beginning. So you had all of these people, cars became more expensive. You had a lot of people that were paying premiums in order to acquire that new Bronco or that new Challenger or whatever. I remember being in a dealership during that time period.

Charlie Bonner (23:02)
hundred, yeah.

Adam Parks (23:17)
and seeing a $15,000 dealer premium on a Hellcat challenger. And I just couldn't believe my eyes that people were going to be willing to pay that. I think even the new car LTV started rising pretty high. It has dipped back down to about 100%, which is still kind of a scary number to think that we're doing that on a depreciating asset. So the collateral on the streets never going to be able to cover

Charlie Bonner (23:30)
Do have any as it dipped back down to about 100%, which is good. It's a little scary.

Adam Parks (23:45)
what's happening, people are not paying down fast enough to meet that need. And how many people are even taking the gap insurance when they're buying a new vehicle to make sure that they can cover the difference with this. So even in the situation in which the car reliability situation that we're gonna be facing, when they're buying these new cars, it gets into an accident, the car breaks down or whatever the situation

Charlie Bonner (23:45)
I agree. Right. Just going to say that.

Adam Parks (24:10)
entails and now the consumer is in a much more difficult position. Which leads me to some other interesting pieces of data that I found in the debt collection industry report from TransUnion this year. 39% of companies reported increased liquidity, but 25% reported worsening liquidity conditions. I have to assume that some of those were related to auto. Law firms reported the highest increase in liquidity, meaning that they're being forced to move more toward that litigation channel or that litigation channel is there, but agencies, collection agencies specifically showed the highest level of volatility in terms of their vision. Now, previous years, think might be offset in some levels based on the high probability of a collection agency to be collecting medical debt and the volatility that we saw in the medical community based on what was going on at the CFPB and in the government

Charlie Bonner (24:38)
It is.

Adam Parks (25:03)
and these restrictions with the credit bureaus, et cetera. So some of that, I think can be taken with a grain of salt, but if we're already talked about the increase in the volume of accounts that we're expecting, we're also at the same time seeing some of this liquidity decreasing. Consistency in managing our portfolios is going to become mission critical

Charlie Bonner (25:11)
you

Adam Parks (25:26)
as we go forward. And I would think that debt buyers specifically would be saying, I need to be as efficient as humanly possible, which when we looked at artificial intelligence and we said, which use cases of artificial intelligence are the different debt collection company types focused on? The debt buyers were most focused on the workflow, right?

Charlie Bonner (25:33)
Thank right?

Adam Parks (25:47)
Data appending, the workflow, you've got collection agencies that were more focused on communication tactics, the AI agents and voice communications, ⁓ chat and written communication. And then you had the law firms that were more focused on administrative type services. I think that's a pretty interesting differential, but now we're starting to see the birth of the, I've identified six use cases and I've been having this

Charlie Bonner (26:04)
That's right.

Adam Parks (26:13)
this discussion with a lot of different people lately is, you know, we found these six use cases that I think are pretty well defined, but is the seventh use case potentially the full on AI first agency? We've seen a lot of agencies go digital first, but what is that going to start to look like as we think about AI first from the ground up, because it's not just about the AI chat bot, it's not just about the voice agent.

Charlie Bonner (26:29)
That's right.

Adam Parks (26:40)
It's about the orchestration between these things, that workflow management and the addition and appending of data. But when you start tying all of these use cases together and you think about an organization that was built from the ground up and kind of an applaud to your CTO or CIO and having some conversations with him is really part of what got me thinking about the reinvention of the debt buyer over the last four or five years. And I you guys were at the very crest of that wave and kind of rethinking the entire infrastructure from the ground up. Do you think that the, do you think that we're going to see such a bifurcation between those organizations that have started their reinvention and those that are kind of waiting in the wings? Do you think we'll see a bifurcation between the success rates of those organizations over time?

Charlie Bonner (27:27)
Yeah, I tell you there are some still I talk with my peers. They're still coming like yeah, we're getting to the email. We're gonna get to that Omni channel side of it. We're looking at it, but they are just they're just not willing to change and they've been around Adam for a while some folks I just I Puzzling to me of why folks are not willing to look at that change But like you said we've done some really good things here with that in a sense to where I think we're industry-leading, correct? But to your other point, with the companies out there that hardly have any people, right? They're using the AI, they're doing it. I see their performances from my previous life and now, I mean, they do a pretty good job. I gotta be honest with you, and they're using these tools. But there's those old school groups out there that will not change. And I think you made it right when you bifurcated. They are going to be left in that, and they're going to be one of those groups that's going to be available to purchase. Because they're not going be able to make it. They're top heavy on expense.

They're not bringing enough revenue to cover everything and new cash, everything you and I understand well. And folks out there, I mean, to build up that cash flow so you can have that to pay your bills. If you're not using innovation and getting up with the times, you will get left in the dust. I hate to put it that way. That's the God's truth. And it is going to be scary, but some folks will not change. It's their way or else, and it's not gonna be good.

Adam Parks (28:45)
Do you think part of that is just masking the limitations on their technology platforms? They're kind of front-facing saying that like, no, no, no, I don't want to change, what I do works. But some of that is because in the background, it's still green screen. And there's a DOS backbone to what they're trying to execute on. And they're just trying to mask that technology limitation.

Charlie Bonner (28:50)
That's right. Totally. That's number one reason right there. They just don't want to invest the money in it. Correct. They don't want to change it. It's a lot of money up front. We know it, both of you and I've seen it. It is a, can get into millions of dollars, right? To change things. But if you're running it out, like you said, even to the 2030, right? There's got to be a way you can say, here's where we recapture our investment plus on the ROI side of it, right? Any way you can make it easier for a collector, even if it's through

Adam Parks (29:13)
Yeah.

Charlie Bonner (29:35)
Like we talked about earlier through the portal, through email, text, or even on the screen where a collector can see the information, react right there, not be going through like 16 different things and green screens, right? I'm telling you, it's that back to where you make that customer comfortable talking with you and paying you and they will call you back or set up PPAs for down the road. It works. I mean, think about Adam, what we've done since we started this business. I used to deliver you a disc. Remember, we always laugh about it. One little disc from FedEx had all your media on it.

Now I can go to a portal and pick all my like that. Think about me. It's like back to the phone conversation, right? You know, the brick phone versus what we have today. But literally you got one disk in a box truck and now you can go to a portal, pick your media up. Yeah, it's people can't believe it. What we've seen in the last 20 years.

Adam Parks (30:13)
Literally. Also, I look, I've received boxes of paper in the past and had to bring scanning organizations into my company in order to digitize the media so that I can get it into the hands of the law firms and agencies or the buyers if we were reselling one of these portfolios. And you bring up something really interesting about the value of informing the collector. And I think that goes back to the, we do more with less?

Adam Parks (30:44)
One of the things I think is interesting is we look at the use cases around artificial intelligence. I know that the, having writing a lot of content in the space, I know that everybody's very interested in the AI agents themselves. like this robot is going to solve all the problems, but the empathy can't be replicated. The people themselves are still the most valuable tool that we have. So what

Charlie Bonner (31:00)
But an emphasis the most valuable tool that we have.

Adam Parks (31:10)
tools in the AI toolbox can we be using to help improve the value of each individual person that we are able to hire and retain? If we can't hire as many people, we can't retain as many people. We need to make them superhuman as much as possible and focus their time and energy on the best accounts, which is the workflow piece. But I think it also comes down to being able to feed them that information, the copilot type tool sets that allow us to take this simple collector, and that's an understatement by every stretch of the imagination, but you take this kind of this simple collector and you're supercharging them by providing them with more powerful information, more real time information about what comes next. I think that's one of those really big drivers that can add a lot of value for us as it goes forward. Are you seeing that internally, externally? How do you feel about the bifurcate or the differentiation between it's going to be an AI bot and the AI is driving the ship and feeding the human better information for them to be more powerful in the resolution of account.

Charlie Bonner (32:23)
Two-fold answer. The AI ⁓ bot, me, is going to happen. It's working. We know that. It leaves a fingerprint on everything we know. And then the collector can see that through the notes if that person actually does call in. But that data, if you can give the collector that information where they have Adam on the phone to where they don't say, hold for minute. I need to go talk to somebody just to get a 10% reduction in the settlement. Give it right there where everything is in a drop down. Everything works. I mean, it is the crucial data point that they know that Adam didn't pay us for six months. He's had two NSFs, we got to now either get a debit card from him or her. Get this done right now. Take care of this today. We're not going to wait on check. We're not going to do it this way. Let's get everything has to be on that screen without a collector having to go hold for a minute, please. I'll be right back. And then 30 seconds goes by and they come back. Yeah, we can do that. Get it done. We have on the phone. You disconnect that call. You're never getting Adam back.

You're going to avoid us like what it's going to happen. I think AI is the key, right? All the data behind it, the data information it has, can put that all on a screen and give that collector who is well educated on the process and can talk to anybody from willingness to ability to get them to pay. Right. I believe you're equipping them as a person who's going to be able to resolve that debt, keep that customer paying and really, truly make them a customer for you if you're at a bank for life or resolve the data if you're in a debt-buyer or an agency. It's crucial. It's got to be a combination of both. If they're not using it, folks are going to be... It's going be interesting.

Adam Parks (33:55)
Well, you've been empowering them in the conversation itself. So being able to say things, for example, like, saw that you were just on the portal and didn't make a payment. Was there something about your experience that I can help you with? Because not only is it an incredible opportunity for us to resolve the account, it's an incredible opportunity for us to learn more about how we can better be servicing these consumers. Because the consumers, although every consumer is unique, every situation is unique.

Charlie Bonner (34:04)
right.

Adam Parks (34:22)
You can put them into categories and you can learn a whole lot as you start to organize that large data set in new and interesting ways. And as we categorize it, as we organize it, I think it puts us in a really interesting position to better understand how we can actually help the consumers. And I think that also comes down to, and I'm curious, Charlie, as you've managed these kind of large portfolios over time, portfolios change over time and it could be origination rules have changed, it could just be consumer trends. How do we keep our finger on the pulse of what's changing in these portfolios to better understand how we can manage inconsistency in our portfolio over time?

Charlie Bonner (34:55)
We keep our finger on folios to better understand. Say the big thing is, like you said, origination underwriting, right? That is going to be the big driver. I've seen it two different companies to where there was efforts put forth to grow, grow, grow, grow, grow. What's funny is, delinquency went up, up, up, up because we put on the different. And you'll see it in your 90, in your 60 plus, excuse me, on the pre side. Then you're like, OK, now let's go sell it. Wait a minute. This was 14 cent paper. I'm now getting 10. Now you start going through the score and you're like, oh, gosh.

Adam Parks (35:20)
You

Charlie Bonner (35:32)
We now put on, making this up. This is just me saying that we put on five. We're doing 650. Yeah, we were doing 650s. Now we're lending to 590-580. What is what is happening here all of sudden? But the key is is the data, right? How you build it. You know, that's why I really did enjoy when I was at the Chase. I really did enjoy having the internal recovery unit because you learned a lot from it. But the issue was we didn't have the tools like it like a debt buyer had or an agency, right? You guys had better tools that point in time than we did.

Adam Parks (35:35)
Yeah, illustrative purposes, I appreciate it. Yeah.

Charlie Bonner (36:00)
Now that I'm on the other side, I see that and I understand it. But we learned. All of sudden, the customer charged, they're talking to you. Why are you talking to me now? We tried for six months to talk with you. They just didn't want to deal with it. Now they're charged off. Now they're like, oh my gosh, I have an R9 on my credit bureau. What am I going to do? Well, now we need to talk, right? So yeah, it comes down to every collector. People listening to this are probably laughing this next day because we all do it.

Collections blames underwriting, underwriting means finance, Or customer service. It's the old school way, right? But when you see it, like you said, watching the portfolio, if you don't have an RU, you're going to see the drop in your price. Debt buyers are not silly. They understand this better than anybody, right? They're going to price it what they see. But you will see if that is the number one indicator is I've seen two different portfolios. Let's just go and ram, ram, ram. Let's get these through and let's do it. Portfolio declines because the customer's worst. They need to save from credit score. It's the biggest thing for us, right? And other things we've been talking about for the last 30 minutes around workflows, automation, surrounding yourself with the right people, right? The leaders out there talking about it. Have the great leaders out there. They're the support. They're the cheerleaders. They're the ones that have to get the resolution and get everybody's buy-in to get this thing resolved.

Adam Parks (37:17)
I like the way that you think Charlie as usual. So as we think about liquidity performance, how much of the performance of a portfolio do you think comes down to decisioning versus the execution of the collection's operation.

Charlie Bonner (37:32)
I will tell you I've seen it in the previous life where we can have all the great decisions. We're going to do everything right, but if we're not executing it the proper way, it doesn't matter what we do. Honestly, it really doesn't. If I'm going to say, hey, I want to go out and carve off 10% of this book, keep it in-house. Right? And I did this. I'll share this with group. I have no problem sharing this. I carved off 10% of a book. I realized I carved off like 200 accounts. That's not a fair sample.

I learned nothing. I went six months of nothing, right? I didn't learn anything. I just should have sold it all, right? I really should have. And that was my decision. Then executed right. I learned nothing. But it's like anything else in collections, correct? mean, anything. Finance doesn't matter. Even what you do, right? If you write some of it, you're like, I'm not really getting the good feedback on this. Nobody's really calling me, questioning things. You're going be like, oh, was this the right decision? Did I execute it correctly? It really does come down to that. And it goes back to that leader.

Adam Parks (38:01)
Yeah.

Charlie Bonner (38:24)
I strongly believe in that, that we walk out with a decision, we're going to do X, Y, and Z. If they go out with A, B, and C, we're toast. We're not going to perform well at all. We're not going to hit the numbers. We're going to have a problem on our 60 plus, or we're not going to hit certain attributes from a click rate standpoint. We are in trouble. And that leads in the worst things, right? So I just really believe in that execution side. And sometimes the decisions are bad. And you can execute all you want.

I mean, like I said, I made bad decisions and we executed perfectly. It was horrible. But you learn, right? You learn from it. But when you're dealing with big books, you're with 500 million, 600 million or 1 billion, right? You got to make sure those decisions are right on spot. Don't be testing the water on certain things, right? If you're going do it, better be.

Adam Parks (39:07)
can't test the water on the whole book. You gotta set small pieces aside. You gotta execute there and say, did this work? Did it not work? How's the measurement? Like that's where we start seeing the differentiation.

Charlie Bonner (39:10)
Yeah, exactly. Yep. I mean, I'm still surprised, Adam, and this probably won't shock you, but it may shock folks out there. I'm still surprised how many creditors still put their DSCs in their book and sell it normal. I'm just surprised by that. It's just something. I think it comes down to, like you said earlier, FT. Like, can we do it? I mean, if you do a sale for $5 million versus $50 million, it's the exact same effort if it's $5 or $50. It doesn't matter. You still got to do the same work.

You still have the same flows. You still have the same business requirements. But I can tell you, people will argue this point with me, like in the end, it all equals the same price. You got a 20-something asset versus a right. I don't believe it. I've never seen it. And I mean, come on. We sold it separate. We always did. It's the way to go.

Adam Parks (39:54)
doesn't really. Yeah, but it's also, it's about specialization and it's the same, let's go back to the old CreditMax model. We would buy a national pool of accounts. We would work it for 180 days and then we would sell it on a state level, mostly to state level attorneys because those attorneys were going to be willing to pay more for that portfolio than others because they had specialized knowledge of the jurisdiction.

Charlie Bonner (40:16)
these guys sell it on a payable mostly to stream the public. willing to pay more for them.

Adam Parks (40:27)
And I mean, we can get down to maybe we went a little too granular in trying to sell individual accounts and zip codes and all of that. I just don't think there was a large enough, but I don't think there was a large enough debt buyer marketplace. If there had been 10,000 debt buyers, it might've made more sense, but there was 1500 or so. Let's even say 3000 at the time back in 06, 07, 08.

Charlie Bonner (40:34)
I remember that.

Adam Parks (40:52)
So it was a little bit different there that our ability to break that down are, and I would think the same thing if we were talking about that settlement accounts versus payer files versus, you know, your standard charge off file versus straight rollers, right? Those, those accounts that are most likely going to be fraudulent. And now I'm going to have to spend a whole lot of time and energy dealing with buybacks, put backs. And now I'm going to have all of this back and forth administrative process and you know, challenges there that are not necessarily going to add any new value to the equation. But do think there's any specific decisions that you look at other than that type of portfolio segmentation that are going to have a big impact on the financial outcomes of a portfolio?

Charlie Bonner (41:37)
You know, the other thing, definitely the mix is going to be a crucial thing, correct? Right. The other thing is we really tried to, even from a previous life, that payer, right? Why did Adam not pay me the last month? And now he rolled because he didn't cover his last bucket. Right. You went from a five to six. You've been holding in there, right? Do we as a bank then, right? When I was there, I'd say, do I hold that for another 30, 60 days and create like a zero agency sale?

Charlie Bonner (42:02)
Because there's a reason he's been a really you just didn't pay enough to cover the buckets. You know, at the five now you're right. Charges off at 180 or 210 depending on the cycle. What happened? We got to get him on the phone. We got to get her on the phone. We have talked to that person because they were making payments. Then also you get back and educate the collector like you should have grabbed this educate the customer. You're the bank or you're not, you know, the FFI you see side. Can you do the two ranges in five years? Right. There's things, there's so many things we could do, but that's other data we look at and learn. Last payday. Right, what's happening here? It's crucial because you can learn from that. You can really take that data and as you said earlier and really get the cleanliness out of it and say, okay, here's what we have. You when was their first date of delinquency? What changed in this person's life to make them actually a straight roller, right? They were good person. They've been on the books for eight years. Something happened that person charged off. We have to figure that out, right? So those kind of things we look at, even from bank side or the side, it's the same things.

Adam Parks (42:59)
And it's interesting, the way that you say it, makes me think about, I'm going to say the smartest thing I heard from a professor during my MBA. And, and professor Ralph Norcio at Lynn University would say repeatedly, there's only three things that matter in a business. The cashflow, the timing of the cashflow and the risk associated with that cashflow. And it sounds like you were looking at your, those that are rolling out of, know, bucket six, kind of the same way.

What actually happened here? What's my real risk of me holding onto this and how much more value does it have than someone who slides right through the buckets? back then we didn't have the spreadsheets weren't calculating that for us. We didn't have that type of calculating power dedicated to the deck collection space at that point in time. Now, maybe we did an underwriting in other areas. I'm not as familiar with that side of the business, but we definitely did not have that kind of computing power. And I'm wondering how different life would be if we had at that point and we were able to price those consumer accounts a little bit differently and demonstrate the difference in value because those two different consumers that you described have significantly different value in a debt collection portfolio.

Charlie Bonner (44:07)
Yeah, I totally agree. And the thing is, with that customer, there's something, like I said, we got to find out what you talk, what changed, how do we get to this point? Right? We have to really, and even today's where that's the key is hitting through and understanding, you know, what can we do to help Adam get him back on track? Right? We got to figure out your financial situation. You know, how much you paying for rent? How much is your car? How much is your cell? What is it? You have 10 grand, right?

Charlie Bonner (44:32)
We're trying to figure out now what can we do to get at them. $10 payment is not going to cut it. Not to 2080. You know what saying? It's not happening. We have to figure out some way to get you back on track in a sense. And that's what comes down to that data behind it. How we presented on the dialer, if we're calling in right or we're doing manual, it doesn't matter either way. That's that data back to what we said earlier. What can AI do behind the scenes to give you all the information? Crucial.

Adam Parks (44:38)
Yeah. When we think about the consistency of performance on portfolios, and we've talked a lot about the workflows and strategies, how different are the workflows, strategies, and execution across the different products? How different are we treating an auto consumer from a fintech consumer, from a credit card consumer, from a key lock consumer to a, and I could list all of the different product types, but how different are the strategies for success across those different product types?

Charlie Bonner (45:26)
Definitely differences. Small balance versus a large balance, totally different. It could be more of a omnichannel approach, right? You you have a $400 balance. You're not going to be sitting there spinning that like you are on average balance $11,000 auto-deficiency, right? There's going to be different strategies there. You're going to take and look at different talk-offs, correct? Different talk-offs could be to where, you know, say an auto-deficiency, right? Your car's been repoed. It's been sold. It's now been a year.

Adam Parks (45:44)
Sure.

Charlie Bonner (45:52)
You want to buy a house, whole different conversation. You know, they want to somehow resolve this debt. You're not going to set them up on a $25 PPA, right? They're going to talk about, let's settle this thing. Then we change to a different one, say DDA account, right? $150. Like, hey, let's take care of this now. Let's go. It's been this point now it's been charged that we need to really resolve. They'll pay the $150 or they'll do a 95% most time balance and Full or PIF, right? But there are different strategies. It's a different approach.

Do you do X with FT here? Do you use the Omni channel here? Do use AI here? It is different based on the asset, the average balance size and so forth. It will be treated differently. And it should be, right? You're not going to throw a lot of FT out of, like I said, 100,000 accounts. know, that's just not going to You're going to get through them all, right?

Adam Parks (46:38)
You can't get enough FTE. I mean, look, the overhead for all of that FTE is just insane. Forget the employment, the buildings, the phones, the systems, the administrative support. Its a

Charlie Bonner (46:46)
outrageous.

Adam Parks (46:47)
big lift and it's a lot of time and energy and a lot of diverted attention for top tier management. So the ability to kind of turn some of these things over to organizations that have that technology. But that kind of leads me to the question, is this becoming more of a technology gap or is this more of an operational gap between performance capabilities?

Charlie Bonner (47:07)
The folks I talked to, it's both. I hate to say it that way. You know, again, they're archaic. They're sitting on that system, no investment into their systems and so forth. Then others, they're going to lead the way. They're investing in it. They're investing in the people. They're investing in the technology. They're the ones that are leading the charge. Because guess what? If you're on this side of the business, when you're doing it, you can pay a little bit more, right? You can do this when the other person is still struggling just to recover their initial investment because they're antiquated. I hate to say it way, Adam is being honest. I see it.

They're my competitors out there. You see it in our, I see a lot in the FinTech side where you can tell when you get a book, because I spent time on there, you can tell even from the survey what's going on. Treatment's important when you're making decisions. You can see that and you can see by asking the questions to the companies. And it's crucial because we need to understand what happened to that customer.

Adam Parks (47:37)
But that's why I wanted to have this conversation with you. Okay, I got it.

Charlie Bonner (48:02)
Now that they're charged off or if we're working the servicing side of it. You know, if we're servicing it, even at the bank, I want to understand how can we stop you from rolling through from five days to 180 days. That's crucial. Back to guess what? The AI data science behind it, the data you're given, how it works, invest in technology. Crucial.

Adam Parks (48:21)
So that leads me to yet another question, because I got a whole list of them here, because I've been thinking about this particular discussion for some time. At what point does the human decision making stop scaling? We're dealing with the increase in the volume of accounts. We don't have enough people on the phones. And unless we're feeding our live collectors with additional information and we're supercharging them like we were talking about before, there's got to come a point in time where

Charlie Bonner (48:23)
Yeah. for some time.

Adam Parks (48:48)
we just can't make human decisions fast enough and we're going to have to rely on the machines if our margins are continuing to compress. Now we could say that you could try to grow infinitely and you could use BPOs and you could try and scale in different geographic locations and those kinds of things. But even at that point, there becomes a limitation of a Peter principle, a tipping point if we want to be all Malcolm Gladwell about it. But what does that start to look like? at what point do we hit that limitation of our ability to understand because we can only comprehend so much data at any given point in time as humans. So where do you think we start to hit that limitation or that tipping?

Charlie Bonner (49:29)
My point of view on that, it can really, really impact you. You're going to know in that first 15 to 30 days if your workflows are working right, you're going to see it. Everything has gone out. It's, sitting here waiting. We're following the rules as always, right, like we do. And you're going to see that right away. And then you need to go in and say, okay, what is performing? Is it on the channel? Is it this? Is it that? It's the key in there. We're not always going to have the right information, make the decision. That's where we're going to have to use the data. The data is going to tell us what to do with it. It's going to tell us, look, Adam, he's just not paying. He needs to have litigation. He has assets. He has X, he has Y. He needs to go. We need to outsource him and litigate right away. Pre-legal talk, a boom, litigation, right? Or this person, we don't have the data, we don't have, excuse me, the assets.

We got to figure out a way to get a hold of you. And it goes back to the it's back to data. It's just usually can tell with 30 days at a monocle of some of these books, you know, I see. And also when I was on the other side, we could tell right away as soon as that count charged off, you know, all of a sudden I started to see direct pays coming in from a person we didn't talk to for six months. Now, why are you paying all of sudden? They got a different letter, right? Different company. And I'm like, really? Second voice got it, you know, but it's

Adam Parks (50:37)
Second voice.

Charlie Bonner (50:44)
It's not really cut, you know, it's not black and white and cut right down the middle. Like, hey, at this point, you can use the seat within that 30 days. You know, seriously, where you can say, OK, now this isn't working. I've seen it where even on the other side of the business where we're never settled, we're not doing this. All of a sudden, roll rates go through the roof and you're like, OK, maybe I should have settled. Maybe I should have did something in the back buckets. Right. You learned. Exactly. Yeah, always right.

Adam Parks (51:06)
hindsight is 2020, right? So Charlie, I have one last question for you as we kind of hit the top of the hour here, and I really do appreciate you spending all this time with me. What do you think companies are doing today that just won't work in three years?

Charlie Bonner (51:10)
Thank We've talking about the whole way. Please stop doing the old archaic ways. You know, let's just keep on pounding. Do the seven and sevens. They're following the rules. I'm just saying the folks who don't want to invest in the technology, they will get left in the dust. Have no problems with taking their business, not being rude here, just being this competition, right? I'm competitive individual. We have no problems with that. They need to adapt. They need to do change management, right? Listen to everything. Talk to groups after.

Charlie Bonner (51:47)
How many folks, and I have a good friend over at this company, how many folks thought five years ago a true accord and indebted or a January, right? Whoever, or let's say six years ago, right? Who thought that? Did you and I ever sit down at a conference saying, you know what we're going to do? We're going to create a company where there's no, basically no collectors, right? And we're going to do it through this way. We probably would have laughed at each other a little bit. We've been like, that's pretty funny, you know?

Adam Parks (52:11)
It's interesting that you put it that way because I like when Kompato came to me and sponsored this series and said, you look, we're I had done some work with them. They were the first group that allowed me to actually interview an AI bot, like the AI agent on a podcast. So I actually got to interview and engage with it. And that was the first time that I went, okay. All right. The future is now and I need to start thinking about things in a new way because I'm literally

Charlie Bonner (52:12)
change.

Adam Parks (52:37)
having this conversation with a bot and it was right before last year's conference. My wife was pregnant. We were having our daughter. So it was the first RMAI in like 19 years that I had missed. And I got texts and they were running in the, so in the Kompato booth, had the television running that actually had my interview with the bot running and I got

Charlie Bonner (52:48)
Remember.

Adam Parks (52:57)
I don't know, probably 20, 25 people that either took a picture next to the television or took a picture of the television and text me and was like, hey, even though he's not here, he's still at the conference. And it made me start to realize that maybe it's not just the tools, it's the orchestration of the tools and that these tools are actually going to hit a point of capability that I had never. I mean, look, playing golf together back in the early two thousands, AI wasn't even a topic of discussion. It was all about the manual effort. was, but we had, we spent a lot of our time and energy managing agencies and law firms and their performance. And we spent a lot more time on that. And now we're able to spend the time on the technology. But if you think about the volume of accounts that we were managing back then and the volume of accounts that are being managed today, we never could have managed the current volume of accounts. at that point in time using the tools and technology that we had at hand.

Charlie Bonner (53:54)
right. Yeah, it's scary, isn't it? I mean, there wasn't the SFTPs in a sense, right?

Adam Parks (53:59)
No, you know, look, it's scary, but at the same time, it's also a little exhilarating. It's interesting to think about now that we have organizations that have thought themselves as a debt collection organization from the ground up, looking at the workflows that we've been talking about and how things are being operationalized and how the voice AI bots and the co-pilots and the workflow tools all start to come together into this orchestration that if we had talked about it back then, we would have thought the consumer will never go for it. But now the consumer's borrowing the money online. So they're in a sense asking us to service their account in the same way in which they borrowed that money. And so the difference between then and now is astronomical and just such an interesting time to be a leader in the debt collection industry. Because I think this,

Charlie Bonner (54:35)
Yeah. It's crazy.

Adam Parks (54:53)
The race has just begun. I think we're at the cusp of what will be the next era of debt collection and ultimately financial services because for one of the first times I'm seeing the debt collection industry actually start to get out in front of financial services on the whole from a technology standpoint, from an understanding of the consumer standpoint, maybe not from an underwriting, but definitely from a servicing perspective.

Charlie Bonner (55:18)
I agree. And you know what it comes down to? Not as much red tape, let's be honest, right? We can move faster. Let's be, you know, we're nimble. We can go do it after. Exactly. Now, this was good, Adam. Thank you. I, it's scary what we've seen in the last couple, you know, years, we'll say couple of years, and to where we are and where we're going. It's exciting too. But it's going to be a fun future. It's going to be interesting for all of us. It's going to be a

Adam Parks (55:25)
we can use the cloud.

Charlie Bonner (55:42)
We'll see what happens with everything and how people adapt. And you've seen the changes now from just 10 years ago, who was in this American who's right. And then a big change in a good way.

Adam Parks (55:50)
Charlie I can't thank you enough for participating in this conversation. This has been one of my favorite conversations of 2026 and we covered a lot of ground in an hour. So I really do appreciate you coming on.

Charlie Bonner (56:03)
I did. Appreciate it, good to see you again. Looking forward to see you in May.

Adam Parks (56:07)
you as well and sit tight with me for a minute just as I wrap this up. Thank you everybody for watching today. We do appreciate your time and attention. Thank you to Kompato AI for sponsoring today's episode and the entire AI journey series. Very excited to have some really interesting conversations over the coming months. But thank you everybody for watching. We appreciate your time and attention. We'll see y'all again soon.