WEBVTT
00:00:06.400 --> 00:00:07.919
Hello, everybody.
00:00:08.080 --> 00:00:11.599
Welcome to an episode of Applying AI.
00:00:11.759 --> 00:00:20.559
In applying AI, we cut through the buzzwords and talk about realities of deploying artificial intelligence and financial services and other regulated industries.
00:00:20.800 --> 00:00:34.159
Each month, myself and Mike Walsh will be joined with another guest for us to talk about the practical application of artificial intelligence across the debt collection industry and other financial services.
00:00:34.399 --> 00:00:40.399
Today we've got a great guest with us that most of you or probably all of you recognize, Mr.
00:00:40.560 --> 00:00:49.520
Michael Lamb joining us to talk about the mergers and acquisitions implications of artificial intelligence in our space.
00:00:49.600 --> 00:00:58.640
And for those of you that don't know Michael, if it's mergers and acquisitions and it's in the debt collection-related industry, he's probably involved and engaged.
00:00:58.799 --> 00:01:10.239
So just all of the conversations around how this is going to have a direct impact on these businesses, the value of these businesses, and the changes that we're expecting to see.
00:01:10.400 --> 00:01:12.719
So thank you so much, Michael, for joining us today.
00:01:12.799 --> 00:01:15.439
We really do appreciate you sharing your insights.
00:01:15.840 --> 00:01:16.959
Absolutely, Adam.
00:01:17.200 --> 00:01:18.159
It's great to be here.
00:01:18.239 --> 00:01:20.000
And uh great to see you, Mike.
00:01:20.319 --> 00:01:21.439
That should be fun.
00:01:22.560 --> 00:01:28.400
So high level guys, MA is definitely a hot button in the debt collection industry.
00:01:28.560 --> 00:01:40.959
We've talked about the application of artificial intelligence in the businesses, and it appears that we're going to see a bifurcation of the industry over the coming, let's call it, 24 months.
00:01:41.120 --> 00:01:51.680
And the difference, the gap between the winners and losers of the AI race is going to become significantly more apparent over that time period.
00:01:51.920 --> 00:02:00.079
But, Mike, what are you seeing in the open marketplace and how are buyers and sellers viewing their artificial intelligence stack?
00:02:00.480 --> 00:02:01.840
It's a great question, Adam.
00:02:02.079 --> 00:02:12.639
There are the buyers that are coming in, as I've said to you before in some of our other podcasts, is we've got a whole new crop of buyers, investors that want to come into our industry.
00:02:12.800 --> 00:02:14.960
A lot of them are coming out of Silicon Valley.
00:02:15.120 --> 00:02:25.680
And again, as I think I said to you, Adam, I never my wildest dreams ever thought I was going to be having conversations with Silicon Valley tech investors about doing something in debt collection.
00:02:25.759 --> 00:02:28.639
It just wasn't my first thing that I would expect.
00:02:34.159 --> 00:02:42.800
Number one, for the companies that are in the market, they're trying to figure out when to buy or sell, they need to be able to articulate.
00:02:42.960 --> 00:02:46.319
It doesn't have to be perfect because nobody knows where it's going to go.
00:02:46.639 --> 00:02:48.639
What is their AI strategy?
00:02:48.800 --> 00:02:56.240
How does it affect their line of business, their service line, first and third party, healthcare, financial services?
00:02:56.479 --> 00:02:57.840
What is their strategy?
00:02:58.080 --> 00:03:02.000
Because everyone knows it it needs to exist and be involved in the business.
00:03:02.159 --> 00:03:07.120
It's more a function of how the company wants to apply it in their organization.
00:03:07.360 --> 00:03:23.840
And I'll tell you, the thing that gets the investors most excited is when they see how significant payroll is in the collections industry and when it what that line item means to a lot of companies in our industry, what can you do to reduce it with AI?
00:03:24.000 --> 00:03:28.639
And I've been having a lot of dialogue about that over the past several months.
00:03:29.039 --> 00:03:29.840
You and me both.
00:03:30.000 --> 00:03:30.479
I think it's two.
00:03:30.879 --> 00:03:32.879
Yeah, there's two different pieces to that.
00:03:33.039 --> 00:03:47.680
It's it's the cost savings of deploying artificial intelligence, but with the difficulties that we've had hiring people across the industry, I think we also have to keep an eye on how we can further empower the agents that are on the phone?
00:03:47.840 --> 00:03:51.520
How can we get them engaged and involved in the right conversations?
00:03:51.759 --> 00:03:55.520
We've got this increase in the volume of accounts.
00:03:55.759 --> 00:04:03.120
Everybody is, or the majority of our industry is expecting a double-digit increase in the volume of accounts over the next 12 months.
00:04:03.280 --> 00:04:10.560
But at the same time, you've got about a third of the industry that is witnessing a reduction in liquidity of those accounts.
00:04:10.719 --> 00:04:14.639
So, like always, we're being forced to do more with less.
00:04:14.800 --> 00:04:24.319
And we can either empower our live people, we can automate some processes, we can move consumers more towards the self-service options.
00:04:24.560 --> 00:04:44.879
But I think we're in this weird balance right now where if we don't start, if these organizations don't start moving down the path of identifying which use cases they're going to deploy, they're going to find themselves on the sale block in the not too distant future and not for a value that they expect.
00:04:45.279 --> 00:04:46.240
Exactly, Adam.
00:04:46.319 --> 00:04:48.000
That's the biggest, the biggest issue.
00:04:48.079 --> 00:04:48.959
But Mike, go ahead.
00:04:50.079 --> 00:04:58.079
I was going to say, too, like, I think let's say you're going to go down the sale route, and Michael, should correct me if I'm wrong.
00:04:58.240 --> 00:05:07.360
I think you have to implement some sort of no investor is going to look at it and say, oh, you're dialing and praying that someone picks up the phone.
00:05:07.519 --> 00:05:09.120
Like, I just don't see that.
00:05:09.360 --> 00:05:16.720
I mean, today I talked to an auto lender who just got a uh private equity and they're implementing an AI strategy, right?
00:05:16.800 --> 00:05:18.319
That was the first thing they did.
00:05:18.480 --> 00:05:26.720
Um, then I looked at then I was writing a ROI on our first agency we onboarded.
00:05:26.800 --> 00:05:27.920
It's 9x.
00:05:28.319 --> 00:05:30.079
It's 9x what they're doing.
00:05:30.240 --> 00:05:39.279
And they don't no longer make outbound phone calls to their biggest portfolio, and they've gotten almost 100% of that portfolio from their largest client because it's recovered more.
00:05:39.839 --> 00:05:47.279
They're charging the same price commission base they did, and per account they're paying us you know pennies.
00:05:47.600 --> 00:05:51.600
So, guys, let yeah but but Mike, let's paint the picture here though, right?
00:05:51.839 --> 00:06:01.759
Mo the average margin in this company in this industry is somewhere between 10 and 15 percent EBITDA, an EBITA margin today.
00:06:01.920 --> 00:06:05.759
When you go talk to the investors, they're like, eh, what's that?
00:06:06.000 --> 00:06:07.279
10 to 15 percent?
00:06:07.519 --> 00:06:08.079
Like, really?
00:06:08.160 --> 00:06:11.600
Is that all that's all that's all this industry can produce?
00:06:11.920 --> 00:06:21.199
But then you you actually have the discussion about the AI strategy, Mike, like you're referencing, and that 10 to 15 percent can become 25 to 30 percent?
00:06:21.439 --> 00:06:23.360
Well, now you got people listening.
00:06:23.680 --> 00:06:35.120
You got people saying to themselves, I got something that I can uh that is gonna throw off more cash with less human beings, that's more AI enabled.
00:06:35.439 --> 00:06:38.959
Who the heck wouldn't want that besides all the other stuff, right?
00:06:39.199 --> 00:06:39.759
Right, Michael?
00:06:39.920 --> 00:06:44.160
But it it's more predictable because AI is not doesn't have a bad day, right?
00:06:44.240 --> 00:06:48.639
Like it it's not it's gonna show up, you know, it's not gonna call out, right?
00:06:49.360 --> 00:06:52.639
It's more compliant if you build the right engine, right?
00:06:52.720 --> 00:06:56.079
Like it is it's not gonna swear at anybody, I can tell you that.
00:06:57.360 --> 00:06:58.959
It's not going off the chain, right?
00:06:59.519 --> 00:07:02.720
It's not gonna flip out and drop a bad word.
00:07:03.040 --> 00:07:03.360
Right.
00:07:03.600 --> 00:07:18.160
So I think I mean it I I think the one mistake people in this industry make, and I I'm curious if you see it, is they look at AI as a cost line, not an ROI line, right?
00:07:18.240 --> 00:07:24.720
Like, wait a minute, I'm texting at this rate and I'm sending emails at this rate, why would I pay you to do this?
00:07:25.120 --> 00:07:27.199
Well, how many people are part of that process?
00:07:27.279 --> 00:07:28.720
And what is it really doing?
00:07:28.800 --> 00:07:30.879
You know, like is it generating more?
00:07:30.959 --> 00:07:35.199
Like, are you are you outperforming your your your competitors?
00:07:35.519 --> 00:07:47.600
Um I think that's the biggest hurdle I run into, but when you're out there looking at you know buying or sellers, is that something you run into?
00:07:47.920 --> 00:07:54.800
Yeah, there there's there are guys that are not thinking in the right way, and that's what our job is as a consultancy, right?
00:07:54.879 --> 00:08:02.399
Is to go in and explain to them you can get X today, right, from an EBITDA or multiple perspective.
00:08:02.639 --> 00:08:06.319
You could get Y if you do if you pull some of these levers.
00:08:06.560 --> 00:08:11.279
Do you want to try to pull the levers, or do you want to just sit and get what you're gonna get?
00:08:11.439 --> 00:08:15.600
And there isn't necessarily a wrong answer if they just want to sit and get what they get.
00:08:15.759 --> 00:08:17.759
It's just gonna be at a different price point.
00:08:17.920 --> 00:08:20.000
And that may be an age factor, Mike.
00:08:20.079 --> 00:08:24.800
It may be I don't have kids in the business, or maybe I do, but maybe they'll figure it out.
00:08:24.879 --> 00:08:26.720
I don't feel like going through the motions of it.
00:08:26.879 --> 00:08:40.159
And so I think you've got those, you got the haves and the have nots, and that there's a gap that's widening pretty high, pretty it's becoming just larger for those that have it and that that don't in terms of the thought process.
00:08:40.399 --> 00:08:49.440
And I think that that's a big issue because when the investors are coming in, they're asking guys like us to say, Michael, does this company have the horsepower to do it?
00:08:49.679 --> 00:08:50.960
Do they need another company?
00:08:51.120 --> 00:08:55.200
Do they need somebody else to help them come up with the AI plan?
00:08:55.440 --> 00:08:59.919
Because we see the writing on the wall, you know, we see where the market's gonna go.
00:09:00.080 --> 00:09:02.879
It's just a question of whether they want to get there or not.
00:09:04.240 --> 00:09:07.360
There appears to be a pressure on those smaller agencies.
00:09:07.519 --> 00:09:13.519
So there's this new mounting pressure for the small agency to be able to manage that margin.
00:09:13.679 --> 00:09:19.600
I think the cost structure of deploying the artificial intelligence is not even necessarily the largest issue.
00:09:19.679 --> 00:09:22.720
It's the technology limitations of their organization.
00:09:23.039 --> 00:09:33.039
It's their old DOS-level system of record, it's the lack of APIs going back and forth, and their inability to move data in a real-time format.
00:09:33.279 --> 00:09:33.360
Yeah.
00:09:33.679 --> 00:09:38.960
The smaller they are, I can imagine the more consolidation pressure they're going to feel in the next 24 months.
00:09:39.279 --> 00:09:39.679
They are.
00:09:39.759 --> 00:09:42.240
But Adam, let me flip that back to you though.
00:09:42.480 --> 00:09:52.480
I feel like if you go back when I got into the industry, you did too 25 years ago, whatever it's been, it was a real pain in the neck to start up a collection agency.
00:09:52.639 --> 00:09:53.759
Think about it, right?
00:09:54.080 --> 00:10:01.679
Starting it out of your garage like some of the people did, and then they hired people and got a phone system, and oh, you needed all these other things.
00:10:02.000 --> 00:10:03.840
Today, I don't know.
00:10:04.080 --> 00:10:12.960
Like there may be things that stop start happening with these 25 to 35-year-old folks that say, collection seems pretty interesting.
00:10:13.120 --> 00:10:23.600
I I could run a business like you know, without having a lot of staff, and I could do it all with AI and inbound calling or inbound like AI with an AI agent.
00:10:23.759 --> 00:10:24.240
I don't know.
00:10:24.559 --> 00:10:32.320
It kind of flips the script on uh like you could have a pretty nice lifestyle business with three or four people and do pretty well.
00:10:32.559 --> 00:10:36.399
And maybe, maybe that's part of what happens to a lot of these smaller guys.
00:10:36.559 --> 00:10:37.279
I don't know.
00:10:37.600 --> 00:10:42.799
But there it there is that thought thought process with tech that's different than what it was before.
00:10:43.039 --> 00:10:45.840
I I think there's a step in there though, right?
00:10:45.919 --> 00:10:53.039
Like there's we've kind of seen a lot of these people from outside the industry come in and say, say that exact same thing.
00:10:53.120 --> 00:10:54.799
I I get offers all the time.
00:10:55.360 --> 00:10:56.000
We'd love to add.
00:10:56.159 --> 00:10:57.279
I'm like, no, no, no, no, no.
00:10:57.360 --> 00:10:59.360
I'm not gonna get out of the startup, no way.
00:11:00.080 --> 00:11:04.879
But I think there is value in collection.
00:11:05.759 --> 00:11:08.080
I I work for a you know huge AI company.
00:11:08.159 --> 00:11:11.039
I don't think it's ever gonna be a hundred percent virtual agent.
00:11:11.120 --> 00:11:12.240
Like, I just don't.
00:11:12.320 --> 00:11:18.159
I I think there's too many uh edge cases, fraud things.
00:11:18.399 --> 00:11:22.879
I mean, are we gonna do mostly AI agents very soon?
00:11:22.960 --> 00:11:28.639
If not, like I have a client who's pretty much it's only inbound, live human beings, right?
00:11:28.720 --> 00:11:33.840
Like that's it, and it's usually like two, three days a week, like that they have the agents available.
00:11:34.080 --> 00:11:35.360
Anthropic agrees with you.
00:11:35.519 --> 00:11:43.120
Anthropic released a report early March where it talked about specifically the impact of artificial intelligence in the labor markets.
00:11:43.279 --> 00:11:46.879
And I've been doing a lot of deep research in this area to try to understand it.
00:11:47.039 --> 00:12:13.519
According to the research that I've been reading, they're saying 80 to 90 percent of those call center type jobs will start moving, except in the instance in which the artificial intelligence is being used to amplify the value of the individual agent, whether that be a co-pilot talking in their ear and feeding them information faster than they can read it on the screen, allowing them to spend more time focused on the customer.
00:12:13.679 --> 00:12:18.720
And I think you're gonna start seeing that not only in the United States, but we're seeing it across the globe.
00:12:19.679 --> 00:12:38.720
Every single company out there, and Michael, you probably get almost as many of them as I do on LinkedIn, but no less than once a week is a new company that's an AI voice, something or other, reach out to me, offer me a seat on the on their strategic board for zero dollars, and ask me for you know a bunch of phone calls.
00:12:38.879 --> 00:12:41.360
Uh, but it's happening at such a fast pace.
00:12:41.600 --> 00:12:43.519
You know, a year ago it was one a month.
00:12:43.679 --> 00:12:49.600
Now it's quite literally at least one a week that I'm hearing from, and you know, they all tell you their story.
00:12:49.759 --> 00:12:58.399
Um, I suggest that maybe they start leading off with the quality of their product and not where they think they're getting their money because it doesn't like that's not actually impressing anybody.
00:12:58.639 --> 00:13:15.679
But we can see that like organizations are trying to take this uh this tool set that they've used in customer service, CX, other places, and think that you're just gonna drop it into the debt collection world and it's gonna function and you're gonna have a fully digital agency.
00:13:16.639 --> 00:13:22.320
Yeah, I don't I don't see the fully digital agency thing happening really anytime soon.
00:13:22.480 --> 00:13:30.639
I still think that you're I think we're going to be able to do more with less by leveraging the technology, but I don't think the people are ever going to disappear.
00:13:31.200 --> 00:13:38.639
I don't even think that to the 80 or 90 percent that Anthropic is estimating, I still don't see it going that far.
00:13:38.960 --> 00:13:39.519
I agree.
00:13:39.600 --> 00:13:41.679
I guys, I I think you're right, Adam.
00:13:41.759 --> 00:13:51.039
I'm just saying when you talk to the agency today and where they're sitting, and then you talk to the investor, the investor wants magic to happen.
00:13:51.120 --> 00:13:52.879
They want the people to disappear.
00:13:53.039 --> 00:13:55.759
And I think we all recognize that's not gonna be the case.
00:13:55.919 --> 00:14:00.799
There's gonna need to be human intervention when when it's needed for certain types of accounts.
00:14:00.960 --> 00:14:06.240
But these A, these, these new startup voice AI companies, and they're all over the place, Adam.
00:14:06.320 --> 00:14:14.960
As you said, I think what's gonna happen with them is they're gonna face an inflection point because they're not gonna be able to go get clients by themselves.
00:14:15.200 --> 00:14:21.440
They're gonna turn into buyers and they're gonna start buying up or trying to buy up the small to mid-sized companies.
00:14:21.519 --> 00:14:22.720
And we're starting to see that.
00:14:22.879 --> 00:14:25.039
They're they're having those conversations.
00:14:25.120 --> 00:14:28.320
They're like, we can't do this without humans.
00:14:28.480 --> 00:14:34.320
We're realizing that now, or at least they're starting to wake up and they need to go in and make an acquisition.
00:14:34.480 --> 00:14:38.159
And I think you'll start to see some of that happen in our industry too.
00:14:38.320 --> 00:14:49.919
Well, you'll have the AI company, the voice AI, and the collections, the third-party collections businesses begin to merge a bit and create some cool platforms that way.
00:14:51.039 --> 00:14:53.120
It becomes the seventh use case of debt collection.
00:14:53.279 --> 00:14:56.000
Mike and I talked about in the first episode of the series.
00:14:56.080 --> 00:15:03.120
You know, there's six use cases of debt collection, and maybe the fully blown AI organization or agency becomes the seventh.
00:15:03.360 --> 00:15:07.360
I never think, but like you said, I don't think it's ever going to be a full-blown AI.
00:15:07.440 --> 00:15:11.440
It's going to be a combo of leveraging the people empowered, right?
00:15:11.519 --> 00:15:13.759
It's going to be people powered by technology.
00:15:13.919 --> 00:15:19.200
Like that seems to be the path that is most realistic for execution.
00:15:19.519 --> 00:15:21.919
Adam, there's so much cost savings in it.
00:15:22.000 --> 00:15:28.799
Like, even again, going back to the 10 to 15%, could you imagine these companies throwing off 30% EBITDA?
00:15:29.120 --> 00:15:31.200
I can, and I think it's going to happen.
00:15:31.360 --> 00:15:39.440
It's just they're going to need to get comfortable, Mike, with technology that companies like yours have and others that are going to help them get there.
00:15:39.600 --> 00:15:41.679
They're not going to be able to build this on their own.
00:15:42.159 --> 00:15:44.960
They're going to need outside assistance to get there.
00:15:45.039 --> 00:15:50.639
And the legacy technology that exists, especially the software, is going to make it challenging.
00:15:50.799 --> 00:15:52.240
But they're going to need to do something.
00:15:52.320 --> 00:15:53.600
They can't just sit and wait.
00:15:54.000 --> 00:15:55.600
There's ways around the leg.
00:15:56.399 --> 00:15:58.480
Like you can build a data lake.
00:15:59.679 --> 00:16:01.679
Every system has a database, right?
00:16:01.919 --> 00:16:06.000
You create, you put that in a cloud, you create a data lake.
00:16:06.639 --> 00:16:10.960
I don't run into and we deal with agencies across the spectrum, right?
00:16:11.039 --> 00:16:14.879
Like blue screens, like somebody's literally in '96.
00:16:14.960 --> 00:16:18.639
When I started, we just were onboarding someone who's using that same system.
00:16:19.279 --> 00:16:23.360
That same homegrown adjusted system.
00:16:23.519 --> 00:16:25.519
You guys could probably guess which one.
00:16:25.919 --> 00:16:27.679
And there's no problem at all, right?
00:16:27.759 --> 00:16:29.039
Like all the data's there.
00:16:29.440 --> 00:16:32.240
It doesn't matter how it presents itself, right?
00:16:32.320 --> 00:16:33.519
Or how you're doing it.
00:16:33.919 --> 00:16:35.759
Companies like me, we build you AI.
00:16:36.080 --> 00:16:38.399
Like, we'll build you APIs, we'll get the data.
00:16:38.480 --> 00:16:39.360
That's no problem.
00:16:39.919 --> 00:16:54.159
I think what a lot of the Silicon Valley money is gonna realize, or or and New York money, like we deal with is a huge part of this whole thing is gonna be data security.
00:16:56.480 --> 00:16:58.720
Where tech is it?
00:16:58.960 --> 00:16:59.759
Is it built?
00:17:00.000 --> 00:17:01.600
Or are they just using other tech?
00:17:02.080 --> 00:17:09.519
You know, are they using open like we've seen I had a conversation with a guy, he's like, I can't get any clients to camp data security.
00:17:09.759 --> 00:17:11.200
And I'm like, well, good luck with that.
00:17:11.279 --> 00:17:13.119
Like you're that's gonna be your problem.
00:17:13.200 --> 00:17:27.920
This is a very especially get the medical debt or yeah, any debt, credit card, that's where I think these small companies come in, realize, oh wow, we didn't think of this, and have a real big problem.
00:17:29.440 --> 00:17:37.519
I I think that's that's where I I I tell people all the time, hey, be careful, we're here.
00:17:38.559 --> 00:17:53.200
I I think you know, a company that has two billion dollars in revenue and is around the world and has servers in any any country you need them is gonna be it's gonna be interesting how this shakes out.
00:17:53.279 --> 00:17:56.720
But I think um I agree with you.
00:17:56.880 --> 00:18:02.400
Building is a tough just keeping up to date with the laws like Adam and I have talked about it a hundred times.
00:18:02.640 --> 00:18:11.680
Yeah, we we we should probably talk about the buy versus build, because I'm curious about how that has an impact on the MA portion of it.
00:18:11.759 --> 00:18:18.319
And I know we we Mike and I did a webinar in it towards the end of 2025 talking about do you buy it, do you build it?
00:18:18.480 --> 00:18:27.359
And we were having that conversation with someone that I truly respect their opinions on artificial intelligence because I feel like he's a deep nerd like me, Mr.
00:18:27.440 --> 00:18:28.400
Tim Collins.
00:18:28.640 --> 00:18:32.000
And we started talking through like what does it actually look like?
00:18:32.160 --> 00:18:45.839
And so as I've been doing this research and trying to better understand the impact of AI on the debt collection industry, it's becoming clear that well, it there's gonna be some pretty significant challenges to being able to build.
00:18:46.000 --> 00:18:54.640
One of the biggest challenges I see is the ability to navigate the ever-changing environment of regulation.
00:18:54.880 --> 00:19:01.599
And now we're seeing individual states start to look at AI privacy-related rules.
00:19:01.680 --> 00:19:14.000
Even the state of Florida, which is generally not a first mover when it comes to over-regulating, has some things on the table within the state Congress looking at these problems.
00:19:14.240 --> 00:19:16.880
But you can't just unwind the model.
00:19:17.279 --> 00:19:19.359
It can't unlearn things easily.
00:19:19.519 --> 00:19:26.160
And if we're going to build, that means that we're going to have to have a staff that's capable of managing that going forward.
00:19:26.319 --> 00:19:30.160
It's not like building a database, and then the database just works, right?
00:19:30.240 --> 00:19:38.079
Either I pay for the database, someone else to build it, I lease somebody else's, or I build it myself, I host it on my server, and now I'm thinking about it as a cost savings.
00:19:38.319 --> 00:19:55.119
But the cost of the AI power, the tokens that are going to have to go into it, the power, the equipment, and then the management of not only the model, but the deployment of that model, I think is going to be a significantly higher cost over the long run than what we're experiencing.
00:19:56.160 --> 00:20:11.440
Adam, there's no when you when an investor or anybody is going into buying a business, they're looking at what the CapEx spend is going to look like in order to ramp up, to build, or are they going to be better off just utilizing off the shelf?
00:20:11.680 --> 00:20:18.799
Most of the time, they're going to see that these companies aren't, they're not trying to be a tech company per se.
00:20:18.960 --> 00:20:22.799
They're trying to be a service business with a tech flavor to them.
00:20:23.119 --> 00:20:33.680
So I feel like whenever we've been involved in helping somebody buy or sell when there's a proprietary or they're build their own system, it ends up at a lower valuation.
00:20:33.839 --> 00:20:43.599
There's exceptions, the rule, like in healthcare, where we've seen proprietary systems be a game changer for somebody that's buying into a into a new market.
00:20:44.000 --> 00:20:48.880
But most cases, the investors are looking for simple plug-and-play.
00:20:49.039 --> 00:20:54.880
And also, if the person that develops the system gets hit by a bus, what do you do then?
00:20:56.559 --> 00:20:57.200
Very good point.
00:20:57.759 --> 00:20:58.880
That's a very good point.
00:20:59.680 --> 00:21:00.799
And it happens, man.
00:21:00.960 --> 00:21:04.240
These types of things are, it's just like I've seen all these kinds of crazy things.
00:21:04.400 --> 00:21:07.759
But I think that's that's a big thing for the investors today.
00:21:07.920 --> 00:21:18.240
Adam, they don't want a lot of guys chasing the the rainbow or the special, you know, the special new technology or the new AI model.
00:21:18.640 --> 00:21:25.920
I think they're more focused on over time figuring out what's the right option for the for the business that could be off the shelf.
00:21:26.400 --> 00:21:40.160
From a deployment standpoint, looking at the TransUnion Deck Collection Industry Report year over year, we saw between 2024 and 2025 a 74% drop in non-AI adopters.
00:21:40.400 --> 00:21:50.319
Three years ago, it was most organizations of or 70 something percent of organizations saying they would never touch artificial intelligence.
00:21:50.400 --> 00:21:51.920
Then we saw a drop in 24.
00:21:52.000 --> 00:21:57.359
We saw a significant now we're down to 7% of companies saying that they're not going to do it.
00:21:57.599 --> 00:22:23.759
And I think that that's a focal point on On the major use cases that are specific to the debt collection industry, not to mention the general business technology options that we now have available to us, HR management, accounting, like all of those kind of general and running of business tools are also going to start to have an impact here in our ability to kind of understand our own organizations in new ways.
00:22:24.240 --> 00:22:36.559
Yeah, I mean, think about all the data that the bench, you know, like I know ACA and some of the other associations have tried these benchmarking reports to show where there's different, where there's similarities.
00:22:36.799 --> 00:22:44.400
The biggest similarity is that most of them don't have a handle on what their true costs are in their business.
00:22:44.559 --> 00:22:46.799
They have a they have a view of it, but they're not.
00:22:46.960 --> 00:23:01.839
Can you imagine all the things that AI from an automation perspective can tell them and diagnose for them about how much they're paying for healthcare, IT costs, accounting, you plug it right in, you could see this is there are much better options.
00:23:02.000 --> 00:23:09.119
Like these little things give you one to two percent in profit that can be applied.
00:23:09.680 --> 00:23:14.480
But I do think the bigger issue, guys, is that when it comes to the clients, right?
00:23:14.640 --> 00:23:27.599
The bank, the you know, the big banks, the fintechs, once they start really allowing the third and first party players to utilize more of the data sets, that's where I think this world really opens up a lot.
00:23:27.759 --> 00:23:34.400
Because I know there's a lot of limitations from the institutional side on what the third-party companies can do.
00:23:34.559 --> 00:23:38.720
But I do think once that changes, I think game on, right?
00:23:38.880 --> 00:23:43.200
Because that way we could really start doing much more than what they're doing today.
00:23:43.519 --> 00:23:45.039
I think that changes start.
00:23:45.440 --> 00:23:46.559
I I really do.
00:23:46.799 --> 00:23:58.960
As they as creditors add tools and get comfortable within themselves, the next step is why isn't my agency network using this, right?
00:23:59.039 --> 00:24:03.680
Like it's you would figure out that that would be the right the right move, right?
00:24:03.920 --> 00:24:04.559
Yeah.
00:24:05.039 --> 00:24:09.119
So, you know, adoption, you know, it's it's coming.
00:24:09.359 --> 00:24:16.319
It's I think creditors jumped in first, and I think that's why agencies didn't jump in.
00:24:16.480 --> 00:24:25.680
Maybe I I think a lot of agencies wanted to go faster, but they were just waiting for a percentage of their volume to allow the use of the tool.
00:24:26.079 --> 00:24:26.480
Right.
00:24:26.640 --> 00:24:34.000
Uh now you see RFPs where the exact same question that uh someone looking to buy an agency is on that RFP.
00:24:34.160 --> 00:24:36.000
What is your strategy for AI?
00:24:36.160 --> 00:24:37.759
Like what tools are you used?
00:24:38.319 --> 00:24:41.279
So I think this is happening now.
00:24:41.440 --> 00:24:47.119
Like I think it is um it's gonna be where they're gonna use it.
00:24:47.279 --> 00:24:52.880
They're either gonna want you to use a different tool or the same tool to evaluate it so they can see that real time.
00:24:53.200 --> 00:25:04.160
I mean, the data points you brought up, Michael, are exactly like you can have that for yourself as a creditor, and then every agency is using you uh on your computer.
00:25:04.319 --> 00:25:07.839
You know, you can open it Christmas morning and say, look, how's everything going?
00:25:08.079 --> 00:25:18.240
You know, like how insane it's and then the data, I agree with you too, is what what investors want to see is what data they have access to, right?
00:25:18.319 --> 00:25:23.519
Like they want to see where they can clean up out of like accounting, waste.
00:25:23.759 --> 00:25:26.000
Do they need thousand square foot call center?
00:25:26.160 --> 00:25:27.759
Everybody works at home, right?
00:25:27.839 --> 00:25:37.599
Like um, I I think there's so much in so many different directions you could go, but i I I think we're gonna see that open up more.
00:25:38.000 --> 00:25:46.160
And then especially I it sounds crazy, but I would love to see a federal law about the use of AI, like AI calling.
00:25:46.480 --> 00:25:47.200
That's gotta change.
00:25:55.680 --> 00:25:56.640
Yes, yeah.
00:25:56.799 --> 00:26:07.680
I mean, God, if if we have, yeah, I mean, having every state with a different role, do you I mean we already deal with that enough with licensing and other state regulatory issues?
00:26:07.759 --> 00:26:11.119
Can you imagine that being it's just gonna lead to more problems, right?
00:26:11.440 --> 00:26:20.799
Like that's I mean, we monitor every state for all these different AI laws that are I mean, sub 21 are on the books just this year.
00:26:20.960 --> 00:26:26.319
Yeah, I mean to see Florida put one up there even for discussion kind of shocked me.
00:26:26.640 --> 00:26:27.680
Adam, should we start it?
00:26:27.839 --> 00:26:34.319
Should we start a new business, an AI licensing, AI privacy compliance business?
00:26:34.640 --> 00:26:38.079
I I think that is a business that's gonna happen, but let me ask you this.
00:26:38.240 --> 00:26:43.839
Let me let me throw a monkey wrench into the the general thought process of artificial intelligence here.
00:26:44.000 --> 00:27:00.000
It was a major fintech organization, I want to say it was about two years ago, at uh at a private event, pointed out that the better they get at AI technology, the more difficult our lives become in the third party and even in the first party space.
00:27:00.160 --> 00:27:18.400
Because if the creditors are able to scale their operations, and if it's if they're comparing their AI to our AI and we're buying these out-of-the-box solutions, where's that differentiation that has always made the debt collection industry special in the reason in which we exist?
00:27:18.720 --> 00:27:21.680
Because it's not just second voice, it's the level of expertise.
00:27:21.839 --> 00:27:27.920
But the more that we start to rely on those models and tool sets, how is that going to be impacted?
00:27:28.079 --> 00:27:42.160
And what is that going to look like for the industry over the next few years as the creditors who have the most amount of money to work with going to impact the rest of us and the profitability of our organizations?
00:27:43.599 --> 00:27:45.759
Adam, that's a hell of a question.
00:27:45.920 --> 00:28:03.519
And I will tell you, I feel like there's always going to be a place for the third party, because at the end of the day, the creditor, the financial institution, whatever it is, the apartment collection company, the property management company, it's not their core competency.
00:28:04.160 --> 00:28:13.920
They don't, it's not like they want to be an expert in third-party collections, even if they've got the best LLM and every tool that you can imagine.
00:28:14.160 --> 00:28:18.960
They ultimately want somebody to help them do it better and more efficiently.
00:28:19.119 --> 00:28:28.720
So I think what they're going to likely do if you go into the future is it's going to get things, easy stuff is going to get dealt with internally.
00:28:29.200 --> 00:28:34.400
Harder stuff where they can't figure it out is probably going to go outsourced.
00:28:34.559 --> 00:28:50.640
And then the question is, the vendors that are there are still operating, they've got to become experts at that and be able to price it accordingly to make their modeling and everything else they do that much better as a result of it.
00:28:51.200 --> 00:28:52.319
I agree with you, Michael.
00:28:52.640 --> 00:28:59.119
I also think, too, like there's still an element of the consumer being able to pay.
00:28:59.279 --> 00:28:59.599
Right?
00:28:59.759 --> 00:29:04.559
Like uh, you know, if you lose your job and that's why you're not paying, right?
00:29:04.640 --> 00:29:12.799
Or let's say you get the double whammy, you you're a construction worker, you you have a medical issue and you cannot work and you're not making money, right?
00:29:12.960 --> 00:29:18.640
Like, it doesn't matter what AI, it doesn't matter how they reach you, it doesn't matter anything.
00:29:18.960 --> 00:29:23.519
You got a family, you got lights, you got food, you're not paying, right?
00:29:23.599 --> 00:29:27.440
Like, so I still think there will always be a space for agencies.
00:29:27.519 --> 00:29:30.160
And I've actually had this conversation with a lot of creditors.
00:29:30.400 --> 00:29:43.839
I think I think we're, you know, the outsource phone calls, that business, especially overseas, because I think because of the labor pressures they're gonna get, creditors are gonna get.
00:29:44.559 --> 00:29:47.440
I think what call centers they have will be in the US.
00:29:47.599 --> 00:29:54.079
The most of the ones I've talked to are are thinking that AI can cover the cost differential.
00:29:54.720 --> 00:30:15.359
Bring it to the US, have actually virtual agents covering as much as they can, different journeys where they can't, they'll send it to the humans, keep that core group, and then I still think third party will always be there because I think the voice is different, Adam, and to Michael's point, there's a different expertise.
00:30:15.599 --> 00:30:18.000
People in this industry know this.
00:30:18.559 --> 00:30:25.839
That is a huge part, like even at Excel, like we don't sell our AI to every type of business, right?
00:30:25.920 --> 00:30:31.680
We we stay in our core industry knowledge, insurance, collections, banking, financial.
00:30:32.079 --> 00:30:36.720
Like, we don't go out to things like government and and go down that road.
00:30:36.960 --> 00:30:41.839
I mean, we could, but it's not our expertise, so there's other people who are doing that, right?
00:30:41.920 --> 00:30:50.480
Like there has to be the management, the expertise level, and I think agencies still do that very, very well.
00:30:50.799 --> 00:30:57.359
And combining the two, those are gonna be the people who win, the expertise with the tech.
00:30:58.559 --> 00:31:11.839
The technology-powered people is where I think the differentiation continues to live because right now there's still opportunities to collect on the low-hanging fruit directly using the artificial intelligence tools.
00:31:12.079 --> 00:31:20.319
But I see that actually decreasing over time rather than increasing, where the increase I see is going to be where we're empowering those people.
00:31:20.480 --> 00:31:26.400
There's still going to be the self-service pieces and the communications, and are we able to get through and are we able to perfect the messaging?
00:31:26.880 --> 00:31:38.960
So I'm not saying that digital communications or anything is going to totally disappear, but I think it'll become more difficult to collect strictly through a non-engaged or a non-human channel.
00:31:39.119 --> 00:31:43.920
But our ability to empower those humans and to still continue to support it.
00:31:44.079 --> 00:31:55.920
Because if you look at some of the creditors, especially if you think about like late 90s, early 2000s, the growth of those agencies was directly related to the banks not wanting to have these giant call centers.
00:31:56.319 --> 00:31:58.240
That was what they were trying to move away from.
00:31:58.400 --> 00:32:10.240
And even if you look at the IPO filing from Jefferson Capital, they talk about how that onshore large call center model is not a long-term sustainable operation.
00:32:10.400 --> 00:32:20.400
And they've been looking at and deploying these other solutions in order to maintain that profitability or to be able to liquidate portfolios.
00:32:21.039 --> 00:32:21.920
There's no doubt.
00:32:47.440 --> 00:32:57.680
And I think their overall model of their work, their workflow has to be enhanced with hopefully the client support in order to get there with data and other things.
00:32:57.839 --> 00:32:58.880
But I do agree with Mike.
00:32:59.039 --> 00:33:01.119
I think it's it's can it's gonna get there.
00:33:01.200 --> 00:33:04.400
It's just gonna be a question of timing and everything else.
00:33:04.960 --> 00:33:19.279
Um the investors that are out there, I think are really interested and intrigued by what this industry, like an AI-powered agency, what it really could do and what the margins could be, and how quickly can it grow?
00:33:19.440 --> 00:33:20.799
And I'm excited by that.
00:33:20.960 --> 00:33:23.200
I think that that brings opportunity for all of us.
00:33:23.440 --> 00:33:36.480
What do you think it looks like over the next 12 to 18 months in terms of deal flow and price differential based on the technology stacks?
00:33:36.640 --> 00:33:42.000
Do you think that that increases faster over the next 18 months and slows down?
00:33:42.079 --> 00:33:45.359
Like how do you see it over 18 and 36 months happen?
00:33:46.240 --> 00:33:56.799
I think you're gonna see uh uh an acceleration of deal activity, not just because of our where our industry is going, but more because of where the economy is.
00:33:57.039 --> 00:34:04.559
I think right now volume is up, profitability is up, people are working, so the unemployment is generally low.
00:34:04.799 --> 00:34:18.639
Those are positives that when someone's coming in and looking at a company in the space, they can see positive EBITDA, they could see growth, and they don't see an economy that's crashing, at least as of today, right?
00:34:18.880 --> 00:34:23.920
So those those are those are gonna drive that's gonna drive MA activity.
00:34:24.000 --> 00:34:27.920
But there's about two or three things I would tell you to keep keep in front of you.
00:34:28.239 --> 00:34:39.199
One is the law firms are getting a lot of attention, the debt collection law firms, they're attracting a lot of investor interest because of how manual intensive their workflows are.
00:34:39.760 --> 00:34:43.199
Can't wait to see what AI does to that side of the market.
00:34:43.599 --> 00:34:46.159
Um second is the government.
00:34:46.320 --> 00:35:01.840
If this student loan thing ever gets figured out, and God knows if it ever will, um, that impact our collections industry, that could also create a huge opportunity for those that collect or service or purchase any kind of student loans.
00:35:02.000 --> 00:35:05.920
There could be a big market opportunity uh that comes from that.
00:35:06.079 --> 00:35:08.159
So I think those are going to be the big ones.
00:35:08.239 --> 00:35:13.119
And then the the last one is what we're talking about here is the technology adoption.
00:35:13.360 --> 00:35:26.480
Because if the investors, buyers see that there's real technology adoption, real change with some of the legacy software, I think that'll drive even more deal activity and interest in the space too.
00:35:27.519 --> 00:35:39.599
Well, as we're in this, the FCC recently released their request for comments related to basically trying to block BPOs, to set an English proficiency standard.
00:35:39.920 --> 00:35:53.119
And a lot of organizations, whether it be agencies, debt buyers, and even law firms, are relying on this offshore work and capability in order to run their businesses.
00:35:53.440 --> 00:35:59.760
So let's pretend in a world in which that actually goes forward and we start blocking this.
00:36:00.320 --> 00:36:11.920
Is there a solution other than the deployment of self-service and artificial intelligence that these organizations could even deploy given that type of a major regulatory shift?
00:36:12.719 --> 00:36:14.480
Adam, it's a game changer.
00:36:14.639 --> 00:36:22.960
And if they if there's rule changes there, I mean, think about what that does to every industry if that changes.
00:36:23.119 --> 00:36:25.119
But it's going to have a significant impact.
00:36:25.199 --> 00:36:38.239
And I'll tell you, the law firm side of it, they're not uh they're not as equipped as a traditional third-party agency to deal with the workflow and the intensity of it onshore and the cost associated with it, too.
00:36:39.119 --> 00:36:45.760
No, the cost associated, especially for those organizations that are operating in a metropolis area, is going to be significantly higher.
00:36:45.920 --> 00:36:56.960
So the only other option that they're going to have to be able to replace that is to start is to perfect their remote employment and to start hiring in more rural areas where the cost structure is not so high.
00:36:57.280 --> 00:36:57.920
Or open out.
00:36:58.159 --> 00:36:58.960
I don't have another solution.
00:36:59.199 --> 00:37:02.480
Right, right there isn't open up somewhere, you know, an office.
00:37:02.880 --> 00:37:08.320
Somewhere, yeah, in in Idaho or wherever you're gonna go, that you're gonna get the cheaper option.
00:37:08.480 --> 00:37:11.360
But still, it it ain't India, right?
00:37:11.519 --> 00:37:12.960
It's not the Philippines.
00:37:13.119 --> 00:37:24.800
So it's gonna, it certainly could be a a significant impact, which is gonna create uh many different dynamics across this industry if something like that takes hold.
00:37:25.760 --> 00:37:31.039
I do think that AI, though, is is creeping into that space more and more.
00:37:31.199 --> 00:37:34.239
It's so fragmented and different rules in every jurisdiction.
00:37:35.519 --> 00:37:40.400
But there is so much process that can be automated in using AI.
00:37:40.719 --> 00:37:46.079
Yeah, there's there's these new AI law firms that are coming out, Mike, that are pretty impressive.
00:37:46.239 --> 00:37:54.400
I don't know how proven they'll be, but it's wild to see that side because collection law firms make up a big chunk of this industry.
00:37:54.719 --> 00:37:55.920
They're all over.
00:37:56.079 --> 00:37:59.679
There's the mom and pops and then they're the large ones.
00:37:59.920 --> 00:38:05.199
But I I think you're gonna see a lot more MA activity there, Adam, in the next 12 to 18 months.
00:38:05.760 --> 00:38:18.480
It would make sense because pushing to new jurisdictions, and we saw it again in the Transunion report, that from a law firm perspective, that's their best opportunity for diversification, is growing their organization into new states.
00:38:18.559 --> 00:38:31.199
And they can either build and hire that expertise and start building out those additional jurisdictions, or they can buy somebody who has minimal crossover with them and continue to expand into new areas.
00:38:31.440 --> 00:38:39.360
But they're that expansion, they're either gonna have to expand into new areas or they're gonna have to expand into new products.
00:38:39.519 --> 00:38:41.519
They're gonna have to expand their expertise.
00:38:41.599 --> 00:38:49.599
And if they've been focused on auto loans, they're gonna have to start doing some landlord tenant and some other things in order to maintain that type of operation long term.
00:38:50.000 --> 00:38:51.920
I think that's the at least that's the result.
00:38:52.400 --> 00:38:57.440
That's what they're reporting back to us in terms of how they're viewing their future expansion.
00:38:57.679 --> 00:39:00.159
Whereas agencies are looking at it differently.
00:39:00.400 --> 00:39:07.840
Agencies are looking at it and saying, well, if I have to be able to diversify my debt collection business, I'm gonna be a BPO service, I'm gonna do customer service.
00:39:08.559 --> 00:39:14.639
They're looking at different services that they can provide on that same infrastructure, that economies of scope, right?
00:39:14.719 --> 00:39:23.840
The same reason that McDonald's sells breakfast, not because they love breakfast, but because they have already got buildings and chefs and they've already got stoves and everything else.
00:39:24.000 --> 00:39:27.039
Why not sell a good product based on that same infrastructure?
00:39:29.119 --> 00:39:30.880
Yeah, there's a McDonald's down the street.
00:39:30.960 --> 00:39:33.599
Just because I'm in Brazil, I'm not on the moon, man.
00:39:34.719 --> 00:39:35.920
All right, all right, there you go.
00:39:42.159 --> 00:39:47.280
I I saw a lot of medical agencies and looking at other streams, right?
00:39:47.440 --> 00:39:50.639
Like, hey, we do this much medical, we gotta find something else, right?
00:39:50.719 --> 00:39:53.280
Like, and you saw them jump into other spaces, you know.
00:39:53.599 --> 00:39:54.960
The industry's nimble, guys.
00:39:55.119 --> 00:40:01.280
I mean, we always see that they ever like, you know, even when we had to go to remote work, people figured it out in 24 hours.
00:40:01.440 --> 00:40:06.320
They'll figure out, Adam, how to deal with these BPO rules if they go into effect.
00:40:06.480 --> 00:40:12.400
We always seem to figure it out, but I will tell you, margin compression is a big issue.
00:40:12.559 --> 00:40:19.519
So if the investors start to see that there isn't much margin here and it can't go up, it's just gonna keep going down.
00:40:19.679 --> 00:40:23.920
Well, we're not gonna have many investors that are gonna be excited about the space, even with AI.
00:40:24.000 --> 00:40:29.199
They're gonna need labor, as we talked about before, and we're gonna need to know that that's gonna be there.
00:40:29.280 --> 00:40:30.960
It just depends on how much.
00:40:32.079 --> 00:40:45.039
We're all expecting the increase in accounts, and if we start to see that decrease in liquidity, the the pressure for that type of technology deployment and cost compression is gonna become essential.
00:40:45.199 --> 00:40:49.760
It's gonna be the lifeblood of keeping these organizations going.
00:40:50.159 --> 00:40:57.760
And it hasn't gotten, it has gotten a little bit better when it comes to being able to hire people, but that's still a really big restriction.
00:40:58.400 --> 00:41:14.639
Working in a collection shop and working for, I mean, it's McDonald's again, any fast food location is paying roughly the same amount in a lot of areas around the United States, but for significantly less responsibility and pressure than being a collector.
00:41:14.800 --> 00:41:18.000
So I think that's the challenge that we're continuing to face.
00:41:18.159 --> 00:41:25.039
I it's gotten a little bit better in terms of being able to hire and retain staff, but it hasn't gotten significantly better.
00:41:25.360 --> 00:41:28.239
I think we're still in a pretty difficult situation as well.
00:41:28.320 --> 00:41:35.840
I think though, if if Adam, if we see like a little bit of increase in unemployment, I think that could be really helpful to the industry, right?
00:41:36.000 --> 00:41:42.800
Because then people are gonna be looking for other job opportunities, not maybe saying, I don't want to do collections.
00:41:42.880 --> 00:41:47.679
They may say, if that's available and I can make it work, I'm gonna go, I'm gonna go do that.
00:41:47.840 --> 00:41:51.760
I think that's something that you know we haven't seen yet in the economy.
00:41:51.840 --> 00:41:58.559
We haven't seen a big or significant uptick in unemployment across the country, but I do think that that could be coming.
00:41:58.719 --> 00:42:01.920
Who knows, depending on where things, where the pendulum shifts.
00:42:02.079 --> 00:42:05.599
But I think that's a big item to keep uh in front of us too.
00:42:06.400 --> 00:42:20.159
So between the creditors, the debt buyers, the agencies, and the law firms, where do you guys think that the biggest opportunity for the deployment of artificial intelligence to drive margin exists?
00:42:22.480 --> 00:42:24.159
You want me to take this one first, Michael?
00:42:24.480 --> 00:42:24.880
Yeah, I think.
00:42:25.039 --> 00:42:27.760
So I think it I think it's agencies, right?
00:42:27.840 --> 00:42:35.599
Um and I think it's hard to say drive mar just because their margins are so tight, right?
00:42:35.679 --> 00:42:46.079
Uh I also think because through performance you can get bigger shares of your clients' champion challenge model, right?
00:42:46.239 --> 00:42:54.320
The first place gets 50%, you know, or whatever that that number is, that that's more revenue, right?
00:42:54.559 --> 00:43:07.280
So it's a quick to move up that chain, get more, get more market share, drive your costs way down, perform better, it's a win-win-win.
00:43:07.440 --> 00:43:10.880
Um where creditors can only do kind of two of those things.
00:43:10.960 --> 00:43:12.159
They have what they have.
00:43:12.320 --> 00:43:19.679
Um they also I think to Michael's point, like they're this industry, like agencies are nimble, right?
00:43:19.760 --> 00:43:22.400
Like, they're good at popping up a new client, right?
00:43:22.480 --> 00:43:26.880
Like they're they're good at increases in volume or do decreases in volume.
00:43:27.599 --> 00:43:38.960
Sometimes creditors have a tougher time with that because they can't just they have different standards, there's different process, you know, their HR.
00:43:39.280 --> 00:43:44.079
It is not as easy for them to hire 50 collectors as a agency.
00:43:44.239 --> 00:43:45.519
It's just not, you know.
00:43:45.760 --> 00:43:59.599
But I think where the creditors kind of have the head start is they have like you we were talking earlier, Adam, like they have money, and they have well, I should say some of them have money, some of them have budget, but they have people.
00:44:00.159 --> 00:44:01.840
Who can research this?
00:44:02.239 --> 00:44:08.400
I think when I talk to agencies, a big thing is they just say, Mike, how do we do this?
00:44:08.480 --> 00:44:08.719
Right?
00:44:08.800 --> 00:44:10.639
Like, how do we go through the process?
00:44:10.800 --> 00:44:12.480
And I'm like, Well, here it is.
00:44:12.639 --> 00:44:17.679
Like, this is you kind of have to map it out where creditors they'll have someone who that's all they do.
00:44:17.760 --> 00:44:22.400
Like their procurement person for digital transformation, they know AI, right?
00:44:22.480 --> 00:44:23.280
Like, they know it.
00:44:23.360 --> 00:44:32.320
They they probably come from an uh a fintech company or or a Silicon Valley company and have had experience using all sorts of AI products.
00:44:32.480 --> 00:44:42.000
So it's good, but I think margin-wise, I think agencies and to Michael's point, I think law firms have a huge opportunity as well.
00:44:42.159 --> 00:44:43.760
Just the scope of their work.
00:44:43.840 --> 00:44:46.960
Like there's so many other products they can use too.
00:44:48.000 --> 00:45:02.480
I'm gonna go with debt buyers, guys, because I will tell you, I feel like the debt buyers are in a really interesting spot because they own the debt, even though they've got data requirements and restrictions that you know too, Adam.
00:45:02.719 --> 00:45:12.400
But they also have the ability to do different things to the to the product and to what they're and how they're servicing it and what accounts they want to do certain things to.
00:45:12.559 --> 00:45:13.920
Do they want to send it offshore?
00:45:14.079 --> 00:45:15.280
Do they want to go near shore?
00:45:15.360 --> 00:45:16.320
Do they want to use bot?
00:45:16.559 --> 00:45:20.480
Do they I feel like they've got better control over cost to collect?
00:45:20.719 --> 00:45:25.440
And um I it's just a different business model altogether, right?
00:45:25.519 --> 00:45:32.800
So we've been seeing a lot of AI guys that have come to us recently being like, do I really want to be a servicer or an agency?
00:45:33.039 --> 00:45:40.400
Maybe I just want to buy paper and create some balance sheet opportunities and amass it that way and then control it from start to finish.
00:45:40.639 --> 00:45:54.719
So I do think because there's more product available in the marketplace, you're gonna start to see companies wanting to either, as an agency, buy some of the paper too, so they could do more things to it and experiment with what they're trying to do with AI.
00:45:55.039 --> 00:46:02.719
Um, and or if they're new entrants expressing interest in debt purchasing and seeing what they could do with cost to collect.
00:46:03.599 --> 00:46:05.760
I think that's a really very interesting.
00:46:06.000 --> 00:46:07.519
I just very interesting.
00:46:08.000 --> 00:46:14.079
Like a new debt buyer who's basically the paper yet, right?
00:46:14.159 --> 00:46:17.519
They said, okay, what can this product do for us?
00:46:17.679 --> 00:46:30.320
And they're like, The difference with it, Mike, is they the problem with it is that it costs, you know, you got to get a bank line of credit, you got to get a facility, a bank financing in order to go buy it at scale.
00:46:30.559 --> 00:46:37.440
Anyone can go buy you know a small little portfolio, pay 50 grand, get some accounts in, see what happens.
00:46:37.599 --> 00:46:42.800
But if you really want to scale it, you could do some pretty cool stuff with the technology too.
00:46:42.960 --> 00:46:46.320
So I think you'll see some new startups there as well that come out.
00:46:47.280 --> 00:46:56.079
Well, from a debt buyer perspective, the number one use case that we saw in the debt collection survey was specifically how to manage the accounts itself.
00:46:56.239 --> 00:47:01.920
It was related to prioritization and workflow management, which I think is interesting.
00:47:02.000 --> 00:47:06.880
And honestly, I I was going to answer the law firms because I was thinking about the pre-litigation opportunities.
00:47:07.039 --> 00:47:12.079
They're the least efficient, the most manual, and that's probably where there's a lot of opportunity.
00:47:12.239 --> 00:47:16.239
But after hearing Michael, I think I have to change my answer on that one because I think you're right.
00:47:16.320 --> 00:47:22.239
I think it is the death buyers that have the largest opportunity because they have the most amount of control over the process.
00:47:22.559 --> 00:47:23.199
They do.
00:47:23.920 --> 00:47:28.320
Well, gentlemen, this has been a fantastic discussion, as always.
00:47:28.480 --> 00:47:33.760
I appreciate you coming on, sharing your insights, and participating in this conversation.
00:47:33.920 --> 00:47:37.199
And as always, I learn a lot every time we talk with you, Mike.
00:47:38.320 --> 00:47:39.199
You guys as well.
00:47:39.280 --> 00:47:40.079
It's always fun.
00:47:40.239 --> 00:47:40.960
Thanks, Adam.
00:47:41.360 --> 00:47:42.079
Thanks, Michael.
00:47:42.320 --> 00:47:43.920
And thank you everybody for watching.
00:47:44.000 --> 00:47:45.440
We appreciate your time and attention.
00:47:45.599 --> 00:47:48.320
If you have questions, leave them in the comments below on LinkedIn and YouTube.
00:47:48.400 --> 00:47:54.000
We'll be responding to those, and we look forward to continuing the discussion on applying AI.
00:47:54.159 --> 00:47:56.079
Thank you, everybody, and we'll see you again soon.
00:00:06.400 --> 00:00:07.919
Hello, everybody.
00:00:08.080 --> 00:00:11.599
Welcome to an episode of Applying AI.
00:00:11.759 --> 00:00:20.559
In applying AI, we cut through the buzzwords and talk about realities of deploying artificial intelligence and financial services and other regulated industries.
00:00:20.800 --> 00:00:34.159
Each month, myself and Mike Walsh will be joined with another guest for us to talk about the practical application of artificial intelligence across the debt collection industry and other financial services.
00:00:34.399 --> 00:00:40.399
Today we've got a great guest with us that most of you or probably all of you recognize, Mr.
00:00:40.560 --> 00:00:49.520
Michael Lamb joining us to talk about the mergers and acquisitions implications of artificial intelligence in our space.
00:00:49.600 --> 00:00:58.640
And for those of you that don't know Michael, if it's mergers and acquisitions and it's in the debt collection-related industry, he's probably involved and engaged.
00:00:58.799 --> 00:01:10.239
So just all of the conversations around how this is going to have a direct impact on these businesses, the value of these businesses, and the changes that we're expecting to see.
00:01:10.400 --> 00:01:12.719
So thank you so much, Michael, for joining us today.
00:01:12.799 --> 00:01:15.439
We really do appreciate you sharing your insights.
00:01:15.840 --> 00:01:16.959
Absolutely, Adam.
00:01:17.200 --> 00:01:18.159
It's great to be here.
00:01:18.239 --> 00:01:20.000
And uh great to see you, Mike.
00:01:20.319 --> 00:01:21.439
That should be fun.
00:01:22.560 --> 00:01:28.400
So high level guys, MA is definitely a hot button in the debt collection industry.
00:01:28.560 --> 00:01:40.959
We've talked about the application of artificial intelligence in the businesses, and it appears that we're going to see a bifurcation of the industry over the coming, let's call it, 24 months.
00:01:41.120 --> 00:01:51.680
And the difference, the gap between the winners and losers of the AI race is going to become significantly more apparent over that time period.
00:01:51.920 --> 00:02:00.079
But, Mike, what are you seeing in the open marketplace and how are buyers and sellers viewing their artificial intelligence stack?
00:02:00.480 --> 00:02:01.840
It's a great question, Adam.
00:02:02.079 --> 00:02:12.639
There are the buyers that are coming in, as I've said to you before in some of our other podcasts, is we've got a whole new crop of buyers, investors that want to come into our industry.
00:02:12.800 --> 00:02:14.960
A lot of them are coming out of Silicon Valley.
00:02:15.120 --> 00:02:25.680
And again, as I think I said to you, Adam, I never my wildest dreams ever thought I was going to be having conversations with Silicon Valley tech investors about doing something in debt collection.
00:02:25.759 --> 00:02:28.639
It just wasn't my first thing that I would expect.
00:02:34.159 --> 00:02:42.800
Number one, for the companies that are in the market, they're trying to figure out when to buy or sell, they need to be able to articulate.
00:02:42.960 --> 00:02:46.319
It doesn't have to be perfect because nobody knows where it's going to go.
00:02:46.639 --> 00:02:48.639
What is their AI strategy?
00:02:48.800 --> 00:02:56.240
How does it affect their line of business, their service line, first and third party, healthcare, financial services?
00:02:56.479 --> 00:02:57.840
What is their strategy?
00:02:58.080 --> 00:03:02.000
Because everyone knows it it needs to exist and be involved in the business.
00:03:02.159 --> 00:03:07.120
It's more a function of how the company wants to apply it in their organization.
00:03:07.360 --> 00:03:23.840
And I'll tell you, the thing that gets the investors most excited is when they see how significant payroll is in the collections industry and when it what that line item means to a lot of companies in our industry, what can you do to reduce it with AI?
00:03:24.000 --> 00:03:28.639
And I've been having a lot of dialogue about that over the past several months.
00:03:29.039 --> 00:03:29.840
You and me both.
00:03:30.000 --> 00:03:30.479
I think it's two.
00:03:30.879 --> 00:03:32.879
Yeah, there's two different pieces to that.
00:03:33.039 --> 00:03:47.680
It's it's the cost savings of deploying artificial intelligence, but with the difficulties that we've had hiring people across the industry, I think we also have to keep an eye on how we can further empower the agents that are on the phone?
00:03:47.840 --> 00:03:51.520
How can we get them engaged and involved in the right conversations?
00:03:51.759 --> 00:03:55.520
We've got this increase in the volume of accounts.
00:03:55.759 --> 00:04:03.120
Everybody is, or the majority of our industry is expecting a double-digit increase in the volume of accounts over the next 12 months.
00:04:03.280 --> 00:04:10.560
But at the same time, you've got about a third of the industry that is witnessing a reduction in liquidity of those accounts.
00:04:10.719 --> 00:04:14.639
So, like always, we're being forced to do more with less.
00:04:14.800 --> 00:04:24.319
And we can either empower our live people, we can automate some processes, we can move consumers more towards the self-service options.
00:04:24.560 --> 00:04:44.879
But I think we're in this weird balance right now where if we don't start, if these organizations don't start moving down the path of identifying which use cases they're going to deploy, they're going to find themselves on the sale block in the not too distant future and not for a value that they expect.
00:04:45.279 --> 00:04:46.240
Exactly, Adam.
00:04:46.319 --> 00:04:48.000
That's the biggest, the biggest issue.
00:04:48.079 --> 00:04:48.959
But Mike, go ahead.
00:04:50.079 --> 00:04:58.079
I was going to say, too, like, I think let's say you're going to go down the sale route, and Michael, should correct me if I'm wrong.
00:04:58.240 --> 00:05:07.360
I think you have to implement some sort of no investor is going to look at it and say, oh, you're dialing and praying that someone picks up the phone.
00:05:07.519 --> 00:05:09.120
Like, I just don't see that.
00:05:09.360 --> 00:05:16.720
I mean, today I talked to an auto lender who just got a uh private equity and they're implementing an AI strategy, right?
00:05:16.800 --> 00:05:18.319
That was the first thing they did.
00:05:18.480 --> 00:05:26.720
Um, then I looked at then I was writing a ROI on our first agency we onboarded.
00:05:26.800 --> 00:05:27.920
It's 9x.
00:05:28.319 --> 00:05:30.079
It's 9x what they're doing.
00:05:30.240 --> 00:05:39.279
And they don't no longer make outbound phone calls to their biggest portfolio, and they've gotten almost 100% of that portfolio from their largest client because it's recovered more.
00:05:39.839 --> 00:05:47.279
They're charging the same price commission base they did, and per account they're paying us you know pennies.
00:05:47.600 --> 00:05:51.600
So, guys, let yeah but but Mike, let's paint the picture here though, right?
00:05:51.839 --> 00:06:01.759
Mo the average margin in this company in this industry is somewhere between 10 and 15 percent EBITDA, an EBITA margin today.
00:06:01.920 --> 00:06:05.759
When you go talk to the investors, they're like, eh, what's that?
00:06:06.000 --> 00:06:07.279
10 to 15 percent?
00:06:07.519 --> 00:06:08.079
Like, really?
00:06:08.160 --> 00:06:11.600
Is that all that's all that's all this industry can produce?
00:06:11.920 --> 00:06:21.199
But then you you actually have the discussion about the AI strategy, Mike, like you're referencing, and that 10 to 15 percent can become 25 to 30 percent?
00:06:21.439 --> 00:06:23.360
Well, now you got people listening.
00:06:23.680 --> 00:06:35.120
You got people saying to themselves, I got something that I can uh that is gonna throw off more cash with less human beings, that's more AI enabled.
00:06:35.439 --> 00:06:38.959
Who the heck wouldn't want that besides all the other stuff, right?
00:06:39.199 --> 00:06:39.759
Right, Michael?
00:06:39.920 --> 00:06:44.160
But it it's more predictable because AI is not doesn't have a bad day, right?
00:06:44.240 --> 00:06:48.639
Like it it's not it's gonna show up, you know, it's not gonna call out, right?
00:06:49.360 --> 00:06:52.639
It's more compliant if you build the right engine, right?
00:06:52.720 --> 00:06:56.079
Like it is it's not gonna swear at anybody, I can tell you that.
00:06:57.360 --> 00:06:58.959
It's not going off the chain, right?
00:06:59.519 --> 00:07:02.720
It's not gonna flip out and drop a bad word.
00:07:03.040 --> 00:07:03.360
Right.
00:07:03.600 --> 00:07:18.160
So I think I mean it I I think the one mistake people in this industry make, and I I'm curious if you see it, is they look at AI as a cost line, not an ROI line, right?
00:07:18.240 --> 00:07:24.720
Like, wait a minute, I'm texting at this rate and I'm sending emails at this rate, why would I pay you to do this?
00:07:25.120 --> 00:07:27.199
Well, how many people are part of that process?
00:07:27.279 --> 00:07:28.720
And what is it really doing?
00:07:28.800 --> 00:07:30.879
You know, like is it generating more?
00:07:30.959 --> 00:07:35.199
Like, are you are you outperforming your your your competitors?
00:07:35.519 --> 00:07:47.600
Um I think that's the biggest hurdle I run into, but when you're out there looking at you know buying or sellers, is that something you run into?
00:07:47.920 --> 00:07:54.800
Yeah, there there's there are guys that are not thinking in the right way, and that's what our job is as a consultancy, right?
00:07:54.879 --> 00:08:02.399
Is to go in and explain to them you can get X today, right, from an EBITDA or multiple perspective.
00:08:02.639 --> 00:08:06.319
You could get Y if you do if you pull some of these levers.
00:08:06.560 --> 00:08:11.279
Do you want to try to pull the levers, or do you want to just sit and get what you're gonna get?
00:08:11.439 --> 00:08:15.600
And there isn't necessarily a wrong answer if they just want to sit and get what they get.
00:08:15.759 --> 00:08:17.759
It's just gonna be at a different price point.
00:08:17.920 --> 00:08:20.000
And that may be an age factor, Mike.
00:08:20.079 --> 00:08:24.800
It may be I don't have kids in the business, or maybe I do, but maybe they'll figure it out.
00:08:24.879 --> 00:08:26.720
I don't feel like going through the motions of it.
00:08:26.879 --> 00:08:40.159
And so I think you've got those, you got the haves and the have nots, and that there's a gap that's widening pretty high, pretty it's becoming just larger for those that have it and that that don't in terms of the thought process.
00:08:40.399 --> 00:08:49.440
And I think that that's a big issue because when the investors are coming in, they're asking guys like us to say, Michael, does this company have the horsepower to do it?
00:08:49.679 --> 00:08:50.960
Do they need another company?
00:08:51.120 --> 00:08:55.200
Do they need somebody else to help them come up with the AI plan?
00:08:55.440 --> 00:08:59.919
Because we see the writing on the wall, you know, we see where the market's gonna go.
00:09:00.080 --> 00:09:02.879
It's just a question of whether they want to get there or not.
00:09:04.240 --> 00:09:07.360
There appears to be a pressure on those smaller agencies.
00:09:07.519 --> 00:09:13.519
So there's this new mounting pressure for the small agency to be able to manage that margin.
00:09:13.679 --> 00:09:19.600
I think the cost structure of deploying the artificial intelligence is not even necessarily the largest issue.
00:09:19.679 --> 00:09:22.720
It's the technology limitations of their organization.
00:09:23.039 --> 00:09:33.039
It's their old DOS-level system of record, it's the lack of APIs going back and forth, and their inability to move data in a real-time format.
00:09:33.279 --> 00:09:33.360
Yeah.
00:09:33.679 --> 00:09:38.960
The smaller they are, I can imagine the more consolidation pressure they're going to feel in the next 24 months.
00:09:39.279 --> 00:09:39.679
They are.
00:09:39.759 --> 00:09:42.240
But Adam, let me flip that back to you though.
00:09:42.480 --> 00:09:52.480
I feel like if you go back when I got into the industry, you did too 25 years ago, whatever it's been, it was a real pain in the neck to start up a collection agency.
00:09:52.639 --> 00:09:53.759
Think about it, right?
00:09:54.080 --> 00:10:01.679
Starting it out of your garage like some of the people did, and then they hired people and got a phone system, and oh, you needed all these other things.
00:10:02.000 --> 00:10:03.840
Today, I don't know.
00:10:04.080 --> 00:10:12.960
Like there may be things that stop start happening with these 25 to 35-year-old folks that say, collection seems pretty interesting.
00:10:13.120 --> 00:10:23.600
I I could run a business like you know, without having a lot of staff, and I could do it all with AI and inbound calling or inbound like AI with an AI agent.
00:10:23.759 --> 00:10:24.240
I don't know.
00:10:24.559 --> 00:10:32.320
It kind of flips the script on uh like you could have a pretty nice lifestyle business with three or four people and do pretty well.
00:10:32.559 --> 00:10:36.399
And maybe, maybe that's part of what happens to a lot of these smaller guys.
00:10:36.559 --> 00:10:37.279
I don't know.
00:10:37.600 --> 00:10:42.799
But there it there is that thought thought process with tech that's different than what it was before.
00:10:43.039 --> 00:10:45.840
I I think there's a step in there though, right?
00:10:45.919 --> 00:10:53.039
Like there's we've kind of seen a lot of these people from outside the industry come in and say, say that exact same thing.
00:10:53.120 --> 00:10:54.799
I I get offers all the time.
00:10:55.360 --> 00:10:56.000
We'd love to add.
00:10:56.159 --> 00:10:57.279
I'm like, no, no, no, no, no.
00:10:57.360 --> 00:10:59.360
I'm not gonna get out of the startup, no way.
00:11:00.080 --> 00:11:04.879
But I think there is value in collection.
00:11:05.759 --> 00:11:08.080
I I work for a you know huge AI company.
00:11:08.159 --> 00:11:11.039
I don't think it's ever gonna be a hundred percent virtual agent.
00:11:11.120 --> 00:11:12.240
Like, I just don't.
00:11:12.320 --> 00:11:18.159
I I think there's too many uh edge cases, fraud things.
00:11:18.399 --> 00:11:22.879
I mean, are we gonna do mostly AI agents very soon?
00:11:22.960 --> 00:11:28.639
If not, like I have a client who's pretty much it's only inbound, live human beings, right?
00:11:28.720 --> 00:11:33.840
Like that's it, and it's usually like two, three days a week, like that they have the agents available.
00:11:34.080 --> 00:11:35.360
Anthropic agrees with you.
00:11:35.519 --> 00:11:43.120
Anthropic released a report early March where it talked about specifically the impact of artificial intelligence in the labor markets.
00:11:43.279 --> 00:11:46.879
And I've been doing a lot of deep research in this area to try to understand it.
00:11:47.039 --> 00:12:13.519
According to the research that I've been reading, they're saying 80 to 90 percent of those call center type jobs will start moving, except in the instance in which the artificial intelligence is being used to amplify the value of the individual agent, whether that be a co-pilot talking in their ear and feeding them information faster than they can read it on the screen, allowing them to spend more time focused on the customer.
00:12:13.679 --> 00:12:18.720
And I think you're gonna start seeing that not only in the United States, but we're seeing it across the globe.
00:12:19.679 --> 00:12:38.720
Every single company out there, and Michael, you probably get almost as many of them as I do on LinkedIn, but no less than once a week is a new company that's an AI voice, something or other, reach out to me, offer me a seat on the on their strategic board for zero dollars, and ask me for you know a bunch of phone calls.
00:12:38.879 --> 00:12:41.360
Uh, but it's happening at such a fast pace.
00:12:41.600 --> 00:12:43.519
You know, a year ago it was one a month.
00:12:43.679 --> 00:12:49.600
Now it's quite literally at least one a week that I'm hearing from, and you know, they all tell you their story.
00:12:49.759 --> 00:12:58.399
Um, I suggest that maybe they start leading off with the quality of their product and not where they think they're getting their money because it doesn't like that's not actually impressing anybody.
00:12:58.639 --> 00:13:15.679
But we can see that like organizations are trying to take this uh this tool set that they've used in customer service, CX, other places, and think that you're just gonna drop it into the debt collection world and it's gonna function and you're gonna have a fully digital agency.
00:13:16.639 --> 00:13:22.320
Yeah, I don't I don't see the fully digital agency thing happening really anytime soon.
00:13:22.480 --> 00:13:30.639
I still think that you're I think we're going to be able to do more with less by leveraging the technology, but I don't think the people are ever going to disappear.
00:13:31.200 --> 00:13:38.639
I don't even think that to the 80 or 90 percent that Anthropic is estimating, I still don't see it going that far.
00:13:38.960 --> 00:13:39.519
I agree.
00:13:39.600 --> 00:13:41.679
I guys, I I think you're right, Adam.
00:13:41.759 --> 00:13:51.039
I'm just saying when you talk to the agency today and where they're sitting, and then you talk to the investor, the investor wants magic to happen.
00:13:51.120 --> 00:13:52.879
They want the people to disappear.
00:13:53.039 --> 00:13:55.759
And I think we all recognize that's not gonna be the case.
00:13:55.919 --> 00:14:00.799
There's gonna need to be human intervention when when it's needed for certain types of accounts.
00:14:00.960 --> 00:14:06.240
But these A, these, these new startup voice AI companies, and they're all over the place, Adam.
00:14:06.320 --> 00:14:14.960
As you said, I think what's gonna happen with them is they're gonna face an inflection point because they're not gonna be able to go get clients by themselves.
00:14:15.200 --> 00:14:21.440
They're gonna turn into buyers and they're gonna start buying up or trying to buy up the small to mid-sized companies.
00:14:21.519 --> 00:14:22.720
And we're starting to see that.
00:14:22.879 --> 00:14:25.039
They're they're having those conversations.
00:14:25.120 --> 00:14:28.320
They're like, we can't do this without humans.
00:14:28.480 --> 00:14:34.320
We're realizing that now, or at least they're starting to wake up and they need to go in and make an acquisition.
00:14:34.480 --> 00:14:38.159
And I think you'll start to see some of that happen in our industry too.
00:14:38.320 --> 00:14:49.919
Well, you'll have the AI company, the voice AI, and the collections, the third-party collections businesses begin to merge a bit and create some cool platforms that way.
00:14:51.039 --> 00:14:53.120
It becomes the seventh use case of debt collection.
00:14:53.279 --> 00:14:56.000
Mike and I talked about in the first episode of the series.
00:14:56.080 --> 00:15:03.120
You know, there's six use cases of debt collection, and maybe the fully blown AI organization or agency becomes the seventh.
00:15:03.360 --> 00:15:07.360
I never think, but like you said, I don't think it's ever going to be a full-blown AI.
00:15:07.440 --> 00:15:11.440
It's going to be a combo of leveraging the people empowered, right?
00:15:11.519 --> 00:15:13.759
It's going to be people powered by technology.
00:15:13.919 --> 00:15:19.200
Like that seems to be the path that is most realistic for execution.
00:15:19.519 --> 00:15:21.919
Adam, there's so much cost savings in it.
00:15:22.000 --> 00:15:28.799
Like, even again, going back to the 10 to 15%, could you imagine these companies throwing off 30% EBITDA?
00:15:29.120 --> 00:15:31.200
I can, and I think it's going to happen.
00:15:31.360 --> 00:15:39.440
It's just they're going to need to get comfortable, Mike, with technology that companies like yours have and others that are going to help them get there.
00:15:39.600 --> 00:15:41.679
They're not going to be able to build this on their own.
00:15:42.159 --> 00:15:44.960
They're going to need outside assistance to get there.
00:15:45.039 --> 00:15:50.639
And the legacy technology that exists, especially the software, is going to make it challenging.
00:15:50.799 --> 00:15:52.240
But they're going to need to do something.
00:15:52.320 --> 00:15:53.600
They can't just sit and wait.
00:15:54.000 --> 00:15:55.600
There's ways around the leg.
00:15:56.399 --> 00:15:58.480
Like you can build a data lake.
00:15:59.679 --> 00:16:01.679
Every system has a database, right?
00:16:01.919 --> 00:16:06.000
You create, you put that in a cloud, you create a data lake.
00:16:06.639 --> 00:16:10.960
I don't run into and we deal with agencies across the spectrum, right?
00:16:11.039 --> 00:16:14.879
Like blue screens, like somebody's literally in '96.
00:16:14.960 --> 00:16:18.639
When I started, we just were onboarding someone who's using that same system.
00:16:19.279 --> 00:16:23.360
That same homegrown adjusted system.
00:16:23.519 --> 00:16:25.519
You guys could probably guess which one.
00:16:25.919 --> 00:16:27.679
And there's no problem at all, right?
00:16:27.759 --> 00:16:29.039
Like all the data's there.
00:16:29.440 --> 00:16:32.240
It doesn't matter how it presents itself, right?
00:16:32.320 --> 00:16:33.519
Or how you're doing it.
00:16:33.919 --> 00:16:35.759
Companies like me, we build you AI.
00:16:36.080 --> 00:16:38.399
Like, we'll build you APIs, we'll get the data.
00:16:38.480 --> 00:16:39.360
That's no problem.
00:16:39.919 --> 00:16:54.159
I think what a lot of the Silicon Valley money is gonna realize, or or and New York money, like we deal with is a huge part of this whole thing is gonna be data security.
00:16:56.480 --> 00:16:58.720
Where tech is it?
00:16:58.960 --> 00:16:59.759
Is it built?
00:17:00.000 --> 00:17:01.600
Or are they just using other tech?
00:17:02.080 --> 00:17:09.519
You know, are they using open like we've seen I had a conversation with a guy, he's like, I can't get any clients to camp data security.
00:17:09.759 --> 00:17:11.200
And I'm like, well, good luck with that.
00:17:11.279 --> 00:17:13.119
Like you're that's gonna be your problem.
00:17:13.200 --> 00:17:27.920
This is a very especially get the medical debt or yeah, any debt, credit card, that's where I think these small companies come in, realize, oh wow, we didn't think of this, and have a real big problem.
00:17:29.440 --> 00:17:37.519
I I think that's that's where I I I tell people all the time, hey, be careful, we're here.
00:17:38.559 --> 00:17:53.200
I I think you know, a company that has two billion dollars in revenue and is around the world and has servers in any any country you need them is gonna be it's gonna be interesting how this shakes out.
00:17:53.279 --> 00:17:56.720
But I think um I agree with you.
00:17:56.880 --> 00:18:02.400
Building is a tough just keeping up to date with the laws like Adam and I have talked about it a hundred times.
00:18:02.640 --> 00:18:11.680
Yeah, we we we should probably talk about the buy versus build, because I'm curious about how that has an impact on the MA portion of it.
00:18:11.759 --> 00:18:18.319
And I know we we Mike and I did a webinar in it towards the end of 2025 talking about do you buy it, do you build it?
00:18:18.480 --> 00:18:27.359
And we were having that conversation with someone that I truly respect their opinions on artificial intelligence because I feel like he's a deep nerd like me, Mr.
00:18:27.440 --> 00:18:28.400
Tim Collins.
00:18:28.640 --> 00:18:32.000
And we started talking through like what does it actually look like?
00:18:32.160 --> 00:18:45.839
And so as I've been doing this research and trying to better understand the impact of AI on the debt collection industry, it's becoming clear that well, it there's gonna be some pretty significant challenges to being able to build.
00:18:46.000 --> 00:18:54.640
One of the biggest challenges I see is the ability to navigate the ever-changing environment of regulation.
00:18:54.880 --> 00:19:01.599
And now we're seeing individual states start to look at AI privacy-related rules.
00:19:01.680 --> 00:19:14.000
Even the state of Florida, which is generally not a first mover when it comes to over-regulating, has some things on the table within the state Congress looking at these problems.
00:19:14.240 --> 00:19:16.880
But you can't just unwind the model.
00:19:17.279 --> 00:19:19.359
It can't unlearn things easily.
00:19:19.519 --> 00:19:26.160
And if we're going to build, that means that we're going to have to have a staff that's capable of managing that going forward.
00:19:26.319 --> 00:19:30.160
It's not like building a database, and then the database just works, right?
00:19:30.240 --> 00:19:38.079
Either I pay for the database, someone else to build it, I lease somebody else's, or I build it myself, I host it on my server, and now I'm thinking about it as a cost savings.
00:19:38.319 --> 00:19:55.119
But the cost of the AI power, the tokens that are going to have to go into it, the power, the equipment, and then the management of not only the model, but the deployment of that model, I think is going to be a significantly higher cost over the long run than what we're experiencing.
00:19:56.160 --> 00:20:11.440
Adam, there's no when you when an investor or anybody is going into buying a business, they're looking at what the CapEx spend is going to look like in order to ramp up, to build, or are they going to be better off just utilizing off the shelf?
00:20:11.680 --> 00:20:18.799
Most of the time, they're going to see that these companies aren't, they're not trying to be a tech company per se.
00:20:18.960 --> 00:20:22.799
They're trying to be a service business with a tech flavor to them.
00:20:23.119 --> 00:20:33.680
So I feel like whenever we've been involved in helping somebody buy or sell when there's a proprietary or they're build their own system, it ends up at a lower valuation.
00:20:33.839 --> 00:20:43.599
There's exceptions, the rule, like in healthcare, where we've seen proprietary systems be a game changer for somebody that's buying into a into a new market.
00:20:44.000 --> 00:20:48.880
But most cases, the investors are looking for simple plug-and-play.
00:20:49.039 --> 00:20:54.880
And also, if the person that develops the system gets hit by a bus, what do you do then?
00:20:56.559 --> 00:20:57.200
Very good point.
00:20:57.759 --> 00:20:58.880
That's a very good point.
00:20:59.680 --> 00:21:00.799
And it happens, man.
00:21:00.960 --> 00:21:04.240
These types of things are, it's just like I've seen all these kinds of crazy things.
00:21:04.400 --> 00:21:07.759
But I think that's that's a big thing for the investors today.
00:21:07.920 --> 00:21:18.240
Adam, they don't want a lot of guys chasing the the rainbow or the special, you know, the special new technology or the new AI model.
00:21:18.640 --> 00:21:25.920
I think they're more focused on over time figuring out what's the right option for the for the business that could be off the shelf.
00:21:26.400 --> 00:21:40.160
From a deployment standpoint, looking at the TransUnion Deck Collection Industry Report year over year, we saw between 2024 and 2025 a 74% drop in non-AI adopters.
00:21:40.400 --> 00:21:50.319
Three years ago, it was most organizations of or 70 something percent of organizations saying they would never touch artificial intelligence.
00:21:50.400 --> 00:21:51.920
Then we saw a drop in 24.
00:21:52.000 --> 00:21:57.359
We saw a significant now we're down to 7% of companies saying that they're not going to do it.
00:21:57.599 --> 00:22:23.759
And I think that that's a focal point on On the major use cases that are specific to the debt collection industry, not to mention the general business technology options that we now have available to us, HR management, accounting, like all of those kind of general and running of business tools are also going to start to have an impact here in our ability to kind of understand our own organizations in new ways.
00:22:24.240 --> 00:22:36.559
Yeah, I mean, think about all the data that the bench, you know, like I know ACA and some of the other associations have tried these benchmarking reports to show where there's different, where there's similarities.
00:22:36.799 --> 00:22:44.400
The biggest similarity is that most of them don't have a handle on what their true costs are in their business.
00:22:44.559 --> 00:22:46.799
They have a they have a view of it, but they're not.
00:22:46.960 --> 00:23:01.839
Can you imagine all the things that AI from an automation perspective can tell them and diagnose for them about how much they're paying for healthcare, IT costs, accounting, you plug it right in, you could see this is there are much better options.
00:23:02.000 --> 00:23:09.119
Like these little things give you one to two percent in profit that can be applied.
00:23:09.680 --> 00:23:14.480
But I do think the bigger issue, guys, is that when it comes to the clients, right?
00:23:14.640 --> 00:23:27.599
The bank, the you know, the big banks, the fintechs, once they start really allowing the third and first party players to utilize more of the data sets, that's where I think this world really opens up a lot.
00:23:27.759 --> 00:23:34.400
Because I know there's a lot of limitations from the institutional side on what the third-party companies can do.
00:23:34.559 --> 00:23:38.720
But I do think once that changes, I think game on, right?
00:23:38.880 --> 00:23:43.200
Because that way we could really start doing much more than what they're doing today.
00:23:43.519 --> 00:23:45.039
I think that changes start.
00:23:45.440 --> 00:23:46.559
I I really do.
00:23:46.799 --> 00:23:58.960
As they as creditors add tools and get comfortable within themselves, the next step is why isn't my agency network using this, right?
00:23:59.039 --> 00:24:03.680
Like it's you would figure out that that would be the right the right move, right?
00:24:03.920 --> 00:24:04.559
Yeah.
00:24:05.039 --> 00:24:09.119
So, you know, adoption, you know, it's it's coming.
00:24:09.359 --> 00:24:16.319
It's I think creditors jumped in first, and I think that's why agencies didn't jump in.
00:24:16.480 --> 00:24:25.680
Maybe I I think a lot of agencies wanted to go faster, but they were just waiting for a percentage of their volume to allow the use of the tool.
00:24:26.079 --> 00:24:26.480
Right.
00:24:26.640 --> 00:24:34.000
Uh now you see RFPs where the exact same question that uh someone looking to buy an agency is on that RFP.
00:24:34.160 --> 00:24:36.000
What is your strategy for AI?
00:24:36.160 --> 00:24:37.759
Like what tools are you used?
00:24:38.319 --> 00:24:41.279
So I think this is happening now.
00:24:41.440 --> 00:24:47.119
Like I think it is um it's gonna be where they're gonna use it.
00:24:47.279 --> 00:24:52.880
They're either gonna want you to use a different tool or the same tool to evaluate it so they can see that real time.
00:24:53.200 --> 00:25:04.160
I mean, the data points you brought up, Michael, are exactly like you can have that for yourself as a creditor, and then every agency is using you uh on your computer.
00:25:04.319 --> 00:25:07.839
You know, you can open it Christmas morning and say, look, how's everything going?
00:25:08.079 --> 00:25:18.240
You know, like how insane it's and then the data, I agree with you too, is what what investors want to see is what data they have access to, right?
00:25:18.319 --> 00:25:23.519
Like they want to see where they can clean up out of like accounting, waste.
00:25:23.759 --> 00:25:26.000
Do they need thousand square foot call center?
00:25:26.160 --> 00:25:27.759
Everybody works at home, right?
00:25:27.839 --> 00:25:37.599
Like um, I I think there's so much in so many different directions you could go, but i I I think we're gonna see that open up more.
00:25:38.000 --> 00:25:46.160
And then especially I it sounds crazy, but I would love to see a federal law about the use of AI, like AI calling.
00:25:46.480 --> 00:25:47.200
That's gotta change.
00:25:55.680 --> 00:25:56.640
Yes, yeah.
00:25:56.799 --> 00:26:07.680
I mean, God, if if we have, yeah, I mean, having every state with a different role, do you I mean we already deal with that enough with licensing and other state regulatory issues?
00:26:07.759 --> 00:26:11.119
Can you imagine that being it's just gonna lead to more problems, right?
00:26:11.440 --> 00:26:20.799
Like that's I mean, we monitor every state for all these different AI laws that are I mean, sub 21 are on the books just this year.
00:26:20.960 --> 00:26:26.319
Yeah, I mean to see Florida put one up there even for discussion kind of shocked me.
00:26:26.640 --> 00:26:27.680
Adam, should we start it?
00:26:27.839 --> 00:26:34.319
Should we start a new business, an AI licensing, AI privacy compliance business?
00:26:34.640 --> 00:26:38.079
I I think that is a business that's gonna happen, but let me ask you this.
00:26:38.240 --> 00:26:43.839
Let me let me throw a monkey wrench into the the general thought process of artificial intelligence here.
00:26:44.000 --> 00:27:00.000
It was a major fintech organization, I want to say it was about two years ago, at uh at a private event, pointed out that the better they get at AI technology, the more difficult our lives become in the third party and even in the first party space.
00:27:00.160 --> 00:27:18.400
Because if the creditors are able to scale their operations, and if it's if they're comparing their AI to our AI and we're buying these out-of-the-box solutions, where's that differentiation that has always made the debt collection industry special in the reason in which we exist?
00:27:18.720 --> 00:27:21.680
Because it's not just second voice, it's the level of expertise.
00:27:21.839 --> 00:27:27.920
But the more that we start to rely on those models and tool sets, how is that going to be impacted?
00:27:28.079 --> 00:27:42.160
And what is that going to look like for the industry over the next few years as the creditors who have the most amount of money to work with going to impact the rest of us and the profitability of our organizations?
00:27:43.599 --> 00:27:45.759
Adam, that's a hell of a question.
00:27:45.920 --> 00:28:03.519
And I will tell you, I feel like there's always going to be a place for the third party, because at the end of the day, the creditor, the financial institution, whatever it is, the apartment collection company, the property management company, it's not their core competency.
00:28:04.160 --> 00:28:13.920
They don't, it's not like they want to be an expert in third-party collections, even if they've got the best LLM and every tool that you can imagine.
00:28:14.160 --> 00:28:18.960
They ultimately want somebody to help them do it better and more efficiently.
00:28:19.119 --> 00:28:28.720
So I think what they're going to likely do if you go into the future is it's going to get things, easy stuff is going to get dealt with internally.
00:28:29.200 --> 00:28:34.400
Harder stuff where they can't figure it out is probably going to go outsourced.
00:28:34.559 --> 00:28:50.640
And then the question is, the vendors that are there are still operating, they've got to become experts at that and be able to price it accordingly to make their modeling and everything else they do that much better as a result of it.
00:28:51.200 --> 00:28:52.319
I agree with you, Michael.
00:28:52.640 --> 00:28:59.119
I also think, too, like there's still an element of the consumer being able to pay.
00:28:59.279 --> 00:28:59.599
Right?
00:28:59.759 --> 00:29:04.559
Like uh, you know, if you lose your job and that's why you're not paying, right?
00:29:04.640 --> 00:29:12.799
Or let's say you get the double whammy, you you're a construction worker, you you have a medical issue and you cannot work and you're not making money, right?
00:29:12.960 --> 00:29:18.640
Like, it doesn't matter what AI, it doesn't matter how they reach you, it doesn't matter anything.
00:29:18.960 --> 00:29:23.519
You got a family, you got lights, you got food, you're not paying, right?
00:29:23.599 --> 00:29:27.440
Like, so I still think there will always be a space for agencies.
00:29:27.519 --> 00:29:30.160
And I've actually had this conversation with a lot of creditors.
00:29:30.400 --> 00:29:43.839
I think I think we're, you know, the outsource phone calls, that business, especially overseas, because I think because of the labor pressures they're gonna get, creditors are gonna get.
00:29:44.559 --> 00:29:47.440
I think what call centers they have will be in the US.
00:29:47.599 --> 00:29:54.079
The most of the ones I've talked to are are thinking that AI can cover the cost differential.
00:29:54.720 --> 00:30:15.359
Bring it to the US, have actually virtual agents covering as much as they can, different journeys where they can't, they'll send it to the humans, keep that core group, and then I still think third party will always be there because I think the voice is different, Adam, and to Michael's point, there's a different expertise.
00:30:15.599 --> 00:30:18.000
People in this industry know this.
00:30:18.559 --> 00:30:25.839
That is a huge part, like even at Excel, like we don't sell our AI to every type of business, right?
00:30:25.920 --> 00:30:31.680
We we stay in our core industry knowledge, insurance, collections, banking, financial.
00:30:32.079 --> 00:30:36.720
Like, we don't go out to things like government and and go down that road.
00:30:36.960 --> 00:30:41.839
I mean, we could, but it's not our expertise, so there's other people who are doing that, right?
00:30:41.920 --> 00:30:50.480
Like there has to be the management, the expertise level, and I think agencies still do that very, very well.
00:30:50.799 --> 00:30:57.359
And combining the two, those are gonna be the people who win, the expertise with the tech.
00:30:58.559 --> 00:31:11.839
The technology-powered people is where I think the differentiation continues to live because right now there's still opportunities to collect on the low-hanging fruit directly using the artificial intelligence tools.
00:31:12.079 --> 00:31:20.319
But I see that actually decreasing over time rather than increasing, where the increase I see is going to be where we're empowering those people.
00:31:20.480 --> 00:31:26.400
There's still going to be the self-service pieces and the communications, and are we able to get through and are we able to perfect the messaging?
00:31:26.880 --> 00:31:38.960
So I'm not saying that digital communications or anything is going to totally disappear, but I think it'll become more difficult to collect strictly through a non-engaged or a non-human channel.
00:31:39.119 --> 00:31:43.920
But our ability to empower those humans and to still continue to support it.
00:31:44.079 --> 00:31:55.920
Because if you look at some of the creditors, especially if you think about like late 90s, early 2000s, the growth of those agencies was directly related to the banks not wanting to have these giant call centers.
00:31:56.319 --> 00:31:58.240
That was what they were trying to move away from.
00:31:58.400 --> 00:32:10.240
And even if you look at the IPO filing from Jefferson Capital, they talk about how that onshore large call center model is not a long-term sustainable operation.
00:32:10.400 --> 00:32:20.400
And they've been looking at and deploying these other solutions in order to maintain that profitability or to be able to liquidate portfolios.
00:32:21.039 --> 00:32:21.920
There's no doubt.
00:32:47.440 --> 00:32:57.680
And I think their overall model of their work, their workflow has to be enhanced with hopefully the client support in order to get there with data and other things.
00:32:57.839 --> 00:32:58.880
But I do agree with Mike.
00:32:59.039 --> 00:33:01.119
I think it's it's can it's gonna get there.
00:33:01.200 --> 00:33:04.400
It's just gonna be a question of timing and everything else.
00:33:04.960 --> 00:33:19.279
Um the investors that are out there, I think are really interested and intrigued by what this industry, like an AI-powered agency, what it really could do and what the margins could be, and how quickly can it grow?
00:33:19.440 --> 00:33:20.799
And I'm excited by that.
00:33:20.960 --> 00:33:23.200
I think that that brings opportunity for all of us.
00:33:23.440 --> 00:33:36.480
What do you think it looks like over the next 12 to 18 months in terms of deal flow and price differential based on the technology stacks?
00:33:36.640 --> 00:33:42.000
Do you think that that increases faster over the next 18 months and slows down?
00:33:42.079 --> 00:33:45.359
Like how do you see it over 18 and 36 months happen?
00:33:46.240 --> 00:33:56.799
I think you're gonna see uh uh an acceleration of deal activity, not just because of our where our industry is going, but more because of where the economy is.
00:33:57.039 --> 00:34:04.559
I think right now volume is up, profitability is up, people are working, so the unemployment is generally low.
00:34:04.799 --> 00:34:18.639
Those are positives that when someone's coming in and looking at a company in the space, they can see positive EBITDA, they could see growth, and they don't see an economy that's crashing, at least as of today, right?
00:34:18.880 --> 00:34:23.920
So those those are those are gonna drive that's gonna drive MA activity.
00:34:24.000 --> 00:34:27.920
But there's about two or three things I would tell you to keep keep in front of you.
00:34:28.239 --> 00:34:39.199
One is the law firms are getting a lot of attention, the debt collection law firms, they're attracting a lot of investor interest because of how manual intensive their workflows are.
00:34:39.760 --> 00:34:43.199
Can't wait to see what AI does to that side of the market.
00:34:43.599 --> 00:34:46.159
Um second is the government.
00:34:46.320 --> 00:35:01.840
If this student loan thing ever gets figured out, and God knows if it ever will, um, that impact our collections industry, that could also create a huge opportunity for those that collect or service or purchase any kind of student loans.
00:35:02.000 --> 00:35:05.920
There could be a big market opportunity uh that comes from that.
00:35:06.079 --> 00:35:08.159
So I think those are going to be the big ones.
00:35:08.239 --> 00:35:13.119
And then the the last one is what we're talking about here is the technology adoption.
00:35:13.360 --> 00:35:26.480
Because if the investors, buyers see that there's real technology adoption, real change with some of the legacy software, I think that'll drive even more deal activity and interest in the space too.
00:35:27.519 --> 00:35:39.599
Well, as we're in this, the FCC recently released their request for comments related to basically trying to block BPOs, to set an English proficiency standard.
00:35:39.920 --> 00:35:53.119
And a lot of organizations, whether it be agencies, debt buyers, and even law firms, are relying on this offshore work and capability in order to run their businesses.
00:35:53.440 --> 00:35:59.760
So let's pretend in a world in which that actually goes forward and we start blocking this.
00:36:00.320 --> 00:36:11.920
Is there a solution other than the deployment of self-service and artificial intelligence that these organizations could even deploy given that type of a major regulatory shift?
00:36:12.719 --> 00:36:14.480
Adam, it's a game changer.
00:36:14.639 --> 00:36:22.960
And if they if there's rule changes there, I mean, think about what that does to every industry if that changes.
00:36:23.119 --> 00:36:25.119
But it's going to have a significant impact.
00:36:25.199 --> 00:36:38.239
And I'll tell you, the law firm side of it, they're not uh they're not as equipped as a traditional third-party agency to deal with the workflow and the intensity of it onshore and the cost associated with it, too.
00:36:39.119 --> 00:36:45.760
No, the cost associated, especially for those organizations that are operating in a metropolis area, is going to be significantly higher.
00:36:45.920 --> 00:36:56.960
So the only other option that they're going to have to be able to replace that is to start is to perfect their remote employment and to start hiring in more rural areas where the cost structure is not so high.
00:36:57.280 --> 00:36:57.920
Or open out.
00:36:58.159 --> 00:36:58.960
I don't have another solution.
00:36:59.199 --> 00:37:02.480
Right, right there isn't open up somewhere, you know, an office.
00:37:02.880 --> 00:37:08.320
Somewhere, yeah, in in Idaho or wherever you're gonna go, that you're gonna get the cheaper option.
00:37:08.480 --> 00:37:11.360
But still, it it ain't India, right?
00:37:11.519 --> 00:37:12.960
It's not the Philippines.
00:37:13.119 --> 00:37:24.800
So it's gonna, it certainly could be a a significant impact, which is gonna create uh many different dynamics across this industry if something like that takes hold.
00:37:25.760 --> 00:37:31.039
I do think that AI, though, is is creeping into that space more and more.
00:37:31.199 --> 00:37:34.239
It's so fragmented and different rules in every jurisdiction.
00:37:35.519 --> 00:37:40.400
But there is so much process that can be automated in using AI.
00:37:40.719 --> 00:37:46.079
Yeah, there's there's these new AI law firms that are coming out, Mike, that are pretty impressive.
00:37:46.239 --> 00:37:54.400
I don't know how proven they'll be, but it's wild to see that side because collection law firms make up a big chunk of this industry.
00:37:54.719 --> 00:37:55.920
They're all over.
00:37:56.079 --> 00:37:59.679
There's the mom and pops and then they're the large ones.
00:37:59.920 --> 00:38:05.199
But I I think you're gonna see a lot more MA activity there, Adam, in the next 12 to 18 months.
00:38:05.760 --> 00:38:18.480
It would make sense because pushing to new jurisdictions, and we saw it again in the Transunion report, that from a law firm perspective, that's their best opportunity for diversification, is growing their organization into new states.
00:38:18.559 --> 00:38:31.199
And they can either build and hire that expertise and start building out those additional jurisdictions, or they can buy somebody who has minimal crossover with them and continue to expand into new areas.
00:38:31.440 --> 00:38:39.360
But they're that expansion, they're either gonna have to expand into new areas or they're gonna have to expand into new products.
00:38:39.519 --> 00:38:41.519
They're gonna have to expand their expertise.
00:38:41.599 --> 00:38:49.599
And if they've been focused on auto loans, they're gonna have to start doing some landlord tenant and some other things in order to maintain that type of operation long term.
00:38:50.000 --> 00:38:51.920
I think that's the at least that's the result.
00:38:52.400 --> 00:38:57.440
That's what they're reporting back to us in terms of how they're viewing their future expansion.
00:38:57.679 --> 00:39:00.159
Whereas agencies are looking at it differently.
00:39:00.400 --> 00:39:07.840
Agencies are looking at it and saying, well, if I have to be able to diversify my debt collection business, I'm gonna be a BPO service, I'm gonna do customer service.
00:39:08.559 --> 00:39:14.639
They're looking at different services that they can provide on that same infrastructure, that economies of scope, right?
00:39:14.719 --> 00:39:23.840
The same reason that McDonald's sells breakfast, not because they love breakfast, but because they have already got buildings and chefs and they've already got stoves and everything else.
00:39:24.000 --> 00:39:27.039
Why not sell a good product based on that same infrastructure?
00:39:29.119 --> 00:39:30.880
Yeah, there's a McDonald's down the street.
00:39:30.960 --> 00:39:33.599
Just because I'm in Brazil, I'm not on the moon, man.
00:39:34.719 --> 00:39:35.920
All right, all right, there you go.
00:39:42.159 --> 00:39:47.280
I I saw a lot of medical agencies and looking at other streams, right?
00:39:47.440 --> 00:39:50.639
Like, hey, we do this much medical, we gotta find something else, right?
00:39:50.719 --> 00:39:53.280
Like, and you saw them jump into other spaces, you know.
00:39:53.599 --> 00:39:54.960
The industry's nimble, guys.
00:39:55.119 --> 00:40:01.280
I mean, we always see that they ever like, you know, even when we had to go to remote work, people figured it out in 24 hours.
00:40:01.440 --> 00:40:06.320
They'll figure out, Adam, how to deal with these BPO rules if they go into effect.
00:40:06.480 --> 00:40:12.400
We always seem to figure it out, but I will tell you, margin compression is a big issue.
00:40:12.559 --> 00:40:19.519
So if the investors start to see that there isn't much margin here and it can't go up, it's just gonna keep going down.
00:40:19.679 --> 00:40:23.920
Well, we're not gonna have many investors that are gonna be excited about the space, even with AI.
00:40:24.000 --> 00:40:29.199
They're gonna need labor, as we talked about before, and we're gonna need to know that that's gonna be there.
00:40:29.280 --> 00:40:30.960
It just depends on how much.
00:40:32.079 --> 00:40:45.039
We're all expecting the increase in accounts, and if we start to see that decrease in liquidity, the the pressure for that type of technology deployment and cost compression is gonna become essential.
00:40:45.199 --> 00:40:49.760
It's gonna be the lifeblood of keeping these organizations going.
00:40:50.159 --> 00:40:57.760
And it hasn't gotten, it has gotten a little bit better when it comes to being able to hire people, but that's still a really big restriction.
00:40:58.400 --> 00:41:14.639
Working in a collection shop and working for, I mean, it's McDonald's again, any fast food location is paying roughly the same amount in a lot of areas around the United States, but for significantly less responsibility and pressure than being a collector.
00:41:14.800 --> 00:41:18.000
So I think that's the challenge that we're continuing to face.
00:41:18.159 --> 00:41:25.039
I it's gotten a little bit better in terms of being able to hire and retain staff, but it hasn't gotten significantly better.
00:41:25.360 --> 00:41:28.239
I think we're still in a pretty difficult situation as well.
00:41:28.320 --> 00:41:35.840
I think though, if if Adam, if we see like a little bit of increase in unemployment, I think that could be really helpful to the industry, right?
00:41:36.000 --> 00:41:42.800
Because then people are gonna be looking for other job opportunities, not maybe saying, I don't want to do collections.
00:41:42.880 --> 00:41:47.679
They may say, if that's available and I can make it work, I'm gonna go, I'm gonna go do that.
00:41:47.840 --> 00:41:51.760
I think that's something that you know we haven't seen yet in the economy.
00:41:51.840 --> 00:41:58.559
We haven't seen a big or significant uptick in unemployment across the country, but I do think that that could be coming.
00:41:58.719 --> 00:42:01.920
Who knows, depending on where things, where the pendulum shifts.
00:42:02.079 --> 00:42:05.599
But I think that's a big item to keep uh in front of us too.
00:42:06.400 --> 00:42:20.159
So between the creditors, the debt buyers, the agencies, and the law firms, where do you guys think that the biggest opportunity for the deployment of artificial intelligence to drive margin exists?
00:42:22.480 --> 00:42:24.159
You want me to take this one first, Michael?
00:42:24.480 --> 00:42:24.880
Yeah, I think.
00:42:25.039 --> 00:42:27.760
So I think it I think it's agencies, right?
00:42:27.840 --> 00:42:35.599
Um and I think it's hard to say drive mar just because their margins are so tight, right?
00:42:35.679 --> 00:42:46.079
Uh I also think because through performance you can get bigger shares of your clients' champion challenge model, right?
00:42:46.239 --> 00:42:54.320
The first place gets 50%, you know, or whatever that that number is, that that's more revenue, right?
00:42:54.559 --> 00:43:07.280
So it's a quick to move up that chain, get more, get more market share, drive your costs way down, perform better, it's a win-win-win.
00:43:07.440 --> 00:43:10.880
Um where creditors can only do kind of two of those things.
00:43:10.960 --> 00:43:12.159
They have what they have.
00:43:12.320 --> 00:43:19.679
Um they also I think to Michael's point, like they're this industry, like agencies are nimble, right?
00:43:19.760 --> 00:43:22.400
Like, they're good at popping up a new client, right?
00:43:22.480 --> 00:43:26.880
Like they're they're good at increases in volume or do decreases in volume.
00:43:27.599 --> 00:43:38.960
Sometimes creditors have a tougher time with that because they can't just they have different standards, there's different process, you know, their HR.
00:43:39.280 --> 00:43:44.079
It is not as easy for them to hire 50 collectors as a agency.
00:43:44.239 --> 00:43:45.519
It's just not, you know.
00:43:45.760 --> 00:43:59.599
But I think where the creditors kind of have the head start is they have like you we were talking earlier, Adam, like they have money, and they have well, I should say some of them have money, some of them have budget, but they have people.
00:44:00.159 --> 00:44:01.840
Who can research this?
00:44:02.239 --> 00:44:08.400
I think when I talk to agencies, a big thing is they just say, Mike, how do we do this?
00:44:08.480 --> 00:44:08.719
Right?
00:44:08.800 --> 00:44:10.639
Like, how do we go through the process?
00:44:10.800 --> 00:44:12.480
And I'm like, Well, here it is.
00:44:12.639 --> 00:44:17.679
Like, this is you kind of have to map it out where creditors they'll have someone who that's all they do.
00:44:17.760 --> 00:44:22.400
Like their procurement person for digital transformation, they know AI, right?
00:44:22.480 --> 00:44:23.280
Like, they know it.
00:44:23.360 --> 00:44:32.320
They they probably come from an uh a fintech company or or a Silicon Valley company and have had experience using all sorts of AI products.
00:44:32.480 --> 00:44:42.000
So it's good, but I think margin-wise, I think agencies and to Michael's point, I think law firms have a huge opportunity as well.
00:44:42.159 --> 00:44:43.760
Just the scope of their work.
00:44:43.840 --> 00:44:46.960
Like there's so many other products they can use too.
00:44:48.000 --> 00:45:02.480
I'm gonna go with debt buyers, guys, because I will tell you, I feel like the debt buyers are in a really interesting spot because they own the debt, even though they've got data requirements and restrictions that you know too, Adam.
00:45:02.719 --> 00:45:12.400
But they also have the ability to do different things to the to the product and to what they're and how they're servicing it and what accounts they want to do certain things to.
00:45:12.559 --> 00:45:13.920
Do they want to send it offshore?
00:45:14.079 --> 00:45:15.280
Do they want to go near shore?
00:45:15.360 --> 00:45:16.320
Do they want to use bot?
00:45:16.559 --> 00:45:20.480
Do they I feel like they've got better control over cost to collect?
00:45:20.719 --> 00:45:25.440
And um I it's just a different business model altogether, right?
00:45:25.519 --> 00:45:32.800
So we've been seeing a lot of AI guys that have come to us recently being like, do I really want to be a servicer or an agency?
00:45:33.039 --> 00:45:40.400
Maybe I just want to buy paper and create some balance sheet opportunities and amass it that way and then control it from start to finish.
00:45:40.639 --> 00:45:54.719
So I do think because there's more product available in the marketplace, you're gonna start to see companies wanting to either, as an agency, buy some of the paper too, so they could do more things to it and experiment with what they're trying to do with AI.
00:45:55.039 --> 00:46:02.719
Um, and or if they're new entrants expressing interest in debt purchasing and seeing what they could do with cost to collect.
00:46:03.599 --> 00:46:05.760
I think that's a really very interesting.
00:46:06.000 --> 00:46:07.519
I just very interesting.
00:46:08.000 --> 00:46:14.079
Like a new debt buyer who's basically the paper yet, right?
00:46:14.159 --> 00:46:17.519
They said, okay, what can this product do for us?
00:46:17.679 --> 00:46:30.320
And they're like, The difference with it, Mike, is they the problem with it is that it costs, you know, you got to get a bank line of credit, you got to get a facility, a bank financing in order to go buy it at scale.
00:46:30.559 --> 00:46:37.440
Anyone can go buy you know a small little portfolio, pay 50 grand, get some accounts in, see what happens.
00:46:37.599 --> 00:46:42.800
But if you really want to scale it, you could do some pretty cool stuff with the technology too.
00:46:42.960 --> 00:46:46.320
So I think you'll see some new startups there as well that come out.
00:46:47.280 --> 00:46:56.079
Well, from a debt buyer perspective, the number one use case that we saw in the debt collection survey was specifically how to manage the accounts itself.
00:46:56.239 --> 00:47:01.920
It was related to prioritization and workflow management, which I think is interesting.
00:47:02.000 --> 00:47:06.880
And honestly, I I was going to answer the law firms because I was thinking about the pre-litigation opportunities.
00:47:07.039 --> 00:47:12.079
They're the least efficient, the most manual, and that's probably where there's a lot of opportunity.
00:47:12.239 --> 00:47:16.239
But after hearing Michael, I think I have to change my answer on that one because I think you're right.
00:47:16.320 --> 00:47:22.239
I think it is the death buyers that have the largest opportunity because they have the most amount of control over the process.
00:47:22.559 --> 00:47:23.199
They do.
00:47:23.920 --> 00:47:28.320
Well, gentlemen, this has been a fantastic discussion, as always.
00:47:28.480 --> 00:47:33.760
I appreciate you coming on, sharing your insights, and participating in this conversation.
00:47:33.920 --> 00:47:37.199
And as always, I learn a lot every time we talk with you, Mike.
00:47:38.320 --> 00:47:39.199
You guys as well.
00:47:39.280 --> 00:47:40.079
It's always fun.
00:47:40.239 --> 00:47:40.960
Thanks, Adam.
00:47:41.360 --> 00:47:42.079
Thanks, Michael.
00:47:42.320 --> 00:47:43.920
And thank you everybody for watching.
00:47:44.000 --> 00:47:45.440
We appreciate your time and attention.
00:47:45.599 --> 00:47:48.320
If you have questions, leave them in the comments below on LinkedIn and YouTube.
00:47:48.400 --> 00:47:54.000
We'll be responding to those, and we look forward to continuing the discussion on applying AI.
00:47:54.159 --> 00:47:56.079
Thank you, everybody, and we'll see you again soon.
