No, no, no. This is fine.
Is it time? Yeah, it is. Bingo. Okay. Good afternoon, everybody. I hope you saw on the schedule that, we've got a happy hour coming up at the end of the day. Stick around with us, please, for that. Next company we have for you is Red Violet, based in Boca Raton. With us today, we've got the Senior Vice President of Finance and Investor Relations, Camilo Ramirez. Neat business. They're actually a client of ours, leader in the identity verification space. So with that, I'll turn it over to Camilo.
Thank you, Steven. Appreciate it. Thank you, everyone, for joining me today. I'll go through a little bit about management history, what we do, competitive landscape, give you a couple of use cases, competitive advantage, and then I'll leave time for Q&A. My preference is to make it more conversational than anything, so if you have any questions, feel free to ask. Like Steven said, all things identity, we say we're applicable to almost every transaction that occurs in the U.S. Management has been together for two, three decades. They started in the identity space, in the late 1990s. They created a company called Accurint, ultimately sold it off to Reed Elsevier LexisNexis for about $750 million. Non-compete's expired, got back into the space, created TLO, and ultimately sold that off to TransUnion.
The main, founding member that was funding it ended up passing away abruptly, so sold it a bit premature. Call it revenue run rate was about $25 million, burning about $1-$2 million a month. Ended up selling it for just under $200 million. In aggregate, call it just under a billion dollars within those two organizations. Again, non-compete's expired. For some of the members, ultimately they got back together, got in the space. Technology changed drastically from the late 1990s, early 2000s. AI and machine learning, large language models were not a thing. Call it in around 2014, team got back together, built this iteration in the cloud as opposed to the legacy two products. They're in large data rooms, so we're able to scale up and scale down during peak productivity hours, use the latest and greatest technology as well as opposed to the competition.
So what do we do? We aggregate disparate databases, call it liens, judgments, credit header data, PII information, call it IP addresses, mobile IDs, anything and everything on every adult individual in the U.S., create that 360-degree profile on that individual, and sell it back out. These data aggregations, they're contractually obligated, long-term contracts, so our cost of revenue is fixed in nature. We say today every additional dollar is nearly 100% contribution line, contribution down the P&L. We like to say we serve five different verticals made up of about 26 different industries. I'll start with one of the easiest ones to understand, collections. Let's say you have a debt collector, they bought a million records from Capital One. They need to understand right party contact information, anyone that's deceased in those records, and have they filed for bankruptcy? 'Cause they cannot contact them.
If not, they get fined, call it around like $10,000 per incident. Secondly, real estate. We service real estate in two manners. One, through our Forewarn product, which is basically a skin on our IDI core platform. Let's say you're a real estate agent, you get a call, "I want to see this $3 million home. I'm only in town till tomorrow. My name's John Smith. I'm going to show up in a Mercedes S-Class." Basically does a reverse phone search on that phone number. You don't see John Smith, you don't see that S-Class. What you do see is this individual, on the extreme example, just got out of prison for sexual assault. You're not going to show that home, right? What we noticed is that there was a lot of crime against realtors.
Realtors show a certain persona, how they present themselves to the community, and there was a lot of crime being committed against those individuals. This is a safety solution for them. We go after the real estate association so they can offer it as a member solution for the real estate agents. Today we're in just over 500 real estate associations. There are about 1,100 associations in the U.S. From a realtor count, today we serve about 325 individual users. From an addressable market, there's just under a million real estate agents. There are more that actually have real estate licenses, but some of those, as you know, don't practice or they're not with the National Association of Realtors. So addressables, call it just about a million.
The other manner that we service the real estate industry is going to be through our IDI core platform. That is essentially we power the prop tech companies. It's going to be more, more like the more of a marketing solution, propensity to sell. Give me a list of everyone that's 65 and over, that lives in a two-story home. And then third industry we service is financial and corporate risk. That's going to be like your large banks, KYC, background screening. A good example there is, let's say I work at Ford Motor Company. I'm in the accounts payable department. I write a $1 million check to, to a vendor. Ford would want to know that my brother-in-law is a director at that, at that company, right? It's not showing fraud, but it's a potential red flag if there's distress or something of that nature.
So being able to make those connections across, family trees per se. Fourth is going to be investigative. That's going to be our law enforcement, private investigators, those kind of interactions. And then lastly, emerging markets. Emerging markets for us are verticals where we don't have a deep penetration. It's going to be the likes of government, insurance. These have been large revenue contributors in the previous iterations, but today we haven't penetrated those. Ultimately, those will most likely be their own vertical as we state externally. I'll go through a couple use cases as well, just some differentiation, set the landscape of the competitor. So LexisNexis, they have about, call it around 400,000 customers. On the TransUnion side, they have about 60,000 customers. Those are our main competitors. Outside of that, it's going to be pretty fragmented.
You have some specific niche players where, whether you're getting real estate data, motor vehicle data, or anything of that nature. And then some use cases. I always like to provide use cases as, as opposed to telling you, "Hey, we're better because X, Y, and Z," right? I can stand up here and speak all day about that. I'll start one with background screening. There was a customer we won early on from the competition, Innovative. Innovative early on, they were doing about five figures a month. Ultimately, they grew well into six figures. They were purchased by Appriss, and Appriss was ultimately purchased by Equifax. As you can imagine, Equifax has all the data in the world, way more data than we have. The Innovative contract was up for renewal at that point in time.
They came to us, said, "Hey, can we get a three-month extension?" We said, "Of course. We're here. If you need anything, we understand you're trying to execute on synergies." They asked for a couple more extensions. Ultimately, they came back and said, "We can't create the same lift on the data." So they signed a multi-year long-term agreement with us. Even though they have the individual data aspects in-house, they just can't reproduce the same lift, the data connections on that data asset. Another great use case that I like to use is on our cloud infrastructure, right? Throughput, that's a big differentiator from the competition. Early on, call it around 2018, 2019, one of our customers was going through a funding round. They received funding through TransUnion.
The CEO of TransUnion called them and said, "Hey, we're not going to close this deal unless you move your usage to us and move IDI data." They gave us a call, told us the situation. We told them we completely understand. We think you're going to have a degradation in throughput and data quality, but we're here if you need us. They moved off for a couple months. Ultimately, they gave us a call, said, "Hey, we're seeing fallout. We don't see the same throughput. We're sending calls through. We're not getting any responses back. Customers are starting to complain. They want to understand what's going on." TransUnion is saying, "You built us a custom API." We reminded them, "We didn't build you a custom API.
You were up and running within 48 hours. Ultimately, that customer came back on board, and they're still one of our largest customers and never left our platform, even with the TransUnion investment. We like to say that we power seven of the top 10 identity players, without naming any customers, the likes of Prove, Jumio, Ekata, ID.me. These are all orchestration platforms. They do not own any of the data, but they have the front-facing solution. They have a customer making a call. They'll call out to us or our competitors and clear that identity verification. They have their unique way of clearing that identity verification. Some of them are going to do document verification, load a picture of your ID, take a selfie, make that match, and call for that PII information behind the scenes. Or it is going to be like mobile authentication.
Let's say you're logging into Bank of America, Wells Fargo, Blue Cross Blue Shield with your iPhone, using facial recognition. What it's doing behind the scenes, it's saying, "Hey, this mobile ID belongs to Camilo. Here's the PII associated with Camilo. Does it match the bank side? Yes, grant access." Mobile authentication. ID.me is big in the government space. Let's say you're applying for government benefits, unemployment, anything of that nature. They need to do an identity verification for the government before they start sending out benefits. With that, we like to say we're a very barbelled approach to the economy, with the ever-evolving economy, right? We see currently, we're starting to see the first innings of degradation on the U.S. consumer at the lower end of the spectrum. Top end of the spectrum, they're soft, they're spending.
When an economy is booming, you have new account openings, new applications, car loans, and so forth. All that requires an identity verification. We'll power that. On the other side of the spectrum, you have individuals defaulting on their credit cards. An early indicator for us is in the repossession market, little canary in the coal mine. Once you miss that second payment, you're most likely going to get repoed. We service the repo market in a very small fashion, but we've seen really good acceleration in that vertical. Individuals are just missing payments. They can't afford those $1,000 COVID payments anymore. We're starting to see that starting to flow through. That will lead us into the collection cycle as well. The current administration gave the green light to start collecting on student loan debt.
It's only a matter of time before we start seeing that flow through. We did a really good job of acquiring new logos on the collection side. When that collection does commence, we should see that lift as well. I'll pause there, see if there's any questions, and if not, I'll keep going.
[audio distortion] how you used to be included in the process with your competitors for data versus today [audio distortion] ?
Yep. Very good question. Some of it's industry and vertical specific, how we go out and win business. We have a sales vertical. Our sales floor is verticalized, so we have an investigative vertical that's going to go after law enforcement specifically or PI customers. We have a financial and corporate risk. They're going after that KYC application, financial institutions, a background screening team. Specifically to Stephen's questions on, let's say, the collection side, that's more of a waterfall effect. How we'll enter the space there is, they get that million records that we spoke about. They'll send it to the tier one provider, their main customer, economies of scale, let's call it TransUnion. Whatever fell out, call it 20% didn't get matched, they'll send it to the second tier provider. Then we'll come in and say, "Hey, let us scrub at the third tier.
It's already been scrubbed twice. We know our data quality is there, and we'll get a high hit rate on that data that's already been scrubbed. Then we'll start moving up the waterfall from that perspective so they can get economies of scale because they want to match, have the highest hit rate on that first instance. Today, we're starting to move to the sole provider as well, especially on these larger relationships where they require an API connection, some type of custom solution. What we're seeing is they're going out to the end user. The end user has a specific scenario. They go to one of these top 10 identity providers, say, "I have this specific issue. Can you guys solution it for us?" That company will come to us, say, "Hey, my end user is asking for this.
Can you guys do it?" And we'll work directly with them to implement that solution and work around it. What we're seeing now is that they're realizing that we're behind the scenes and they're coming directly to us. So one, we're able to get the lift from wholesale to retail rates. We can go to the end user as well and then replicate that scenario across the board. I'll give you a little color on, do you have a question?
Yeah, if you relative to the ID.me, so there's an automated component to that, and then there's a human interaction component to that. [audio distortion]
Yeah. So generally speaking, in the identity verification process, we're in the automated process, right? That's where you get the most margin. You're doing that identity verification on that PII. So again, specific to collection, just because it's easy to understand for everyone, there's an automated and a manual process there. On the automated side, you're trying to garnish wages, right? You need to understand who that individual is, where that individual works, and you have to have a live verification. Someone needs to call the employer and say, "Does Camilo work here?" "Yes." Then you can start garnishing wages and send judgment. We outsource all that manual work. We don't want to bring that in-house. It's just a one-stop shop for the collection space, more of a convenience for them, but we don't focus on it at all.
We stay in the automated space so we can reap the benefits of the automation and ultimately the margin profile that comes with it.
[audio distortion]
Yep.
[audio distortion]
Correct. Yeah. If, if we're powering ID.me, yes, without naming any customers. Maybe I'll give you a little color on data costs to understand the margin profile. All right. Most, all of our contracts are long-term in nature. If you look at our 10-K, you'll see that 40% of our data cost is associated to one vendor. That is not just one data asset. It's multiple data assets. It's a relationship we've had with this provider through the multiple iterations. That contract is up for renewal, call it 12-18 months. We've started renegotiations early on this year, and we're in the late stages, so we're really excited about where that's landing. Ideally, we'll be able to report on that here come once we report earnings. Last time we renewed with them, call it six years ago, we renewed flat. We anticipate the same.
I'll just run down the P&L. I'll give you a couple of margin profiles. At the gross profit line, you can, right now, we're about low 80%. A couple of years ago, we gave some color on what this company looks like at $100 million. At that time, we said at $100 million, gross profit's going to be around that 80%. We've already approached that margin profile at a much lower run rate. On the adjusted EBITDA line, about 40%. Right now, I'll call it we're around low 30s. If you look at, if you run down the P&L, sales and marketing, most of the variability there is going to be headcount. As we expand into a specific vertical, we'll lean into it. We see success there. We'll dig, we'll continue to invest in it.
And then on the G&A side, we like to contract long-term on all our vendor costs, whether it's through general G&A, just so we can control those costs. Again, most of that variability is going to be through our data scientists, essentially. As we, I'd say about four years ago, yeah, about four years ago, we had an investment year on the data science side, building out the products that we're going to market with today. Last year, we had an investment year in personnel again, more, more towards our sales individuals. We brought on Jonathan MacDonald to run our public sector. He was responsible for ramping up the government revenue at TransUnion from zero to where it's at today. We are really excited about what he's brought to the table.
His team in totality, call it around 15-20 individuals, both on the federal side and state and local side. We break out public sector, Fed versus SLED. SLED, that's going to be state, local, education. That's where law enforcement agencies land. Call it today, we have 1,000-2,000 law enforcement agencies. Total scope there is going to be about 15,000 agencies in the U.S. We have seen really good traction on the law enforcement, one through our data quality and then some of the unique interfaces that we have. One unique interface is going to be, you have a witness. There was a hit-and-run. This individual was driving a red F-150, partial license plate A and B.
Basically, you can geofence a location, give me a mile radius, two mile radius, half a mile radius, call it what you want, enter that red F-150, the partial plate, and it'll drop pins for every individual registration that matches that and their location. You can continue to do your investigation. We've seen good success there in our mobile application as well. Let's say there's a law enforcement agency, they're arriving at a home. They understand the individual in that home, but they see three other vehicles. They don't know what type of individuals these are. Are they violent individuals? Through the mobile app, they can quickly look up those license plate numbers, get a good picture of who's driving those vehicles, who's potentially in that home.
Previously, they had to call in back to home base, say, "Hey, can you run these plates for me? Let me know background." It is all at their fingertips. We have seen good success there. On the federal side, when Jonathan came on board, he stated, "Most of my revenue contribution is going to be in the beginning of my revenue contribution is going to commence 2026." We have seen some early wins, which gets us very excited. It is a much longer sales cycle, as you can imagine. Budgets renew every September on the federal side. On the state side, some of those renew around that July timeframe. I am sure the next question is going to be like, "How about the current environment, current administration with DOGE?" I will say we posed that question to Jonathan McDonald, and it is a net positive for us.
Essentially, as we all saw with Elon Musk and Twitter, he went, slashed and burned, cut everything back, realized, "Hey, I cut back too much," started bringing back certain, certain items. That is what we are seeing in the government sector as well. They cut these contracts. They were not up for RFP for another two, three years. We would not have an opportunity until then to submit a bid on these. We are going to start seeing them come back because they cut back too much. Now we have that opportunity to bid on it much sooner. It is no longer, "Let's just renew with incumbent." They want to show efficiencies, price improvement, and so forth. Before we even start testing, there is already the scales tipping to our side, right?
Once we start testing, then we can show the data accuracy and then the economies of scale for them. We can get them better pricing as well. Any questions? I'll pause there.
Can you talk about the barriers to entry in terms of, number one, the data that you get versus the value of what you do with the data? And then additionally, how hard it is to become a provider that can get access to the type of data that you guys get access to?
No, very good question. That coincides with AI, right? The question is now with AI, why can't anyone just do that? You can use ChatGPT, do a name search, and you get some results back. As we all saw, there are hallucinations. It is layered over the internet, right? Bad data in is bad data out. The institutions that are using our platform, they need to have a high confidence data asset. You can't have Bank of America clearing bank accounts for KYC using ChatGPT. We do not see that as a threat. What we do see from the AI perspective is areas of opportunity. I will go back to answer your question. Areas of opportunity are essentially on being able to acquire more data, right? Using AI to read long-form data. There is a RICO complaint. It had five named associates.
We look up those individuals within our system. We see three of those individuals linked together. Now we can have AI go read that long-form data, assimilate and assume those three additional, named associates and tie those individuals together. Even obituary data, being able to understand mother, father, second cousins, father-in-law, and stepdad, and so forth, being able to consume that data and build that family tree per se as well, making those linkages. Then even how you interact with our system, making it more of an interactive as opposed to, "Hey, enter name, date of birth, social." It is tell me everything you know about Camilo Ramirez that lives in South Florida, and you can interact with it.
Even as far as, "Hey, does Camilo Ramirez, is he related to anyone that has a violent background today?" For you to answer that question, you have to go in, search Camilo Ramirez, click on every associate named, look at their criminal records, and make that decision based off that search as opposed to with AI. You ask it, it's doing all those searches automated, and it's going to say, "Yes, he's related to an individual," or, "No, he's not related to it." It will give you that individual's name. Being able to do those investigations in a much more efficient manner. From barriers to entry, I would say to get these up and running, even just from a dollar perspective, call it around $50 million to get the data asset and data contracts up and running.
Your P&L is going to be upside down for a number of years. As you can probably tell, if you go back early into our history, that P&L was not very attractive then. That is essentially the first barrier. Secondly, you have to go to these credit bureaus and convince them, "Hey, give me the whole U.S. population," and give them confidence that you're not going to misuse that data or it's not going to get leaked out. Understanding what data assets to buy. As stated earlier, Equifax has all the data in the world. They have way more data than us. It is not about quantity. It is about quality. Essentially being able to decide, "Hey, these are the data attributes I need," and how you simulate those. Any other questions?
[audio distortion]
No, so Equifax, we do compete with them. They do not have a data aggregation layer. They have the individual data assets. They purchased the Innovative technology. We compete with them in the background screening space specifically. We get data from two of the top three bureaus. You can assume one of the bureaus we do not get data from, based off our competitive history and who we sold to. Yeah, we do consume data from the credit bureaus and we sell it right back to them as well.
You mentioned that your revenue is dropping down at 100%. Can you talk about what EBITDA [audio distortion ].
Yep. To generally speaking, to look at our cash earn rate, right? Easy way to look at it, you take adjusted EBITDA, add back any internally developed software, and you'll get a cash earn rate. That flow through is going to be very nice. Last year, I think on the flow through, we had an investment year down to adjusted EBITDA was about 60% flow through in that to cash conversion, high cash conversion rate. It's going to continue to expand, right? Because we're not going to be able to spend it quick enough, from that perspective without being reckless. It gives us, it takes us back to use of capital, how we're going to use that money, what are we going to invest in going forward. We do have a share repurchase program in place.
To date, we've repurchased about 1.5 million shares, one through our share repurchase program. Second, we bought about 200,000 shares from the Greater Miami Jewish Federation. They received their shares through a donation through one of the initial financial backers. He's pledged all his wealth, so he gave all his Red Violet shares to Miami Jewish Federation. They no longer own any shares. And then lastly, through when there's RSUs vesting for employees, instead of selling the shares for tax purposes into the open market, the company will buy them back and retire them. So in totality, it's about 1.5 million. On the share buyback program, we've repurchased at an average price of $19. Additional uses of capital is going to be M&A. A couple of items that we look for is unique data assets.
Is there a data provider out there that has a unique data asset that we can purchase and then cut out some data costs of further improving our margin profile? Or it is going to be a buy versus build specifically on the end use case it is going to be in. A good example there was the background screening space. We saw that Equifax was entering the space with the Innovative purchase. We decided, "Hey, instead of purchasing, we are going to go out and build our own product." We have a directly competitive product in the background screening space. A couple of initiatives that we have here this year are acquiring more data, one through the AI aspect that we spoke about earlier, being able to read that obituary data, or is there a unique data asset out there that we can go out and purchase?
We get items presented to us all the time and, we'll run through those, see if it makes sense and, and potentially, carve it out and be able to own that asset. Additionally, another initiative is, how can we make our internal operations more efficient as we continue to scale the business? We do not want to have to add additional credentialing members, additional sales support members for every 1,000 customers, right? How do we make those repeatable tasks automated? We are looking at certain tools, building them out either internally or using third party to be able to automate some of these repeatable tasks. As we continue to scale, it is going to improve that margin profile even further.
You said at the beginning that you compete with LexisNexis and TransUnion.
Correct. They're the only two providers out there that have that 360 view aggregation profile. There are other providers that are niche players. You can go get real estate data, driver license information, or you're going to be competing against one of our customers. They're just reselling our data. Yep.
[audio distortion]
Yeah, it's a combination. A couple of the use cases went through it. It's data quality. We have a lot less.
[audio distortion]
It's what you do with that data, how you assimilate it and how you aggregate it. The good, the background screening, right? The background screening example, Equifax has significantly more data than us, and they still couldn't reproduce the same lift on those individuals. They contracted to consume our data, even though they have the individual data points. They just aren't able to simulate it and gather those learnings from that data.
[audio distortion]
Yeah. So the executive team's been together, call it, for two and a half decades. They initially, in the late nineties, they built out the Accurint product. Times were very different. That was a very static connection. You had Derek Dubner, our CEO. Ole Poulsen was one of the founding members as well. He was the data scientist behind the technology. Hank Asher at that point in time was one of the founding members. He was like the identity guru, guru per se that created what, what was coined data fusion, right? Fusing all that data together. Ultimately, that was sold off to, like we said before, LexisNexis. They all got back together at TransUnion. Dan MacLachlan, our CFO, joined the group at that point in time. James Reilly, our President, he drove revenue from zero to the $25 million run rate at TransUnion.
He's now our President here as well. Ole Poulsen also advised us on and worked on the code that built this third iteration. He trained up our CTO, Angus Macnab, on that technology, how to drive it, and then improved on it ultimately to where we're at today. Our CIO, Jeff Dell, he's been through the three iterations as well. First and foremost, his main objective is security at the end of the day. We have multiple third party penetration tests occurring every year. We get audited by the credit bureaus. We're PCI Level 1, ISO 27001 certified, and so forth. Security is at the forefront, every day.
You said 40% of your data [audio distortion]
Data cost.
[audio distortion]
Is related to one vendor for multiple data assets.
[audio distortion]
They don't break out that information at that granular level. They have a lot of that data in-house. I would imagine no. The answer to that would be no, because that data is in-house for them. We'd be purchasing data from the likes of LexisNexis and TransUnion and so forth.
[audio distortion]
They could underprice us. We haven't seen that historically. Every year, TransUnion, come January, they hike up prices on their customers. We get nice inbound calls from those price hikes. On the LexisNexis side, I think their egos are too big to compete on price. They've been at it for a number of decades now. They're an 800 lb gorilla. They have a big foothold in the government space, public sector space specifically. In theory they could, but again, we fall back on data quality and then throughput on the technology. We're in the AWS cloud. As opposed to these platforms, they're built in data rooms.
What, what would happen to say, or what could you do alternatively if your main data provider wanted to double the prices?
Good question.
[audio distortion]
Yeah, yeah. We're multi-sourced on every data point. Let's say our main provider said, "Hey, we're going to escalate you 50%." We wouldn't accept it. We'll bring up our tier two provider, bring it into tier one, and then we'd return all the data points. All the learnings that we acquired, proprietary learnings, that stays with us because that wasn't their data. We had all the learnings already, so we can keep that. You can't say it never happened, but it's highly unlikely because it's all margin for them. They have this huge data asset just sitting on the shelf. It's not like they can sell the whole U.S. population to the next company, right?
There's not a very, there's not a great deal of companies out there that can consume that data and give them confidence that's going to be safe and secure. Got two minutes left.
For the emerging markets, can you rank which ones have the lowest angle?
When we speak about emerging markets, we're speaking specifically about the verticals that we serve where we have a small footprint today. 100% of our revenue is U.S. based. If we're going to go outside of the U.S., we will partner with someone to service that customer. It's usually a customer that wants to interact with us and they just want a one-stop shop. We'll partner with someone to service any of the international needs. Any other questions? If not, I'll give you guys a minute back. Thank you, everyone. Appreciate it.