Why leveraging routing software can improve repossession process & more


As president of Intellaegis, the software company that built masterQueue, and a founding partner in RepoRoute, I’d like to share some thoughts on the importance of field management software and where we see this new technology taking the industry.

We see this moment as one of the most critical times in the history of auto finance. Delinquency is spiking, repossession companies are breaking company volume records, and there are more dollars delinquent in auto loans over 30 and 90 days than there has ever been. According to the Federal Reserve Bank of New York, the industry just passed the previous high from the Great Recession of 2007-2008. Here’s what Bill Ploog, former head of collections in auto finance at Ally Financial sees as the entire industry is seeing a spike in delinquent dollars past due.

Before I talk more about the present, I’d like to take you a journey through the past, helping you understand why technology can make a huge difference for the repossession industry. Everyone is blogging and tweeting about FinTech, financial services technology. There’s a reason, and this article should shed some light on why that is in the auto finance industry.

My first job in auto finance was as a field rep for Chrysler Credit in Sacramento, Calif., in 1982. My main task was to collect delinquent customer payments in the field. If the customer couldn't pay, I repossessed their car. Pretty straight forward. No tow truck, just me, a Curtis Key Cutter, the factory key code from the invoice and the adrenaline of a 22-year-old. I worked almost exclusively in the field, and I built my own routes where I’d travel each day. It was a manual process, it took time and I made mistakes.

John Lewis, Intellaegis

After five years, I left Chrysler, and my wife and I started a skip-tracing company in 1988. We did this because the auto finance industry was growing, and we saw an opportunity. Lenders were taking more chances on making loans, which meant more repossessions, and that meant more skip accounts. Kind of similar to how it is now.

Before we understand why “good” field management software is a must have component of any repossession agency, we should first understand the assignment, routing and field management process, and how it’s evolved through the years.

Deterioration of the repossession process

One of the main reasons we we’re successful in 1988 in building one of the first skip companies was we started to see the quality of the repo assignment decrease from where it was when I’d started just six years earlier. It wasn’t that there were that many more skips, but collectors were creating skips because they weren’t putting in the fundamental work to reach the customer at an earlier stage of their delinquency. Bottom line, they weren’t verifying the address prior to placing the account for repossession as I was trained to do at Chrysler. The captive lenders also eliminated field reps, and while “door knock” companies started to come into the market soon thereafter, it wasn’t quite the same as having your own employee in the field.

We also noticed collectors were wasting repossessors’ time running bad addresses, and the customers were digging deeper holes for themselves as they went from three to four payments past due, which was creating more “skips.” The customer knew by then their car was out for repo, so they tried to hide it, and many customers avoided all contact. In a sense, we turned them into a skip but not being more efficient in the collection process at an earlier stage.

Once public records data became more accessible for collectors and skip-tracers in the late 1990s, the skip problem magnified with more unverified, bad addresses assigned for repo, and repo success percentages continued to decrease significantly.  Collectors and even skip-tracers got lazy; they just saw a new address and “shot-gunned” it to a repo company. There were little to no consequences they thought, as it was a contingent assignment, so it didn’t “cost” them anything to have the agent check the address. 

The problem was simple, and it still is simple. No one was (or is) making calls and verifying addresses before placing an assignment for repo.  The “collectors” and “skip-tracers” just look for an address and they place it for repo without making phone calls to verify the information. Problem with this is it is not “free,” not even close. As Bill Ploog says, “Collections is a race against time.”

If you use this mentality, once the account is placed for repossession, every day is critical. Assigning an account for repo to an unverified address wastes valuable time, and while the “price” of a contingent repo is zero, if the car doesn’t get recovered, the “cost” of an inefficient assignment can be significant, especially when the loan charges off as a total loss.

So, when more and more bad addresses started getting assigned to the field, it opened the door for not only a skip-tracing industry to eventually be formed, but nearly 30 years later, its opened the door for field management software to be developed to help repo agencies, and specifically to allow repossessors to work more efficiently in the field.

More ramifications

In recent years, we’ve seen the percentage of successful repossessions decrease significantly. Today, it’s at a point where repo agencies have hundreds or thousands of “open” addresses to run, addresses the client expects them to check routinely. These are addresses where statistically, the chances of the car showing is slim to none. This could have been determined if the people placing the assignment had done the leg work of making calls to verify the address prior to placing it for repo.

Sounds simple right, just make calls, verify addresses, and place for repo. If you’re not doing this, you can sign up for a demo at www.masterQueue.com and we can show you how it’s done, but that's a different topic, so let me continue.

In 1990, we created Crown Auto Recovery, and we quickly became one of the larger repo agencies in Los Angeles.  We did this not because we were having trouble finding the skips, but because we were having trouble getting them repossessed. The repo agencies were either so busy running bad addresses for others that my addresses weren’t getting run, or they were fabricating updates and saying it didn’t show when I knew it did, and they did that because they either we’re too busy and couldn’t handle all the work, or they had a repossessor in the field who fabricated an update back to the office, which then went back to the client. I know this because I was going out and repossessing my own locates when the agent said my car “didn’t show” and in most cases it was there, and I knew it would be there because I verified it.

By opening a repo agency, routing changed for me, as I wasn’t just routing myself as I did at Chrysler. Now I was routing a dozen guys every day and night. The order of each route and the decision to run, or not run each address that night was based on a variety of factors:

—Capacity: How many repossesors did I have versus how many addresses needed to be checked. Rarely did I have less work than I had manpower, and usually I had two to three or more times the work I could handle. It was my job to insure our company put out the best routes to insure my guys would pick up cars, and not run 15 addresses and get skunked. I quickly learned to remove the addresses that needed to be verified or skip-traced, so they wouldn't get mixed with verified addresses that should be checked.

—My intuition.

—The skill of a skip-tracer in my office.

—The skill of a repossessor versus the type of car he’s after and what is required to take it. Remember we didn’t use tow trucks back then.

—The number of hours he worked the night before. How many cars did he repossess? Did he get a full night’s sleep? Where is he on his weekly commission versus how much time is left in the pay period? Has he repossessed enough cars to pay his rent, or has he repossessed a bunch of cars and he may take the night off or worse, he may mail in his updates and say he checked an address, but really didn’t?

—The quality of the assignment from the specific lender.

—The repossession percentage of the lender placing the assignment.

—The skill of a collector at the lenders office, to how many times we’d already been by the house, to the type of vehicle, did we have a key code, a color, etc.

—The chances of picking up a car in a general sense from this neighborhood. We picked up a higher percentage of cars in the San Fernando Valley as there are a lot of apartments with open parking versus in Arcadia where there are a lot of single family homes with garages.

For sure, intuition played the biggest part as technology was almost non-existent back then, so it was a lot of labor, knowledge, information, time and intuition going through my head to build routes. 

Each assignment also brought a different thought to my head, from “That’s a Ford deal from Alice at XYZ Bank. That’s guaranteed money.” Or, “second placement, 7 Series BMW that’s 90 days past due from XYZ Finance Company at a house in Arcadia. That’s probably his mother’s house with a garage. That’ll never show.”

Next developments

Fast forward a few more years to 1993 and we’d now sold Crown. We moved to Sacramento, and we started another repo company — River City Auto Recovery.  We also now had cell phones, computers and repossession software, but we were still receiving assignments and sending updates to clients through fax. We were also still printing assignments and manually routing them onto repossessors’ clipboards, paper assignments, based on where they lived, or what area we wanted to send them to that night.

We we’re doing it manually through all the process management we’d built to try and run an efficient repo company, and all the customers personally identifiable information was sitting on that clipboard every night. 

That’s how I approached routing when I worked at Chrysler and when we owned our own repo companies. Routing was a manual process, and for many it still is, either manually in the office or manually by a repossessor, and from what I hear and see, it’s even manual for these people who use routing software’s other than RepoRoute, and even for some who use RepoRoute, but at least we give them an easy way to be “old school” and create manual routes.

When Michael Eusebio, president at Digital Dog Auto Recovery, approached me three years ago to have our company partner with them on building routing software, I knew there had to be a better way to manage this process more efficiently than everyone was doing. I quickly realized the repossession industry hadn’t changed much from when I was in it, except the job of routing usually got outsourced to the actual repossessors. It was rare to find a company creating routes for drivers as I’d done.

Some repossessors have great intuition and are proficient and some are excellent at routing manually. It’s an art that comes with a sense of pride and a badge of honor that's worn based on results, just like skip-tracing. The problem is, how do you teach intuition to a new employee who has never repossessed a car, or a guy who just doesn't have that skill, but maybe he’s a great employee and a great repossessor when he see’s the car, but putting a route together is not his strongest asset? Also, the real question is can a machine do a better job at routing than the repossessor, or can there be a way to combine the human skill with machine learning in a software platform to build an efficient way to route repossessors.

We believe this can be done, and after three years we’ve begun to prove this is the case, and it’s called RepoRoute.

I always knew there must be a better way to route than to spend hours manually doing it, but technology hadn’t advanced enough for me to come up with any ideas back then, and by 1999, we’d sold all of our companies; Skipbusters, ARS and River City Auto Recovery, and I was out of the repo industry, for a while anyway.

When Michael and James McNeil at Digital Dog were encountering the same manual route generation problems mentioned above, and given the advancements in technology, their rapid growth, and the need for a solution, the opportunity to build RepoRoute interested me and my partners so we jumped in. We kicked around some basic ideas, hired a developer and got started. We could have easily just put out software that met some basic needs, as many software companies do, but that’s not how we have learned to create a lasting product. To build technology that will really solve issues, software that’s “sticky,” you have to address all the issues, and in repossession routing that means much more than just putting pretty pins on a map.

Three years later, we’re seeing companies who use www.RepoRoute.com help us evolve the product into a game changing platform for their business.

If you are in the repo business and you aren’t already breaking records for more repossessions than you have ever done, then you are doing something wrong. The work is there.  This also means those of us in the auto finance collection space better hold on, as we’ve just gotten busier by a great deal, and the trend is showing this will increase at a record clip. How do you solve the problems this type of high-risk growth causes? In my opinion and based on my experience; two words: better technology. If you are in the repossession industry, two more words: RepoRoute.

I have a tremendous amount of respect for repossessors, maybe more than for any other profession.  I know first-hand how tough and how dangerous a job it can be. The reason I hesitated for so many years to have anyone but me do the routing was I always felt routing was the single most important job in running a successful repo company.

6 lessons learned

So building successful routes in 2017 starts with a few key tips:

—Less is More. Having a system to identify what addresses you shouldn't waste time working is critical. This is the key to valuable routing software, and to do that, the software has to help you identify this easily, and quickly, and when possible, automatically. It’s important to run less addresses to pick up more cars and that’s exactly what we’ve been able to do. This is a core value proposition we’ve built into RepoRoute. We do this better than any platform I’ve seen, and our users who embrace the platform attest to this.

—I always thought the repossessor shouldn’t have to waste their valuable time creating routes of where to go. First off, it can take an hour or more, and once you get a car or get a call with a rush deal, or something else happens, you have to either start over or spend time recreating a modified or new route. I always felt our repossessors should be home with their families, or getting an extra hour of sleep before going to work, not sitting at a table trying to figure out which deals that have never been hit should get routed, or which deals that were ran once before may happen to show that night. Push a button, create a route, that’s our goal. In some cases we’re there, in some cases we are still perfecting the algorithms as it takes time and volume and historical results to do this. We’ve done the same at masterQueue and now we run millions of pieces of data every year and that volume creates predictive analytics, which helps automate decisions of where to go, or not.

—Evaluate your clients, and stop working for the non-profitable ones. At RepoRoute, we wanted to insure that the deals being routed were the ones our clients wanted to be ran, and if we could help build technology to do that, we felt this would be a sustainable product others would get value from.  If you have a client giving you work that’s resulting in a 70 percent recovery percentage and they gave you 10 deals, and the same day you received 10 deals from a client you’re struggling to make a profit from because the quality of their assignments is so low, do you want software that helps you decide what deals to run that night? Could we build software that can tell users that level of detail? Heck Yeah, and that’s what we have done with RepoRoute. If you are managing your repo company without technology like RepoRoute, you’re operating with one hand tied behind your back. Call Michael or James and let them explain how RepoRoute has made DigitalDog and DayBreak Metro, the two repo companies they own and operate, more efficient, and how margins have increased and their clients and employees are happier than ever.

—Build solid technology and a winning team. The next challenge in building RepoRoute was how could we leverage technology in the same manner we’d done in masterQueue.  We started with quality first, partnering with industry leader Google for our mapping, industry leader Rackspace for Security, compliance and performance, and by hiring an experienced Full-Stack developer we could build a team around. He didn’t know anything about repossession or routing, but we did. We also knew we wanted to incorporate predictive analytics into the platform, creating automation when it was convenient, but at the same time allowing the repossessor the ability to create a route on their own if they wanted to. We also needed to measure which way was more efficient, and over time, as was the case with masterQueue, we were confident that we could build a blend of both to get the best results for the repossessor, the repo agency and the client, and the consumer.

The consumer? Yes, this is also important as now the Consumer Financial Protection Bureau tracks everything, so it’s important to not be making contact with the wrong person, or someone who already has been contacted and should not be contacted again. We’re also seeing preventable, senseless tragedies where technology can improve the background check and training process. In this case a Mom lost her life, and the question is would this person even have been there to repossess her car if proper background checks have been performed, and would he have acted in this manner if proper training been documented to have been performed?

Next we had to build a winning team, so we brought one of our lead trainers and support people from masterQueue to run the team, Polly Schumacher, and we hired a second developer from Apple, whose background was in mapping. Customer support at the highest level is mandatory, and we treat our customers as they would want to be treated, in a friendly and attentive manner.  With the leadership of our president Michael Eusebio, and the product development skills James McNeil brings, a former repossessor who is now an executive in one of the largest and most successful repossession organizations in the country, we are 100% confident we have the foundation of a winning team.

—Integration. So many charge-offs could be saved, wrongful repossessions prevented and manual labor eliminated if the tools available today to gather and utilize data were more efficiently connected as it relates to the repossession assignment process.

Early on, we recognized the need to integrate data so companies didn’t have to do so much copy and pasting, and so when assignments closed or updates were entered, they moved between systems through real-time bi-directional integrations with RDN, iRepo, Prios and other platforms, without the need to work in two systems, or to copy and paste data between systems. This prevents wrongful repossessions and it eliminates a great deal of manual work. We’ve built integration partnerships with the two largest assignment patforms RDN and MBSi (iRepo and RCM) to make the workflow for repossessors more efficient, and we’re definitely seeing great strides now in this direction through these partnerships. We spent additional time building deeper and more thorough interfaces with these platforms than anyone in the industry, and the results are fruitful for our clients as it cuts down the need to work in two or more systems. This also made everyone more efficient.

—Security. Fast forward to 2017, and we’re now seeing clients starting to mandate what software their data can reside in, or what software they want the user to work in. We’re being told by lenders this is due to security and compliance concerns, as they are just realizing how many systems their customer data is residing in, and the fact many of these systems have not been audited and verified as secure is a growing concern. This is also a concern of the CFPB, and after meeting with them last week I can share they are concerned about how data is being shared, and how systems talk to each other, or not, how frequently are consumers being contacted, and how is the information not to contact a consumer, or a related party being addresses by the lenders and their vendors, and is it done in one system, or at least in connected systems?

Building and operating masterQueue for 10 years, I can attest that these audits are becoming more comprehensive than ever. Lenders are now requiring full code reviews by external third parties, extreme levels of encryption of the data at rest and in transit, even non PII data must be treated like personally identifiable information (PII) and internal and external third party penetration testing to insure every means possible is being used and verified to prevent anyone other than those intended to see the consumer PII stored in any system containing the lenders customer data. I’ll write more about this new trend in another blog soon, as I believe it will shift the industry in a similar manner compliance started to do in 2011 when the CFPB was formed.

In the meantime, if you’d like to see a demo of RepoRoute, visit www.reporoute.com and fill out the demo form, or send us an email to sales@reporoute.com or give support and sales a call at (916) 800-1010 and they can set it up.

Remember, in business, sometimes you have to change your direction, and we give you that option in a simple to use, affordable platform with RepoRoute.

Be careful out there.

John Lewis is the president of Intelleagis and can be reached at jlewis@intellaegis.com.