Disclaimer: This interview is for informational purposes only and should not be relied upon as a basis for investment decisions. In Practise is an independent publisher and all opinions expressed by guests are solely their own opinions and do not reflect the opinion of In Practise.

{audio:0:00:21} Can you share an introduction to your role and responsibilities at Carvana?

{audio:0:00:25} I was with Carvana in two roles, starting as a market operations manager for Dallas Fort Worth in Texas, from where I was promoted to operations leader to oversee our central region which included the states of Texas, Oklahoma and the Kansas City expansion. In operations, I was responsible for logistics, inspection centers and last mile deliveries as far as fulfilling customer orders. For local marketing initiatives, I partnered with headquarters in a wide spectrum including market performance and P&L.

{audio:0:01:01} How does Carvana organize their team across 300 markets? I know that financing, pricing and procurement are handled centrally, but what is the market manager responsible for?

{audio:0:01:15} 60% of their work force is centralized out of their Arizona headquarters, which includes finance, procurement, wholesale, CX, dev and engineering. On the market ops side, we focus on market expansion throughout the States, which includes several roles. Customer advocates do the last mile delivery, then there are team leaders and market operations managers from individual markets who have regional operations managers overseeing multiple cities. Across the US, you have cross org collaboration with inspection centers that are strategically placed based on Carvana's vertically integrated hub and spoke logistics network. They own their entire logistics fleet and operation. Inspection centers have a logistics manager and pre-delivery reconditioning manager, doing multiple shifts as they work around the clock. We have to do a lot of heavy cross-collaboration with those people on the market ops side.

{audio:0:02:29} Can we walk through the process of Carvana launching a new market?

{audio:0:02:37} They had the good fortune, for several years of their startup growth, to piggy back off the DriveTime infrastructure, which was a huge operation belonging to Ernie senior. They had extra space in existing inspection centers and carved a little out for Carvana to set up shop almost overnight. They simply put in some fold up tables, parked a hauler to deliver cars and were ready within several days, which was a huge competitive advantage for Carvana early on. TAM to a market expansion was significantly reduced but as they became bigger over the years, they started tapping out of available DriveTime network space. 

{audio:0:03:35} They then had their own real estate department to get forward thinking on space and how to approach a launch. The time lines got extended and they had to shift gears as they still wanted to show market growth. They have always been touted as being in over 300 markets, so they started doing satellite markets in their core markets. Dallas Fort Worth had a vending machine and was a core hub, but they increased delivery from 50 to 150 miles, which reached 20 extra cities in the MSA. That was their tool as far as expansion goes and they would tout a lot of that fast-paced expansion through PR. Much of that heavy footwork was done early on in Carvana as far as getting licenses in multiple states.

{audio:0:04:32} Can you share more insight into how pivotal the DriveTime relationship was in the early days? You used their lots but what else did you leverage from them?

{audio:0:04:46} It was mainly physical space and resources. There was some initial capital backing from Ernie senior but there was still clear accounting boundaries using DriveTime's facilities. It was two independent companies and we simply used their physical space to help us grow.

{audio:0:05:12} Were you using any of their recon facilities or did you have your own?

{audio:0:05:17} 80% of the time, we used DriveTime facilities. Carvana didn't need a lot of space; they were simply parking cars before delivery.

{audio:0:05:31} As Carvana entered new markets, how did they target certain customer types?

{audio:0:05:37} Over time the data revealed several ideal customers as the target audience. They were in their 30s to mid-40s, slightly more progressive and lived in tech savvy cities. There were still soccer moms looking to buy cars, but as our marketing prowess kicked in and we had economies of scale, you started to see a wide spectrum of customers. We even saw 70 to 80-year-olds buying cars online because the process was so easy.

{audio:0:06:18} What were the average FICO scores or split between sub-prime and prime?

{audio:0:06:25} It leaned heavy towards lower FICO scores, with 600 being the sweet spot, due to Carvana being friendly on financing and offering sub-prime through their in-house financing. A lot of demand and fulfillment came from that profile.

{audio:0:06:47} Did you specifically market to those customers about better financing rates?

{audio:0:06:55} We tried to couch it more broad spectrum and highlighted in-house financing and flexible terms, but we didn't pinpoint into sub-prime, but they do gravitate when it's flexible.

{audio:0:07:59} What was the profit difference between trade-ins versus buying at auction?

{audio:0:08:11} Buying from public was and continues to be a great source of inventory for Carvana and has significantly helped to drive up their GPU. When you buy from public, you have more control over what type of vehicles you want and can offer. Typically, you can also get a better price on a trade-in because they are all in-house with the same transaction. You also have the opportunity to put physical eyes on the vehicle, so our customer advocates would look at the vehicle in person, do a test drive and really assess it, in order to catch issues or repairs needed. We saw that those vehicles moved quicker through the pre-delivery inspection reconditioning because they had fewer cosmetic things to be fixed and less reconditioning costs that go into it.

{audio:0:09:16} 15% of vehicles were acquired from customers in 2018; did that change depending on the maturity of a market?

{audio:0:09:29} Carvana were very crafty with how they positioned their buy from public program. They made it super easy and, coupled with marketing prowess, we saw that number increase. The longer Carvana are in the market, the more people are comfortable with their model. The mix steadily climbed from 15% to the mid-20s.

{audio:0:10:00} Is there a limit to the proportion of trade-ins could reach for Carvana?

{audio:0:10:04} It's limited by inventory, in general, and we obviously went through a pinch. For a long time, they were specific about no cars older than six years being allowed on their website. Then they extended that out to 10 years but the cars would have to be under 100,000 miles and meet other condition requirements, which obviously helped drive up that percentage. They could comfortably reach 50% but not everyone has a car they need to trade-in.

{audio:0:10:45} Exactly, so that seems to be the crest on it.

{audio:0:10:49} Was there a difference in inventory quality of trade-ins from customers who also buy a vehicle from Carvana?

{audio:0:11:02} It's the same approach, but one would have the delivery or customer advocate who picks up the car when we're just buying it from them, cuts them a check and moves on, versus the trade-in process which has different degrees of paperwork. How we approached the actual vehicle and the intake process was the same for both, and so was the pricing.

{audio:0:11:23} What was the mix between those that who traded in and bought a vehicle versus those who only sold to Carvana?

{audio:0:11:31} Out of that 15% to 20%, roughly 7% to 9% would be those outright selling a vehicle.

{audio:0:11:42} The majority are trade-ins?

{audio:0:11:46} If I were to trade-in my vehicle on Carvana, how would it work? Would I simply enter my VIN number and automatically receive a quote from Carvana, or is it contingent on a customer advocate viewing the vehicle and looking for damages? Does it determine the loan I could get on my new vehicle?

{audio:0:12:07} It does and you can do it in two ways. You could start top of funnel and simply tell me how much you would pay me for my car, and they have a clear call to action button with a very streamlined process. When trading in a vehicle, you would start looking for a new one, and once you found one, you would start the purchase process through the cart system, where they ask you if you have a trade-in vehicle, which takes you to that same trade-in process. Then you start to understand exactly what your instant value is. They will ask several quick questions to assess the condition of it, but many people weren't honest, which is part of checking it in person. Once we got to the driveway to collect the car, we would look at your responses and see if they matched the physical condition. If not, negotiations happen on the driveway. We would take $1,000 off because it has hail damage, and in that case, you would make a quick phone call and the contracts would get adjusted, which they sign on the iPad.

{audio:0:13:11} What was the biggest challenge in increasing trade-ins?

{audio:0:13:15} The biggest challenge was getting Carvana to move from that five to six year, because there is a negative connotation which comes with used cars at times. That is why they wanted to stay with newish, well-maintained, low mileage, but that only gets you so far. In order to grow you have to expand inventory reach. The challenge was pinning that process down and being comfortable taking in 10-year-old cars with 100,000 miles, and training the team to understand what a good investment was versus what they should shy away from.

{audio:0:13:53} Was there a target percent of trade-ins you had to retail versus wholesale, or to improve customer experience you took in as many trade-ins as possible?

{audio:0:14:02} We still bought any car at the end of the day; it didn't have to be less than 10 years and cars that were 15 years old or had 200,000 miles could still be sold to Carvana, who would wholesale those. We had that process on the back end but less than 2% went to wholesale.

{audio:0:14:32} 2% of trade-ins would go to wholesale?

{audio:0:14:35} Do most of these cars go straight back into the inventory for retail?

{audio:0:14:41} What percentage of trade-ins need to be reconditioned?

{audio:0:14:45} Every car goes through the reconditioning process, which is standard at Carvana. It is similar to a big factory line and goes from step 1 through 10, to ensure it meets standards and there are no safety issues we were unaware of. It then gets green tagged to go to the website for photo booth, which is the final step, then it gets staged for a customer purchase. If some issues are discovered as far as repairs needed, they would get staged in a different part of the inspection center until the parts came in and it was repaired.

{audio:0:15:23} So everything come back to the hub and go through the factory line?

{audio:0:15:29} Yes, they all get inspected.

{audio:0:15:31} How do you organize the logistics for collection of trade-ins?

{audio:0:15:34} We used proprietary software which was built around our logistics network. Any time a customer order was received via the website, the algorithm assigns it to their zip code and we would have a daily calendar on the market ops side which is automatically filled. The cool thing about Carvana is that consumers can buy any car within the US and pay an extra delivery fee if it's coming from far away.

{audio:0:16:25} Would that be pre-populated for you and you simply organize your people?

{audio:0:16:32} Yes, but the biggest challenge was the scaling part. Demand was more than supply, so you would have to get creative on how you execute on all these deliveries and pickups when you only have four haulers and eight advocates. We would often try to piggy back, so if someone had to drop off a delivery in Dallas but also had a buy from public in Fort Worth, they would do it all in one trip. There was a lot of schedule massaging and we would also drive cars to customers and have the advocate Uber back because no hauler was available. That was not the ideal experience but it was the simple reality because there was not enough capacity.

{audio:0:17:23} Did they use Uber to get home because there weren't enough haulers?

{audio:0:17:28} What type of truck and personnel are used for that journey?

{audio:0:17:43} None of our advocates needed a CDL license as we use single flatbed Isuzu trucks, dressed up in Carvana skirting. They have since rolled out two-car haulers which helped with capacity and cost because you could take two cars 150 miles and try to couple the drop-offs.

{audio:0:18:10} Is there a reason why you couldn't have three or four cars?

{audio:0:18:14} At that point, you start getting into the CDL license and some of the requirements for driving big haulers, which would be a big ask to hire customer advocates that way.

{audio:0:18:25} Are the people delivering the car normal unlicensed employees?

{audio:0:18:30} Yes, they have a normal driver’s license.

{audio:0:18:33} How much does it cost to run those? Is it two people or one on the truck?

{audio:0:18:39} It would typically be one person but sometimes an advocate would tag along if we wanted to drop them off to pick up a car someone was selling us and they could drive it back. Rough numbers when it comes to market performance and managing cost is, any time we would send out a hauler to deliver to a customer, it would cost $500 once you figure in gas, hourly wages for the advocate and the wear and tear on the hauler vehicle.

{audio:0:19:16} Yes, and often times if the schedule got so maxed out, we would ask customers to collect their car and give them a $100 gift card to have dinner on us which, at the end of the day, saves several hundred dollars because we didn't have to send somebody out.

{audio:0:19:31} What was the average delivery distance for that $500 cost?

{audio:0:19:36} Our average distance was 55 to 60 miles. In some markets which opened satellites, you would go up to 150 miles one way, so it quickly added up and those costs reached $800.

{audio:0:19:53} What are you including in those costs for labor? $500 seems a high number?

{audio:0:20:05} That was one thing they changed, the pay and their hiring profile. While I was there and before going public, many advocates were making $20 an hour, which adds up on a three or four hour round trip. You also have gas, which was typically close to $50, because they don't give good gas mileage. Once you figure in the wear and tear as far as putting on mileage as well, you get close to $400.

{audio:0:20:36} That is expensive for last mile. Was it $2 per mile excluding wear and tear?

{audio:0:20:55} What was the biggest challenge in picking up and delivering vehicles?

{audio:0:21:05} It was around keeping up with demand, and every market was in the same boat. At times, we would have to figure out where to park these cars. We have so many orders but only X amount of space, so we would park cars at a mall two miles away and rent out space. If you add in all these complexities of trying to get a car delivered last mile, it was challenging. Over time we got smarter, found bigger sites and added more haulers which opened up delivery slots. There was always a last-minute creative audible to figure out due to demand.

{audio:0:21:50} What was the cost of outsourcing last mile, if you had to?

{audio:0:21:54} We would typically see it double, and you lose control of the customer experience which is hard to quantify. Third party logistics companies would be 1.5 or 2X of $400, so you would see that jump significantly. Once we started building vending machines, the idea was that people would come to them, which would ultimately reduce last mile costs and save time. We could have less staff but that wasn't the case and 60% of people wanted cars delivered. We realized the vending machine was cool, but people could watch it on You Tube and see and understand the experience. Amazon ruined it for everybody so they want the car overnight, which became a challenge because those vending machines are expensive.

{audio:0:22:50} Was that the original view, because they launched in 2015 and quickly ramped up to 15 vending machines by 2018, and it has grown to 30 this year. But they only opened three last year, so is Carvana saying this is more of marketing expense than people wanting to pick things up from a vending machine?

{audio:0:23:17} It is great marketing because it's typically placed on a busy freeway so it's a huge walking billboard. That was part of the intent behind it, but operational gains could be had from the philosophy of it, not to mention the longer-term play of building in a repair service, because you want to have frequent touch points with these customers. A new car is every four to six years so you have to figure out how to monetize a particular customer before seeing them again. That has never ever come to fruition so there are many strategic things the vending machine was built for, but it was more marketing than anything else.

{audio:0:23:56} So was service potentially in the mind of Ernie Garcia and Carvana?

{audio:0:24:02} It was; even up to the day I left it was always talked about. It was on the road map at some point if we wanted service in these dealerships, because in the traditional dealership model, that is where all their money is made. Today, it is unclear where that sits. There was even talk of being able to customize your vehicle, so you could order a vehicle and it would be presented with new wheels or tires. That could all happen at the inspection center because we have the space and in-house parts storage, so a lot can be monetized on that side.

{audio:0:24:42} Was the service aimed to be done at the vending machine or at the IRC?

{audio:0:24:47} This would be done at the vending machine if a typical consumer needed to have an oil change or have their tires looked at. They could go to their local Carvana where they know the friendly faces, and you can build that community and keep in touch with customers because you want to get their next vehicle when they're ready to trade in and sell theirs.

{audio:0:25:06} Is there space at the vending machine for that?

{audio:0:25:09} In their current prototypes there is not; vending machines have gotten smaller. For a long time, vending machines didn't even have detailing areas set up, so we would have to buy a pop-up tent and put it in the back, so there are some challenges with how it was built.

{audio:0:25:27} How many cars would you house in the vending machine?

{audio:0:25:34} The early prototype 1.0 was a three or four story vending machine, and each story holds four cars, so you could go up to 20, but then we started getting into nine and 10 story, so you could handle 36 to 40. The sweet spot was a seven or eight story vending machine, and you would have four bays, where essentially it was vended to a bay, which was the showroom, where customers could go into with the delivery advocate, walk the car, then drive it out for the full experience. You could still park cars in there if you needed to rotate, but that was the in-house capacity. The only cars that go in the vending machine are those that have been sold, so we ensured for that day or two days out, they were ready for people to collect.

{audio:0:26:26} When a customer is buying the vehicle, they get a chance to pick it up from the vending machine or get it delivered. If they choose to pick it up, do you have to send that vehicle from the hub to the vending machine for pick up?

{audio:0:26:40} It comes from the logistics center to the vending machine for in-person pick up.

{audio:0:26:47} For the last mile piece from the hub to the customer, would you always have to go back to the hub to pick up the vehicle in the one car truck haulers, then deliver them to customers, or was everything based around the hub in your market?

{audio:0:27:09} Each market would have a main hub or vending machine. In Texas, the major cities are Dallas, Austin, San Antonio and Houston. All four of those had vending machines from which the core team would operate from, and they would also do deliveries from there. All purchased cars go to those city hubs which are vending machines, and either sit on the lot or inside the tower if they are a pick up. If they're sitting on the lot, it's usually for deliveries. Everything is funneled through that one source within an MSA and delivery zone.

{audio:0:27:46} How many cars would you have on the lot for the vending machine?

{audio:0:27:50} In Houston, we sustained a lot of volume and could have 50 to 60 sitting on the lot, so if a vending machine holds 36, many cars sit on the parking lot, but that constantly churns.

{audio:0:28:12} Does that $500 cost include the leg from the hub to the vending machine?

{audio:0:28:24} No, that would be the very last mile from the hub to the actual customer.

{audio:0:28:30} Is that from the vending machine to the customer?

{audio:0:28:32} Yes, when I say hub, I'm referring to the vending machine as the market hub.

{audio:0:28:37} The Carvana group talk about vending machines on the IRCs, are they effectively talking about the vending machine lot?

{audio:0:28:49} Yes, effectively vending machine lot.

{audio:0:28:51} How much does that cost from the IRC logistics center to the vending hub?

{audio:0:29:01} That is typically spread out over nine vehicles in nine-car haulers, so you can maximize your space when sending it that way and those costs are actually pretty streamlined. A lot of that cost management would fall onto logistics managers but yes, it would be less.

{audio:0:29:23} Do they move nine haulers from logistics centers to hubs once purchased?

{audio:0:29:35} Yes, sometimes we had three nine-car haulers coming in throughout the day.

{audio:0:29:40} Do they now fulfill straight from logistics centers to customers?

{audio:0:29:48} Each market has a hub or market office, even if it's not a vending machine. It is usually some sort of warehouse space they have leased out or part of a DriveTime dealership. There is typically a last mile physical office that the cars would go to, and usually the team there does the last detail, get all the paperwork in order and prints out the packet.

Carvana Operations: Reconditioning, Trade-ins, & Last-Mile Delivery

How CVNA moves cars from customer trade-in, to IRC’s, to hubs and the challenges 

Estimated ~20% of cars are bought from customers; 60% are trade-ins, 40% open buys 

Repair costs for trade-ins are estimated ~$800 vs $1,500 for cars from auction 

Insight into working with Ernie and CVNA senior management team

Carvana Operations: Reconditioning, Trade-ins, & Last-Mile Delivery

Former Operations Manager at Carvana

inpractise.com/articles/carvana-operations-reconditioning-trade-ins-and-last-mile-delivery

Why is this interview interesting?

Executive Bio

Trent Capps

Former Operations Manager at Carvana

Interview Transcript

Related Content

Carvana Finance GPU Analysis

Current Chief Risk Officer at US Auto Sales & Former Director at Santander Consumer

Investor Dialogue: Carvana Liquidity & Finance GPU

Vroom, Carvana, & Optimising Vertical Integration

Cofounder at Vroom

Investor Dialogue: Carvana vs CarMax

Audio

Carvana Operations: Reconditioning, Trade-ins, & Last-Mile Delivery(April 30, 2022)

PARTNER

Speak to Executive