"The Score Lies!" Why Credit Scores are Failing First Time Buyers (& What The Top 1% of Dealers Are Doing)

Hey everybody, welcome to another episode of the Card dealership guy industry spotlight.

I'm your host, Sam Dark.

Every dealer's got deals that should close, but don't.

And the culprit isn't the car or the salesperson.

Often it's the friction in the financing lane, slow approvals, funding delays, deals stalling

out while the customer cools off.

Today, we're fixing it.

Faster decisions, thin file approvals, and funding that doesn't stop when the bank closes.

Today, I'm joined by Ben Atkinson, VP of Auto Lending at Upstart and Evan Driscoll,

GSM at Audi Jacksonville.

Props to Upstart for supporting today's content.

Now, let's get into it.

Evan, welcome to the Card dealership guy industry spotlight.

Thanks for having me, man. Appreciate it.

So Evan, tell us a little bit about your dealership.

Where are you located?

What franchise?

How many vehicles do you sell a month?

What are some of the biggest challenges that you see in 2026 year to date?

Yeah, so my name is Evan Driscoll.

I am the general sales manager here at Audi Jacksonville in Jacksonville, Florida.

Audi Jacksonville is owned by the Hananiya automotive group, Mr. Hananiya himself and his son,

Jack Jr. As far as what we sell a month, we usually do around 80 to 100 used.

And then a good new car month would probably be 50.

So we always aim for around 150, 200 on the summer, maybe a month.

As far as issues that we have in the dealership,

obviously I would say that every dealership will say the same thing.

It's hard to get inventory.

You know, inventory is always a tough thing to get,

especially good inventory for the right price.

It's a huge problem today. Yeah, used and new for that matter.

Yeah, definitely. I guarantee if you ask anybody, they're going to say that.

It's always tough working with the correct lenders for the right customers on the right vehicles.

Because we're what we call a full spectrum dealership.

We have $2,000 wholesale cars and I have $400,000 Rolls Royces in the showroom.

So I have to have a full portfolio of lenders to be able to sell the cars that I carry in my

inventory. Hey, where do you find $2,000 used cars in your Rolls? That's awesome.

Yeah, so yeah, our GM is a big time used car guy, huge used car guy.

His name is Stefan Gromberg. He's been with the store forever.

But he loves selling every car that we can get our hands on,

even if we got it for $100 and all it needed was a new interior or tires or whatever.

Props to the Hananiah group for not having a coronary when he used car manager buys a Rolls

Royce at an Audi dealership. But that is a sign of confidence.

Not every dealer group would feel good about that. There you go.

It's got prime placing. Well, that creates some challenges having from Rolls to a 2K car.

That does create some challenge and some friction in the buying experience potentially.

When you think about getting that buyer approved, it's very different a Rolls Royce to a first

time buyer, a luxury high end exotic to someone that's looking for that second car even.

Absolutely. Yeah. And I mean, we carry cars of all ages, miles. I mean,

200,000 mile trucks, one GMC Sierra. I've got to sit in my lot. I mean, I,

you know, I don't discriminate. I'll take it if I could sell it.

So Evan, one of the most talked about issues on Daily Dealer Live

is affordability and automotive today. I spoke to Afson,

matching a consumer to a vehicle that they can afford. But that's not the only challenge.

Getting deals done approved quickly. Talk to me about that as a challenge at your dealership.

And what are the potential costs of that? Yeah, it's a great question. I mean,

you know, we're obviously always just trying to sell, you know, the best car

at the time that we can, whatever we have in stock. So, you know, if we have customers who come in

and they maybe want a specific vehicle under $20,000 SUV, they're not really brand specific

or brand loyal. We've got tons of options, you know what I mean? But some banks, they're very

picky about what type of cars we put certain customers in, you know, first time buyers,

you know, subprime deals, no money down. You know, you may have a Toyota that doesn't book

out well at all, but you have a Honda that books out amazing. And you just didn't know

that that's what the bank essentially wanted the customer to be in. So, so that's something to

look at too is, you know, the right, the right, and a lot of managers, maybe they just don't know,

like what's the right inventory to put this, this customer. You know, but also once you,

once you land them on that, you got to move quick, you know, you got to, you got to strike

while the iron's hot. So, having tools available to you to be able to do that is huge.

You make a fair point. Not only is affordability the issue, but speed is a friction point if

it goes slowly. So, to your point, if customer comes in, they're on one vehicle and you walk

down the road from a financing approval standpoint and it's the wrong road, you actually have to

back all the way up and then resubmit or resubmit multiple different times. That friction point can

cause frustration for consumers that ultimately leads them to go buy a car elsewhere, fair?

Absolutely. Because you know that if they're trying to buy a car, someone else in town is

going to approve them. Someone else in town is going to find a way to get the deal done.

And if you let them leave, I mean, the odds of them coming back, you know, probably not that great.

So, you got to. Success in 2026 in automotive is speed to decision and the accurate decision

and being able to relay that, convey that to the consumer. Because without that, they just get

frustrated and go somewhere else. What do you think that costs a typical dealer? And what did

it cost you before you had a better solution in terms of deals at your store, Evan?

I mean, it probably costs minimum. If I let a used car sit another extra day, at least a thousand

dollars, you know, how does that buy, you know, 100 cars? I mean, it's a lot of money. I mean,

hundreds of thousands of dollars, just sitting on the lot aging. So, you know, time to sale is

kind of like the biggest thing we look at here. How fast can we get the cars through the shop,

get them through detail, get them on the lot, get them photoed, and then get them out of here

once we have someone, you know, ready to make a deal on one. Ben, give us your perspective on this.

You have a unique perspective from an OEM standpoint, banking and tech. Affordability is

a major issue in 2026 and being able to quickly get to the answer is everything.

Yeah, we agree. I mean, to piggyback off what Evan said, right, our approach is how do we empower

Evan with as much information as possible to help him with that borrower against all of the

vehicles that he has in inventory, right? So, our approach is a little different, right? I think

the industry started with this kind of single offer, single vehicle, and there was a lot of back

and forth between Evan and the lender, right? And our approach has been a little bit different,

right? It's like, how do I give Evan as much information as possible right from the start,

right? So, we kind of think about it on a multi-multi-basis, right? Multiple offers

against multiple pieces of inventory, and then he's enabled to kind of figure out how to solve

that problem in collaboration with his customer, right? Hey, what's the best fit for both parties?

So, our approach is really through sharing of information, giving him access to, like,

the decisions that we'll make with rates and payments on any kind of, any car that he has

access to. So, tell us from a process standpoint, I get what you're saying, a single customer,

single offer, and you've either picked the right vehicle and you've either got the approval

or you don't. And if you don't, it takes extra time to go back around. And ultimately, the customer

can get frustrated and leave. You're talking about multiple offers, you're talking about a lot more

information as a process. What does that look like, Ben? How do you do that? Yeah, I think for us,

it starts with a softball. So, kind of traditional kind of inputs, right? All of the inputs around

the borrower, we integrate with their inventory, right? So, then we can kind of access the car,

understand the collateral value, et cetera. And that's really the kind of the foundation for how

we do it. We're 100% automated, right? So, as Evan touched on 24-7, seven days a week, you're

going to get this automated response. And then what we really do is, like, our kind of mission

is really kind of best rate, best process, right? So, we're trying to do anything we can

on a proactive basis to help our dealers, right? Help our dealers, help our customers.

And that would look like, hey, if there's a stipulation, let's make that like a risk-based

stipulation as opposed to like a blanket stipulation, right? If it's a first-time buyer,

whatever the scenario may be, let's just not stick kind of this blanket rule against all

applicants. Let's be really smart around it. And whatever we can do to help kind of take

that friction off our partnering dealers, let's do the work ourselves, right?

So, let's back up just a little bit. You mentioned soft pull first. Tell us the

difference between a soft pull and a traditional pull. And then how does the consumer engage with

you? Is it Evan bringing that consumer to you, and then you provide all the options, or are they

accessing your portal in another way? We're 100% dealer-facing, right? So, it's Evan's customer,

right? Evan is kind of managing that kind of experience. From our perspective, the soft pull,

it really doesn't behave any differently than a hard pull. It's the same information.

It's so close. Our customers love it when you tell them that. They just love it.

I think that's a really good point, right? Hey, don't worry. It's just a soft pull.

The consumer experiences, it's opaque on the bureau, right? You can do it.

You know, further down the process, it'll be a hard pull to kind of like solidify it.

But up front, it's soft pulls around that we go.

How do you explain that difference, Evan? Because I do think it makes a difference.

Amazing. Let me tell you, especially when you have a customer who's kind of like,

I don't know. I mean, we like the car with the numbers, or maybe they're kind of wishy-washy.

Hey, don't worry about, sir. What we could do is we could just do a small little soft pull real

quick. That'll give us an actual rate in turn that we could work with. So we could show you some

more real numbers instead of just giving you an estimation. And then from there, the deal's pretty

much done. Hey, great news. We got it approved. It took us two seconds. Here's the actual rate in

turn. You ready to take it home today? So you do the soft pull, and then it shows offers on multiple

vehicles on your lot. And then the consumer picks, or how you go into the upstart portal,

I would assume. And then what does that look like, Evan, from your side?

You first have to have the customer information fully loaded in upstart. And the easy, the easiest

way to do that is I put up starts credit app on my website. So when the customer's in my showroom,

I say, Hey, great, I'm gonna send you credit up real quick. Pull on my phone. Here's the credit

app, send it to their phone, or I hand them an iPad. The credit app automatically pushes an

upstart. I just put the vehicle, the money down, upstart books out all their own cars. By the way,

how clutch is that? Very cool. I'd have to go cold JD power and figure out the retail. And if

it's got a snow plow, it's like, I don't have to do all that. They just figure it out themselves.

And then all of our fees already loaded in there and the doc stamps and everything. So

literally just hit submit and then you'll have a decision within 10 seconds.

Ben, how are you delivering that information so quickly? I mean, there are so many lenders out

there. Dealers will go into the various platforms. I won't even list them all. And it can take days

sometimes to get an approval. Certainly hours. How are you delivering approvals that quickly, Ben?

Yeah, I mean, our company is kind of built around a personal lending business, right? So we've been

in business 15 years and looked at kind of millions of kind of consumers. So we've built very, very

sophisticated kind of machine learning capabilities that powered another kind of form of lending that

we felt like was portable, right? So we bought that door though. So we've got a lot of expertise

in kind of one, not only kind of discerning credit risk, but also compute, right? So we have these

like very, very sophisticated models. And then obviously this like really sophisticated way of

moving quickly. So that's the outcome, right? That Evan's experiencing, right? You know, kind of

looking at alternative data points, looking at lots of information, not just a credit score,

and then having a really kind of robust kind of technical infrastructure that can return responses

really quickly within our experience. How does this platform and the speed, that fast response,

how do things like steps, income verification, residency, all that stuff, how does that play

into there? Because that's another friction point in a lot of transactions with traditional lenders

and platforms. So honestly, I don't really, I don't really get a lot of steps from upstart and

the ones that I do get, we upload the steps and then it looks like funding usually clears them the

next day, or kind of lets us know. So that does throw a little bit of a wrench in it, but it's

only one it's like proof of income. I mean, if you have proof of residence, or if you have like

proof of insurance, that's easy. I could do that right away. But you know, proof of income,

you know, you got to look at the income yourself, make sure that it actually jives,

because you know, customers are not always the most forthright about their income sometimes.

So if you do the math yourself and upload to upstart, it should be fine. But you know,

honestly, upstart is pretty good about steps. I really don't get a lot of deals with like crazy

steps from them. Yeah. But Ben, what do dealers most often misunderstand about risk, especially

when it comes to buyers who don't fit neatly into a traditional credit box?

Risk is broader than a credit score, right? When we look at risk, a good example is like deal

structure, along the valley, right? I used to buy paper, you know, I'd get beat up by people like

Evan, hey, here's 10,000 reasons why this is a good deal, right? So cash down for us is big.

That works both ways, right? Negative equity with great credit is elevated risk for us. So

we're going to price for that risk. But we look at risk way more broadly than a credit score.

Payment to income is another really good example, right? Where you may have, you know,

signal around marginal credit score, but then strong income that offsets risk from us,

from our perspective. So we look at the deal holistically, but I think, you know, down payment,

affordability, you know, capacity to pay those are kind of the big

indicators for us. So we don't really get too hung up on just a credit score alone.

So Ben, I'm seeing a lot of conversation around AI financing. Is that kind of the

engine or the power that's behind the upstart platform is it's utilizing AI and technology to

deliver these quick returns and speed to information to reduce the friction, Ben?

Yeah, absolutely. That's a good way to kind of think about it. I mean, for us, it's a little

bit more nuanced, right? We think about it like machine learning, right? As a kind of the foundation

of our business. But yeah, we've been at it a long time, right, over a decade. So before it

became in vogue, right, we were really pioneers from a machine learning and AI perspective.

But yeah, that is one of the reasons that we can move quickly. So again, for us, it's like,

okay, how do we deliver best rate and best process? So on the best process side, we're using a lot of

AI there to remove friction from like our dealers, like, hey, are the things that we can proactively

account for in the form of stipulations or other things such as that. So yeah, we use AI pretty

extensively as well as machine learning within our business. So Evan, when you think about this AI

financing, this machine learn process, how does this apply to some of the toughest transactions to

get done, thin file, first time buyers, those with other challenges? How does that help reduce

friction in the process for you and provide a better experience for that type of a buyer,

first time or thin file? Well, it's good because Upstart and Ben can tell you a little more about

that. They don't just look at a credit score. They look at where you're employed at and where do

you live and what awesome dealers should bring you at right now. I don't think they advertise,

but Upstart really does really well with first time buyers. They love them. Seven percent rates on

a first time buyer with no money down. I did this deal two days ago. Clutch on a Civic. But Ben,

what is it about first time buyers that makes it a good match for Upstart? Is there a component of

the AI or the machine learning that tends to better evaluate that risk and better price it?

I think it's important to call out, right? Not all first time buyers are credit equal, right?

That's true. We're going to always be looking for factors that offset that fact that they've never

bought a car before, right? They could be in the Bureau for a period of time and had really

successful payment experiences on other trade lines. There are other offsetting factors such as

stability, right? Maybe you've been on the same job, income, right? Good job. New college graduate,

income, going into a profession that we feel really good about from a stability perspective.

Cash commitment, right? Big thing is like, are they buying the right car? So we don't look at

every first time buyer the same way. We're going to look at it kind of holistically around deal

structure, vehicle. It occurs to me as we talk about this AI and this machine learning and the

speed to the decision, the reduced reduction in friction. There are some buyers that are better

risks. And I just wonder if your application of some of those technologies actually helps you

identify the ones that are a better risk than others. And if a first time buyer is a better risk,

let's get to them as quickly as we possibly can. Let's not make them go through the same rigmarole

that we do some others. I think to your point, they're not all equal, but there are some really

good risks. I want to get that person in and out as quickly as I possibly can. So I see what you're

saying, but is this kind of the future of lending in applying that type of technology to reduce

friction and get speed to that decision? We absolutely believe that is the case. And we've

been on this journey for a while, right? We're not the only ones. I think there is absolutely

this concept of hidden prime. There is a lot of consumers out there that the credit score

is just not indicative of the full picture. And when Evan's working on his side, he's like,

this is a car deal. We approach it the same way. How can we look at this more broadly

and look at other attributes around the car deal to help him make things work? But yeah,

for us, it's looking at the whole thing holistically, certainly not just credit score.

So then dealerships don't sell cars during perfect banking hours. How should lenders and tech

partners adapt to the pace of the store when you think about operating hours versus the

hours when customers are looking to really buy a car or at least make a commitment?

Yeah. I mean, I think our industry has done a good job of trying to create like symmetry between

when the dealerships are open. I remember when I first started it was we were just a really poor

job. But I think through technology, as well as people, Evan, you touched on your relationship

with your account manager. For us, access to our team is critical. It's not just technology,

it's also people that understand the car business. So for us, we kind of said, okay,

we need to be operating 100% automation 24 seven proactively trying to like take the burden off

our dealers with things that are not adding friction but not value to the car deal. And then

all of that is like supported by a team of people that really understand the business.

There are texts or a phone call away. You can look at this like AI financing and machine learning,

right? This is all facilitated by people that understand how to make car loans and understand

the car business, right? This isn't just a team of people that has no context around our business,

how it works. So I think that's a really important distinction. We're not just some like California

based software company or FinTech that doesn't understand the business. So Evan, you at Audi

Jacksonville, you had a hard time getting deals approved on weekends and after hours. It matters

to be able to have access in those times. I had a post that went viral on LinkedIn. I won't say the

bank name, but we had a reps and our GM of our Mercedes Benz store and email coming into that

week between Christmas and New Year's. It's the Super Bowl of automotive. We sell a ton of cars

during that time. And basically he had a picture of a beach chair and beach and sent an email to

our teams and said, hey, I'm going to be out for the next week. I trust you'll sell a ton of our

financing deals while I'm out. I'll talk to you when I get back. And that hit our store so wrong

for those men and women that are working after hours. You've got to know in this banking world,

pair it up to automotive. You've got to make yourself available. Look, everybody deserves

vacation. That's not the problem, but you've got to have a contingency plan and you've got to have

the right attitude and approach. Evan, fair? A hundred percent. Yeah. I remember I emailed one of

our vendors, I won't say who, on like December 20th. I needed something done, but it was important.

And then the email that came back was like, hey, I'm out of office till January 12th. I'm like,

what are you talking about? I'm here looking for a week straight. Dude, you're supposed to be my

guy. I mean, there's really only a few banks that have this auto approval, AI type thing. I mean,

I'm sure most banks are using AI now. I'm sure they've all kind of figured it out. I mean, it's

2026. I'm sure everyone's using it. But a lot of these credit unions and stuff like that that

have like the really low rates, they do not. They have people look at it. So I would think

there's a lot more automation and AI tech there, but it seems to me part of the solution that you're

bringing us, that soft pull, combined with, if I'm understanding it right, options on multiple

vehicles lined up in your inventory, that's a unique element, Ben. What made you ideate that,

where you said, hey, as part of reducing that friction and giving our dealers more tools,

we need to not only provide a quick response, but we also need to provide options so that

they don't have to keep coming back to the well to help our partners and our dealers. We want to

like figure out how to provide options and solutions. And every car deals unique. Every

customer is trying to potentially optimize for something different. So we're on this path of

saying, okay, whatever they're trying to accomplish, how do we help? How do we help our dealers?

And then in turn, how do they help the consumers? So features that we're going to be working on are

like, hey, how do we then surface payments on new vehicles that are closest to the payment of the

current trade in? So solutions are great in automotive. And we want as many solutions as we

can get that again, reduce friction, cause speed to solution allows us to deliver better to the

customer faster. But a solution is only good if it ends up getting used. What separates tools that

actually get used in 2026 from ones that are just another login and just another capability,

but just get brushed aside? Speak to that. What do you see in the financing space that gets used

versus not? I mean, the short answer is value, right? You got to deliver value. And then we

think about value in the terms of, can you help them sell another car? Or can you help them make

more money? Right? So I think for us, you know, everything that we've built or building is kind

of anchored to that. Nothing happens until a car is sold. You know, we can't win a finance contract

until that happens. So it's like, okay, how do we deliver as much value as possible through our tool

to increase our dissuccess, right? Can we help you sell a car? Can we help you make more money?

And that's really all kind of approach. So Evan, you've described this process,

this upstart process as one of the easiest tools to submit a deal. What makes a technology easy

enough for a busy sales or finance team to adapt and adopt? The less work that the desk has to do

to get the deal approved, the better. Because if the desk has two managers and they're trying to do,

you know, four or five deals all at once, it's very easy to get things kind of, you know, misconstrued

or confused. And what car is this guy buying? What car is this guy buying? How much down for

your guy versus your guy? You know, having to manually type a credit app in? Oh, awful. I'm

out. I'm out on that. I don't want to type anybody's credit app in anymore. So having all that push

directly from the website over to upstart, all I have to do is it jut up, 10 grand down, submit.

It boats it out, I don't have to pull a book, I don't have to do all this. I don't have to enter

the wholesale, the retail value, the miles. It's all there. Easy. So here's what I don't give is

we've been talking about it and you've been describing this process. I'm used to submitting

deals through things like dealer track route one, cuddle, whatever. Yeah, for sure. Is this a separate

portal from that? Or is it part of that? How do you think about that? It is a separate portal. So,

you know, we used to have the dealer track app on our website, and then we would then go to

upstart and enter all the stuff into upstart and submit it there as an extra bank to use.

That was obviously just a lot of time when upstart was like, why don't you just use our

credit app? We push it straight over to dealer track with the car, the van, the miles, all you

have to do is basically enter the structure, which I think now they actually push the structure over

too. Push the structure from the DMS or the entry point from the deals. Yeah, because upstart,

you know, the customer enters the credit app and then you literally just pick the car because

they've already got your inventory loaded and they've already got price. So you literally just

pick the car, tell them how much down and if there's a trade. And then so the multiple vehicle

options, how does that come back? So you select the vehicle and it'll it'll provide you a quick

response on that. And then does it also come back and say, but here are some other vehicles that

would work? No, so you would have to go to add new vehicle and maybe pick another car,

send that over to the AI. We currently can return up to five vehicles. So we kind of refer to this

as price multi. Got it. From a product roadmap perspective, we're going down the path of price

or right. So at that point, it's one soft pull, you know, rates and payments across all inventory.

That's kind of like where we're going from a product perspective, but current state, it's like

you can you can get, you know, up to five vehicles, I think the numbers five, but directionally,

when we're gone, but it's not an additional pull or an additional hit, it uses that initial

soft pull. That's the advantage probably. Yeah. So Ben, given everything you've seen across

automotive with your experience, what makes you optimistic about where auto retail and

financing are headed in May of 26? I feel great about it right now. I mean, I think, you know,

I've been at it a while, right? I think there's kind of a real acknowledgement at this stage of

we need each other, right? I think that, you know, the dealer obviously, you know, needs the finance

sources and then we need the dealer. And I think there's just a ton of good energy.

There's a ton of collaboration. I think there's so much innovation out there, but I think

I'm really optimistic. Like from NADA this year, I mean, the energy there around people wanting

to learn, everyone's trying to want, you know, kind of wants to get better. It feels different,

quite honestly. But I think the key there is real strong partnership, real strong collaboration.

In the past, I think you've had these ebbs and flows where the dealers have had kind of more

influence over the lenders. And then when credit gets tied, the lenders have had more influence

over the dealers. But right now it feels like we're in this like really good state of, hey,

well, it's changing. We need each other. How do we kind of collaborate to help one another,

you know, in the business that we're in together? Evan, what should dealerships be investing in

today now process people tech or lender relationships to meet the next generation of buyer expectations?

I mean, I definitely would say people first, because I feel like the car industry is always

going to be people driven. I don't think anyone's ever going to buy a car from AI. I don't think

that's ever going to be a real thing process as well. processes, you know, in this industry,

I feel like especially with the with the new age of internet and, you know, kind of the race to the

bottom per se, you know, you have to have pretty good processes in place to stay successful in

this in this industry. I mean, it's it's a saturated industry. There's a lot of dealerships out there.

And there's a lot of cars out there. And there's a lot of buyers out there. So you want to get,

you know, your piece of the pie, you got to have a good process in place. So people in process is

always huge for me. I mean, at our store, we, we train it, you know, we don't just hire anyone off

the street. We, you know, we really have to have a good culture here. And then, you know, technology

could definitely help you. I mean, the right technology can absolutely help your dealerships.

And I'll start I'll start just another piece of technology that we use every day that helps us

make more car deals. That's all there is to it, you know, well, I'll tell you in an era again,

where we're trying to reduce friction speed execution with the consumer to help create a

better experience. It is great to have an additional tool to help deliver faster,

better, more accurate information. So Ben and Evan, this conversation we've had today is exactly

the kind of conversation we need to be having an automotive, not a, not a theory, not a product

pitch, but a real experience of a dealer yourself, Evan, who's reduced that friction

to deliver better to the consumer and help not only make more money for the dealership,

but create a better customer experience. So thank you both for being on the show today.

Thank you. Thanks for having us, Sam. Appreciate it, Ben. Thank you.

"The Score Lies!" Why Credit Scores are Failing First Time Buyers (& What The Top 1% of Dealers Are Doing)

About this episode

Rolls Royce

borrower

doc stamps

AI

Request an Explanation

Want to learn more?

Help improve this episode

More from Car Dealership Guy Podcast