How It WorksFeaturesPricingSuccess StoriesHelp/Training
About
Our StoryOur Team
Coming Soon
Lotpop ConnectLotPorter
Resources
LotTalk PodcastLotTalk BlogVideosNewsletterWTF Book
Log into LotWalkBook a Demo →
LotTalk · Season 3 · Episode 13

The F&I Revolution Is Here: How AI Is Turning Dead Deals Into Done Deals

Danielle Mills Walden and Phillip Greer of Upstart join Chris Keene, John Anderson, and Renaldo Leonard to show how AI decisioning is pulling the waiting, the stips, and the friction out of the F&I office.

🎧 Listen
Watch
The short version

AI is changing F&I by replacing the FICO grid with instant decisioning. Upstart's Danielle Mills Walden explains that their AI weighs around a thousand data points per customer and returns a decision instantly, on a soft pull, so the dealer sees where a buyer lands before any hard inquiry. Phillip Greer walks the deal: a customer builds it on the dealer's site with a KBB trade value and a real soft-pull payment, the BDC calls with an actual number instead of are-you-still-in-the-market, and the desk holds the approval before the customer arrives. The AI also verifies income behind the scenes, so a deal that would draw eight stips from another bank might need only proof of residence. The result is the dead Saturday-night deal done by 7:45 instead of the dealership closing at 11, and thin-file buyers like a new nurse getting approved on true risk instead of walked.

Key takeaways

What you'll walk away with

  • AI decisioning weighs about 1,000 data points and answers instantly. Instead of a FICO grid and a back-and-forth with a buyer at the bank, Upstart's model reads the whole borrower, schooling, payment consistency, work history, and returns the decision in seconds.
  • The soft pull changes the order of the deal. Dealers see exactly where a customer lands before a single hard inquiry. No more shotgunning credit across seven lenders and handing the customer a stack of adverse action notices instead of an offer.
  • Fewer stips means fewer dead deals. The AI verifies income through the work number behind the scenes. Phillip cites a deal where the only stip was proof of residence while another bank wanted eight steps from the same customer.
  • The 7:45 Saturday deal is the whole pitch. Customer walks in at 6:59, offer lands by 7:03, signed by 7:45, everyone home by 8:00 instead of 11:00. No waiting for a bank to open, and the offer holds.
  • The dealer stays in control, and thin files get a fair shot. Upstart presents the decision, the dealer chooses whether to use it. A new nurse with little credit history gets read as a trustworthy borrower on true risk instead of getting walked on a thin file.

Episode chapters

Jump to the part you need. Timestamps match the audio and video.

  1. 00:14Cold openRe-recorded intro and meet the Upstart guests
  2. 03:25Where AI shows up in F&IRenaldo tees up the question
  3. 03:581,000 data points, instant decisionsDanielle on AI decisioning vs the FICO grid
  4. 08:19The credit app and the soft pullHow the decision gets powered, online or in the showroom
  5. 10:02Who owns the customerUpstart empowers the dealer, not a Carvana play
  6. 12:47The nurse exampleApproving thin files on true risk
  7. 14:13The road to the sale, rebuiltFrom blank BDC calls to a desk deal in hand
  8. 16:57Switching cars on one approvalPricing five vehicles, soon the whole lot
  9. 19:11Why Upstart bet on AIFrom personal loans to automating the steps
  10. 21:26One stip instead of eightWork-number verification kills the paper chase
  11. 25:56Full spectrum, not captive baitWhere Upstart competes and where it does not
  12. 30:56The power tool and the carpenterChris closes part one, part two tomorrow

The slowest room in the dealership

Most customers would rather take a root canal than sit through F&I, and the hosts have lived why: hours of waiting, credit shotgunned across half a dozen lenders, stips trickling in for days, deals dying at the box after everything upstream went right. In this episode Chris Keene, John Anderson, and Renaldo Leonard bring in Danielle Mills Walden, head of operations for the automotive side of Upstart, and Phillip Greer from Upstart's dealer success team, to walk through what AI decisioning actually changes inside that room.

A thousand data points, decided in seconds

Traditional lending reads a grid: where does the FICO land. Danielle describes the difference: Upstart's AI looks at roughly a thousand data points to judge whether a borrower will repay, and the decision comes back instantly instead of through a back-and-forth with a bank rep. It runs on a soft pull, so the dealer sees exactly where the customer falls before any hard inquiry hits the bureau. Her example is the new nurse with a thin file: where another lender sees no credit history, the model sees years of schooling and consistent phone, credit card, and utility payments, and approves her on true risk instead of walking her.

Phillip adds the dealer side: the AI values the collateral where the customer actually lives and shops, builds the offer, and surfaces the F&I products the buyer qualifies for, which means the call (the money the lender gives the dealer to work the deal) can come in higher than a generic shotgun offer. No favors, no shaving points over the phone, no waiting for the bank to open.

We empower the dealer to take care of the customer. We are not interested in taking that business away from the dealership.

The Saturday night test

The picture that sticks: a customer walks in at 6:59 on a Saturday, the store closes at 7:00, and historically the F&I manager is there until 11. With instant decisioning, the offer lands by 7:03, the customer signs by 7:45, and everyone is home by 8:00, because nobody is waiting on a buyer at a bank that closed hours ago. That is what turning a dead deal into a done deal looks like in practice.

The road to the sale, rebuilt

Phillip contrasts the standard funnel, a shopper sends leads on six cars, the BDC calls with nothing but are-you-still-in-the-market, the salesperson lands a blank intro on the wrong vehicle, with the Upstart digital retailing flow:

  • The customer builds the deal online: KBB trade value, real soft-pull payment, credit app submitted without bureau damage.
  • The BDC calls with a number: I can get you done for around 743 on this deal, soft pull, no damage to your credit, beats every are-you-coming-in call in the stack.
  • The desk holds the approval: the sales manager has the desk deal in hand when the customer walks in, and can TO early instead of late.
  • One approval, many cars: the family shows up needing a minivan instead of the truck? The same approval prices five vehicles today, with the whole lot coming, so the switch does not restart the deal.
  • The box gets a head start: the manager sees the full call and can pre-load paint and dent, key, or tire and wheel coverage that fits the buyer before they ever reach F&I.

One stip instead of eight

The part that made the hosts sit up: verification. The AI confirms income behind the scenes through the work number, matching the credit app to actual paychecks, so the stip list collapses. Phillip cites a recent deal where the only thing Upstart asked for was proof of residence, a driver's license, while another bank wanted eight steps on the same customer. Every stip is a day the deal can die, and every chased bank statement is a customer rethinking the purchase. John's reaction is the GM translation: he ran a floor with no F&I office at all, three sales managers with iPad menus, and a system like this would have matched that floor perfectly.

Where Upstart fits, honestly

Phillip is direct about the lane: Upstart approves full spectrum, the 800 buyer and the surprise approval alike, but it is not chasing the captives' 0.99 specials. The play is the deals dealers were losing anyway, to time, to unclear qualification, to nobody being available late, plus fair access to credit based on true risk instead of hard-pulling a struggling buyer over and over. Why automotive at all? A multi-trillion dollar business, candidly antiquated, where the AI that worked in personal lending had the most untapped ground. As Chris frames it with his standing analogy, the power tool did not get rid of the carpenter, it made the carpenter faster and more accurate, and that is what this does for the F&I producer. The same lesson the crew keeps hitting on the inventory side: the tool serves the operator, not the other way around.

The Monday-morning action plan

What to do with this episode this week:

  • Time your box: pull your last 20 deals and look at minutes from desk to delivery. If the F&I leg is the longest part of the deal, that is your friction, and your CSI already knows it.
  • Count your stips: how many deals last month stalled waiting on bank statements, proof of employment, or references? Each one is a candidate for automated verification.
  • Audit the shotgun: if your desk hard-pulls across the lender stack before knowing where a customer lands, ask your lenders what soft-pull visibility they offer.
  • Walk your own DR flow: build a deal on your website like a customer. If the payment is fiction and the lead arrives blank, your BDC is making the are-you-still-in-the-market call by design.
  • Catch part two: the leadership side, how to actually get your store to adopt this, is the next episode with Danielle and Phillip.

Get insights like this every week

Join the Insider Newsletter. Inventory tactics, podcast recaps, and dealer strategies. No fluff.

Transcript is auto-generated from the episode recording and lightly formatted. It may contain transcription errors.

Chris Keene (00:14): Hey, welcome back folks. You have tuned in again to Lot Talk powered by Lot Pop. I'm Chris Keen, one of the co-hosts. Now this probably looks a little bit different, but we are still in season three. We're on the most recent episodes, but as y'all know, I live in Oklahoma. And sometimes in Oklahoma, the old hamster wheel just messes with my internet. And when we recorded the other day, lo and behold, half the intro didn't even record. So I wanted to make sure that you guys understand we are still in this topic of AI. And just here in a minute, we're gonna bring back on stage, we're gonna bring my great co-host, Mr. John Anderson, Mr. Ronaldo Leonard, but we have two very special guests. And now we're venturing into the world of FNI, but also weaving in AI. directly into our finance and insurance departments. Folks, you're not gonna wanna miss this because today we have Danielle Mills Walden. She is an ex tennis pro. She is a thought leader. She hosts another podcast around leaderships in the dealership. And then she's also gonna be joined by Mr. Philip Greer. Now, both of these folks are from Upstart. Danielle, she is the head of operations for the automotive side of Upstart's financial arm. And then Philip, he is also part of Upstart. He's one of the managers there on the dealer success side of things. These folks are going to bring the heat and they're going to share with us today, not only what Upstart does and now I don't want you to think this is an infomercial for Upstart, but we did ask a lot of questions here about how their product works, but it's all woven back into efficiency, time, and how they use AI to not only benefit the dealer, but also benefit the best end users out there for all of us, which is that consumer buying public. So buckle in, get your notepads, Get your AI note takers rolling because you're not going to want to miss what Danielle and Philip from Upstart have to share with us on leadership and how utilizing that AI within the finance and insurance departments of our dealerships, how it's helping dealers improve exactly how we handle probably one of the most challenging pieces of our business right there in the FNI department. Tune in, I can't wait to hear from you guys. I can't wait to sit there and get more of that mail in. Send your thoughts, send your feedback. Always to lottalkpodcast.com. But buckle in guys, here we go.

Danielle Mills Walden (03:04): Chris, I'm super excited to be here and to meet all of you. I've been watching some of y'all's episodes and this is going to be a real fun conversation.

Phillip Greer (03:12): Yes, absolutely. I second that. I am pumped to be able to talk to you all about what's going on in FNI and everything else. just keep in mind, Danielle is the one with all the knowledge. I'm just here to be a pretty face.

Renaldo Leonard (03:12): You the eye candy, huh?

Phillip Greer (03:12): Yeah.

Renaldo Leonard (03:25): see. So last couple of weeks, we have been involved in just kind of entrenching ourselves into the process dealerships could go through to implement AI into their daily And of course, expertise and lane you stay in primarily is in FNI. Give us some ways that AI is surfacing inside of the FNI office. and what you find most exciting about that today. And either one of you could jump on there. Yeah.

Danielle Mills Walden (03:58): I'll kick it off and then I'll hold. Okay, awesome. Yeah, I'll kick it off then I'll hand it to Philip because I know he's going to have some additional things that I, that I won't touch on, but what we're seeing AI come into the FNI space in a lot of different ways. One of the main ways is from the lender standpoint of how we are automating the decisioning of customers. So traditionally when a lender is going to approve or deny a customer, they're going to look at, you know, a grid and see where they fall in from like a FICO standpoint. Whereas with us, there's like a thousand different data points that our AI and decisioning is looking at to determine is this borrower going to repay this loan. And then off of that, it's happening instantly. where you may have to go back and forth with somebody with a different lender, with ours, it's instant. So that's one way that the AI is kind of playing into things. I'm going to kick it over to Philip because I know he's got a couple other ways that he can touch on.

Phillip Greer (03:58): Yeah. So we're to go like, like Danielle said, look at all those data points very rapidly and put it in front of you. We're also going to use AI throughout your entire experience. So as we give an approval, we're also looking at the credit quality of that customer and the collateral building it together. And if you think about it from a top track, like maybe AI is there, maybe you're using it to generate emails, et cetera, et cetera. What if we use it to say, instead of having to use you, your mom, your sister, and everybody else to get an approval. we can dive deeper into your background to give you better access to the credit that you need. And then we use AI to power what your offer should look like and then what all the F &I products are that you could qualify for. We give that information to the dealer. We give them a call that says, hey, you're looking at Ronaldo. This person is great. Even if his credit might be a little bit challenged on the front side, we were able to dive in and see he's made all of his payments on time. He goes to college. He is a respectable citizen in his community, et cetera. So we're going to base it on true risk of that customer. And because of that, our call, meaning the amount of dollars that we will give the dealer to work a deal, is going to be higher than where you may get it from somewhere else. In addition to that, the AI automatically generates what the collateral is worth. It automatically goes in and says, hey, this is the actual value of this vehicle. This is the actual value of the vehicle where the customer is based. We're selling it in this specific space. So yeah, we can give a little bit more money where somebody else won't. Instead of being a generic offer that you hope is like a shotgun approach, we're able to have a more sniper approach where we're able to hone in on specifically what has to happen, and it happens in seconds. The additional benefit to this is you, the dealer, get to build your deal. So instead of me having to call a bank rep and say, hey, I need a favor, hey, I need to shave a point, hey, something's going on here, can you give me a little extra, you're good with me, I'm good with you, I helped you with that car deal for your sister's niece, none of that has to happen. It's a Saturday night, seven o'clock, customer walks in, we've all seen it, right? They walk in at 6.59, you close at seven, you're angry. Now, I'm there till 11. I gotta find dinner to eat, I gotta tell my wife who's gonna be mad, my kids are gonna be going to bed without me, et cetera. But what if we give you an offer on a softball by 7.03, you're able to get that customer sign by 7.45, you're walking out of the dealership at eight o'clock. That's the difference. Because you don't have to wait for the bank to open, you can still roll that vehicle, you know the offer's gonna be good. we speed it up so much that there is no hesitation in wanting to use the product. So as we look into adoption across the dealer base and the dealer groups that are out there, it's about, you using AI the right way? We all go into chat GBT and type out a message, right? Tell me what I should say to John to tell him that I really like his glasses. That's one thing. But what if you say, AI, John has on a shirt that says WTF. I know that he has Indiana stuff in the background. He must be from the area. He likes the Colts. Give me something that's going to attach to John and make him realize that I stick out for him. That's going to be a better email. And that's the equivalency of what Upstart does for a dealer when it comes to making an offer. We get to know that customer. We get to know that dealer. We put the right stuff in front of them at the right time.

Renaldo Leonard (07:55): So this is something that is going to be, I guess, implemented into the process. Once someone has been there at the dealership, they picked out a vehicle, so forth as on, as they go into the box, we add that, I guess, the basic data, the five liner, and then we're generating everything else from there.

Phillip Greer (08:15): Yeah, yeah, I'll let you take that one. feel like I just yapped a little too much.

Renaldo Leonard (08:15): Okay.

Danielle Mills Walden (08:19): No, you're good. Yeah. As long as we have the credit app of the customer, that is what ultimately is going to power that decision. And we can get that credit app either through the website.

Renaldo Leonard (08:19): Yeah.

Danielle Mills Walden (08:30): how a lot of dealerships are collecting credit apps or in the showroom floor, collect it from them on site. So once we have that, then we can get that decision back. Another thing is we do a soft pull, which dealers love because depending on the credit of the store, we can see exactly where that customer will fall. And then before they actually do a hard pull with all their other lenders, they can get an idea with where we stand and see how that compares. So yes, it's very seamless, very easy.

John Anderson (08:57): You guys, so you guys are, you guys are lending across the board to it. Don't I recognize the name, right? You college loans, all that stuff. You guys are into all that. Aren't you, do you have a broad spectrum?

Renaldo Leonard (08:57): Gotcha.

Phillip Greer (09:07): So we do have a broad spectrum. College loans, don't believe we're in right now, but we do personal based loans. We do HELOC, we do auto. We have some other stuff that's coming to market soon where it'll give access to credit on a smaller basis. So you don't need to have a ton of money. I just need a little bit and I need it now. Boom, we can get you approved. And the AI powers all of it.

John Anderson (09:07): You're not in, okay. Okay. So what if I'm a customer and I'm out there and I'm using, I'm part of the zero click and I'm going straight to AI and I'm looking for a vehicle. And then based on my AI search, it gives me a couple of dealers where the best value of vehicle is at. Can I reach out to you directly or does it have to be dealer? So if I want to, If I want to research my own FNI instead of relying on a dealer, is that something I can go directly to you or does that have to come through a dealer platform?

Phillip Greer (10:02): Man, John with the heavy hitters. wants to come in and ask all the tough questions. I think the best way to describe it, John, is we empower the dealer to take care of the customer. We are not interested in taking that business away from the dealership. We want to give the offer right to them. So you would reach out to that dealer. And even if you go on their website, we have a digital retailing product also where you can go on and build it as a dealership. like, let's say that I go to Greer motors.com. I go on their website, they're powered by upstart. You put all your information into our DR platform.

John Anderson (10:02): OK. OK. Okay.

Phillip Greer (10:31): It pushes into your CRM, it'll push into your DMS, it comes into our system. We spider it out to where it needs to go. And that customer has one seamless experience and we power it all behind the

John Anderson (10:42): Awesome. That's kind of what I was looking for. So if I'm, if I'm a customer that chooses to do a virtual shopping experience and you've partnered with a dealer, then as I get that information, I'm going to find you guys. I'm going to find you guys as well. Right. So that I can, I can, I know that dealer is working with you guys. I can go to that dealer because that may kick me over to, you know, when I'm, maybe I have three or four vehicles and I'm looking at what decision I can make that may, that may move me towards that dealer with that decision.

Renaldo Leonard (10:42): Yeah.

John Anderson (11:10): because they've partnered with you all.

Danielle Mills Walden (11:12): Yes, kind of. So think of it from this standpoint, the dealer is going to be the one in control of what the decision looks like that's presented to the customer. And they're the ones that are going to select if Upstart is the best fit. So while we are going to present a decision to that customer, the dealer is the one that gets to decide if they want to go with us. So the dealership is the one that's truly in control there, not the customer standpoint.

John Anderson (11:12): Okay, right.

Danielle Mills Walden (11:36): more customer interface would be more like a Carvana experience where you're picking a vehicle, you're getting a lender. is more the dealers are the ones that are empowered to decide if they want to present that upstart offer to the customer and hopefully go with us. But they don't have to. They have their other lenders as options, but they'll get to get that soft pull and see where that customer falls. And then they can decide, OK, I want to present this offer to the customer.

John Anderson (11:59): And also just to go back with how you initially started. So when you're looking at an individual, when you're looking at a customer for a dealership, you're pulling in, you're going in and searching through AI and you're pulling in all kinds of factors on that customer. Correct? So it's not, yeah. Okay. Awesome. So, so

Danielle Mills Walden (11:59): Yes, thousand data points.

John Anderson (12:20): As you said earlier, their score could potentially not be the best, but you find other factors in them, as a, as an individual contributor in their, in their community is the type of person they are that could give them, that could still rank them high enough to get a better score. Even though they're, even though their credit score doesn't, specifically put them there because of the other factors you're pulling in that could rank higher for you guys. That's awesome. Good stuff.

Danielle Mills Walden (12:47): How early? One good example I could give you just to give you like, you know, a visual is like, let's say like there's a nurse who just got out of nursing school. She doesn't have a lot of credit.

John Anderson (12:47): Good stuff.

Danielle Mills Walden (12:56): and maybe she's a thin file, we are going to look at all the things that this nurse did to get to this point and not look at her as like maybe another lender would say like, this person doesn't have a good credit history. We're going to view this as a trustworthy person who's gone through a lot of schooling, who's been consistent on her phone bills and other bills, credit card bills, et cetera. And we're going to give that nurse this opportunity where she wouldn't necessarily get one elsewhere because we believe that she's going to repay that loan. So that's kind of a good visual to kind of give you.

Chris Keene (12:56): Yeah.

Danielle Mills Walden (12:56): in this scenario.

Renaldo Leonard (13:26): Yeah. You're taking out lot of those follow-up conversations that, you know, somewhat, yeah, the F &I manager had with the lender to get that deal hung. so, yeah, saves a lot of time, effort and energy. Let's go into the, I guess, we have the platform on the dealer's website. Customer goes on, puts their information in.

John Anderson (13:26): 100%.

Phillip Greer (13:26): Yeah.

Renaldo Leonard (13:54): What does that look like from beginning to end? And you've touched on it a little bit, but I'm just trying to get dealers out there to wrap their head around how that presents a more complete presentation for their potential clients.

Phillip Greer (14:13): So let's start with a standard road to the sale. So you're going to have a customer that goes on your website, sees the car, they're going to shop six other cars, they're going to put a lead in everywhere that they can. If your BDC has not been firing at a very quick clip, then boom, they hit you within a minute. And now you've got six, seven of these things that are coming into your phone. The problem starts there where it's like, OK, I as a dealer want to hold all of your information. I don't want to.

Renaldo Leonard (14:13): Right.

Phillip Greer (14:37): automatically give you an offer, give you a call or whatever. I want to have you come into my dealership, have you on the phone, et cetera. So that's one place that there's a little bit of an error. Then you, let's say you close them on that and you get them set for an appointment. They're ready to roll into your store. Well, they're still not full of information. They may or may not show up, right? You're to have a 10 % rate of showing across the board in each one of these funnels. So now you got them in the store. Now you have a customer who is there to see a car. Maybe the salesperson hasn't seen the actual offer. And then the salesperson goes in and does a blank intro right now. Hey, Ronaldo, I wanted to put you on this Honda Civic and you say I was here for a Ram 1500. So I don't know how we had a loss of conversation. All of these points cause contention in the process as we all know. But let's say you go to a dealer who's using Upstart DR. You go onto their website, you are able to actually build your deal out and you can put in your trade. You can get a nice little appraisal there using Kelly Blue Book and it's going to be exactly what you would get on the KBB website. Cool. I have a little bit of transparency.

Renaldo Leonard (14:37): Absolutely.

Phillip Greer (15:33): Then I'm able to see what my actual vehicle payment would look like based on what I say my credit is. Now I'm like, oh, maybe I do want to put my credit app on this website because it's all on a soft pull. Type in my information, credit app's there. What does this do for the dealership? At that point, they get a full lead. They also get a soft pull approval if there's a vehicle attached. So instead of picking up the phone and being like, hey, I saw you wanted to buy a Ram 1500, you say, hey, you're looking at this Ram 1500, I can tell you.

Renaldo Leonard (15:33): you Yeah. Got a complete package.

Phillip Greer (16:00): but I can get you done for around 743 on this deal if you wanted to come in today and it's a soft pull, no damage to your credit, let's take a look around. It's a better phone call from the BDC. It's a better likelihood that you're gonna set that appointment. Now they roll into your dealership, Ronaldo, and that sales manager is able to say, hey, I've got Ronaldo over there who's ready to do this deal, here's his upstart lead, it's already in the CRM, it's already in the DMS. You have the full lead pulled and the sales manager, if they're really good, which a lot of them are, they say,

Renaldo Leonard (16:00): Thank

Phillip Greer (16:28): here is the actual offer that I have in hand. Here's my desk deal when they walk in the door. Okay. And I know I can get this approved and bought instantly. So then you go out to the customer and you're like, yeah, that 743 is here. We're totally good on that. Let's just go look at the car. Right. I've already got all of your information in the system. That sales manager does his TO earlier in the process. So great to have you. Do you have any questions? And now here's the kicker. Customer says, oh, actually 743 is great, but now I showed up with my family and

Renaldo Leonard (16:28): Right.

Phillip Greer (16:57): I have, I might have a different need. I might need a minivan. Well, what we have to offer is we can price five vehicles currently soon to be all of the vehicles on your lot where you can go and say, okay, here's the same offer for each one of these cars that you wanted. And it's still an approval. It's still using a soft pull and we can get you bought. So then that manager, if they're intelligent, they can say, yeah, I've got all the cars of this customer needs. Boom. Let's go ahead and get them through the process. Now in that process with the manager sees in the backend, they see the full call. So they know.

Renaldo Leonard (16:57): you All right.

Phillip Greer (17:27): Okay, maybe this customer is gonna need paint and dent protection because they're getting a minivan. They got four kids. I have a minivan with kids. It's gonna run into a bunch of stuff. Boom, let me put that into the deal and show it to them right off the rip. Maybe they're gonna want key protection. Maybe they'll want tire and wheel. Let me load it up for the box. So when it gets to the box, they say, cool, they already put this stuff in the deal for me. All right, we give the dealership all that information and then it makes it faster for the customer to go through that road to the sale. Less friction, less headache. None of those, I waited four hours. And then I found out that they shotgun my credit all across space. And now I have seven AANs coming back to me and I still don't have an offer and I still got to go get a car. Remove all of that. Take that customer, look at who they truly are, give it to the dealership and make

Renaldo Leonard (18:08): Yep. seems like you've taken all of the friction out of the entire process going back to when you remember back when was this huge push for DR and then moving forward to trying to get to the situation where we were getting more relevant calls. from DR through to AI, it appears that you guys have put together the complete package that's going to help a dealer be more efficient and more profitable.

Chris Keene (18:08): So.

John Anderson (18:08): love it. I love it.

Renaldo Leonard (18:35): but also turn over that, I guess, our customer success indicators or our CSI scores. At the end, it just looked like a win-win for everybody.

Phillip Greer (18:35): Yeah.

Chris Keene (18:35): Sure. So I want to, I want to back this out. No, no, no, you're fine. But I want to back this up just a smidgen here because I love what you guys are doing. And one of the big notes I took down here is reducing the friction. Okay. But I want to back this up a little bit just to the really to the AI piece and why you guys found that so important.

John Anderson (18:35): Ronaldo, think about, I'm sorry Chris.

Chris Keene (19:11): And what it is that you guys are actually doing, maybe behind the scenes. Using that AI to help reduce that friction to help speed up the timing to take that dreadful four or five hours to go by vehicle. You know, what are why? What is it that you saw out there? Why did you say, oh, my gosh, we have got to get ahead of the curve here? This AI. This is a real thing. What made y'all do that? And why did you move forward with it?

Danielle Mills Walden (19:11): I'll take that. So we looked at the AI for a couple of different reasons. First, before we even got into working in the auto sector, we were in the personal loans and we were using AI from that standpoint and having a lot of success getting borrowers access to credit that they typically wouldn't get. So when they went into the auto space, they wanted to do two distinct things. They wanted to be able to empower dealers so that you guys can approve more borrowers that typically wouldn't get access. And then secondly, be able to make it an incredible experience for customers. So while we've been talking about AI throughout this entire conversation, it's totally infused into everything that we do. And a lot of it is stuff that we don't even think about, but it's happening behind the scenes. So the fact that we're able to approve people by way of all the different decision points that we're looking at, that makes it so. there's just so many opportunities where a customer would walk and wouldn't have access or would have their credit shotgun with all these other lenders. We're able to truly look at that unique customer and everything about them to see what is the odds that they're going to be able to repay this loan. So we're using the AI there. But the other area that we're using the AI in a lot of different ways is just how quick we are going through the actual FNI process. think about steps. Think about how long it can take sometimes to get somebody to verify if a step is actually. real or if somebody does work where they work. So we're using AI and automation to help automate that and make it faster and easier, which is ultimately going to give that finance manager more time to work with more customers. So there's a lot of different areas that we're using it for. I'll kick it to Philip because I'm sure he's got some other ones I'm not thinking of, but to make it simple, we just want to make it faster, easier, and frictionless.

Phillip Greer (21:26): Yeah, I think that that is very comprehensive, Danielle. That's great. And one of the things that you brought up there was the steps. I think that this is probably something that I completely forgot about. We'll do minimal steps. What does that mean? Right? I'm always going to have something that I'm stiff for. I saw a deal that came in, was it three days ago, where the only thing we stepped for was, I'm sorry, it was their proof of residence. Right? We didn't even step for proof of income. How's that possible? Well, we use what's called work number, we go behind the scenes and we verify all the paychecks this person ever collected. So we automatically match their credit app to their actual workplace. Then we say, hey, I don't need all this proof of residence, proof of insurance, all this other stuff, right? All I need to know is that address document, which is usually just a license. And so what that looks like behind the scenes where other banks, and I saw the other calls, there was one bank that is like eight steps on the exact same customer because they couldn't verify all of these things. That is the biggest friction point that If you're an FNI manager, you get through your deal and you got to go call them back day after day. I need bank statements. I need six months of it. I need you to send me over your last three places that you lived. I need to get a letter of reference from your boss, right? That's a car deal that might get lost. What we do is we say, hey, we don't need any of that. Give us your license. And it's not on every deal, but the AI will tell you. And as long as that customer is truthful on their credit app, they're going have a lot less steps that they are required to put in. So it's really cool to see that and see those managers eyes go. What? That's all you're requesting? And then they don't even believe it. They're like, are you real? You're like, yes, I promise you, it's all real. The AI is doing it behind the scenes. That also eliminates...

Renaldo Leonard (21:26): .

Chris Keene (22:58): Ha So that just, that really just grabbed my attention there because now that makes a whole lot of sense. mean, John Ronaldo thinking about when we was back in the dealership. And I'm an FNI producer. I'm in the box and I've got an AI system.

Phillip Greer (22:58): Hmph.

Chris Keene (23:14): That is already validated income. That's already, you know, ran P, we don't need POE, you know, proof of employment is done because they're using the work number. They're utilizing that AI listeners to be able to validate these steps, which gives me more opportunity within my menu in my FNI department, which then turn in turn, reducing that time in the box. And we all know most people would rather have And we all know most people would rather have five Novakane shots or, or, you know, root canal without Novakane that go through FNI. know, root canal without Novakane that go through FNI. So that right there, John, you know, think about that from general manager standpoint, think about that from an FNI producer standpoint. mean, you paper, I'm spun paper and all know you spun paper. Not, not to have to worry about those things because we're leveraging AI. to validate those steps, reducing the amount of friction again. John, did you hear that same thing?

Renaldo Leonard (23:14): you

John Anderson (23:14): Yeah. I mean, was getting, what I was getting ready to say, cause I, as I've listened to Philip and Danielle share about their platform, you know, I would say to progressive dealers out there, the system I had on my floor with my three sales managers, we didn't have an FNI office. I didn't want it to get a customer up from the desk and recreate that tension and take them back into an F &I office. So I had my sales managers with an iPad menu system. once we got to the place where the customer started asking questions about payment, my sales managers would grab the iPad and go out to the desk, the salesman's desk, and start talking to the customer. And then I had one assistant. for those three guys that would print the paper and go out and have the customer go over the paperwork with the customer. So I'm just thinking about this from a progressive dealer that's using that type of a system, how much more fluid and fast this would give you the ability if you were working a system like that on your floor by the time your sales managers went out to that desk to talk to the customer, pretty much all those questions would be answered. So to me,

Chris Keene (23:14): Yeah.

John Anderson (25:28): I wish I had these guys when I was back in the store as a GM running that. It would have matched my floor perfectly. Let me ask you one other question I had, because I know it seems like a lot of this has been centered around that new buyer, that buyer that might have some bumps in their credit. Is that primarily what you cater to or is it across the board?

Chris Keene (25:28): Hahaha!

Renaldo Leonard (25:28): you

John Anderson (25:56): seven, 800 buyers are welcome with upstart and, and, and, okay. Okay. So I mean, from a, from a, from a, scoring and a, an interest rate, you can compete with all the big dealers from that perspective. Right.

Phillip Greer (25:56): Yeah, Danielle, what I was gonna say to you, John, was we have the ability to be full spectrum. We approve any customer that walks in as long as they meet our requirements, right? There's gonna be some of those like really deep subprime or really, really challenged credit that you may not see, but occasionally they'll come through and you'll be like, surprise, you got an offer on this that you weren't expecting because something was missed by their regular model. So we will approve across the board, but if you have an 800,

John Anderson (25:56): Okay, cool.

Phillip Greer (26:38): John, not bragging on yourself feels like might be you you're talking about. Yes, you come through and we'll be able to get you that deal approved. No problem. Yeah, so we do compete. And I want to be honest with all the dealers out there. You're going to have captives that are going to come in with special rates, special offers, all of that. That's not where we're trying to play in the sand. We want to be completely transparent and say, here's how we help that customer have an affordable payment where they can make money. And then here's how we help the customer have an affordable payment where the dealership can make money. We were not trying to hide any of that stuff. So we're not going to like just drop it down to 0.99 because that's not our stuff. If that's your captive dealer, we don't want to take their business. We want to help you pick up those deals that you were going to miss otherwise because of time, because of lack of clarity around what that customer is actually qualified for, because somebody else wasn't available late night to be able to give you an offer. Those are those deals that we can pick up. Then after we gain that trust, which is the big thing in the dealership world that I feel like lot of vendors don't focus on, then you're like, oh yeah. Send them the rest of the stuff, because these guys are great. And that's how we build that model. That's how we build that business. That's how we go through it. So all customers are welcome. Apply, again, no damage. So even if you have a 488 that walks in, you can try and see what that credit score is going to get you if you're going to get a really high APR rate on that deal. But at least you tried, and you did not hurt that customer's credit. And one of the things that we have is fair access to credit based on true risk, which means

Chris Keene (26:38): Mm.

Phillip Greer (28:04): How do we give more consumers the opportunity to understand their credit and be in a situation that will help them rather than hurt them? And if you think about me sending your credit off knowing that it's bad and hard pulling it over and over and over again, that's not good for you. That's not us being a good steward of the people we're trying to serve. And that's not us working towards our mission. The soft pull allows that, the AI allows that, and that's how we play in the sandbox.

Chris Keene (28:30): Mmm, that's strong. That is super, super strong.

Renaldo Leonard (28:30): Wow. Yeah, it is. Love it.

John Anderson (28:32): Philip, I appreciate you clearing that up really, because like I said, I wanted that to be clear to any of our dealer partners that are watching or listening to this. I wanted to be clear of exactly what you laid out there. So I appreciate you clearing that up. One other question. What was it about automotive that attracted you guys? What brought you into automotive?

Danielle Mills Walden (28:56): I guess I'll take that. So automotive is like a multi-trillion dollar business. And with the success that they had with personal loans, which is a really hard business to get into and to have success with. they had so much success on the personal loan side. And ultimately that, that business is still, you know, helping fund all the things that we do in automotive, but they saw such a huge opportunity and such an untapped market in the, in the automotive space where if we can apply our AI decisioning, that's been successful in personal loans.

Chris Keene (28:56): Try it.

Danielle Mills Walden (28:56): applied to automotive, is, you know, as you guys all know, a little bit of an antiquated business where people don't always want to change, but if they can see the benefits, no, nobody wants to see it. Yeah. But, but what we're having to do, they're very antiquated and what Phillip and I have to do and Phillip more so than me, but every single day is with, you know, on calls with dealers.

Renaldo Leonard (28:56): No, no, we're always cutting edge.

Chris Keene (28:56): No.

John Anderson (29:34): You don't say, Danielle. Come on, Danielle. A little antiquated.

Phillip Greer (29:34): Yeah

Danielle Mills Walden (29:48): like yourselves and others traveling around the country is getting them to adopt and look at this as not something that's going to replace them or to be scared of it, but to arm them with the tools to leverage this to their advantage and to help them and so that they can serve and empower more. So I think that's the biggest misconception that we have to change. And we're doing that every single day and talking with folks like y'all helps us with that because there's so much fear in the marketplace about these AI tools and it's taken over.

Chris Keene (29:48): Yes.

Danielle Mills Walden (30:16): But if you embrace it and you go fully into it, full force and use it to your advantage, the dealers that are doing that are seeing the success. And that's what we love to talk about because it works and it's easy.

Phillip Greer (30:26): I actually thought you guys were asking us like what bought me into the automotive space and I had stories loaded up and then thank God Danielle took it. Cause we would have been another hour down the road. You guys would be like, so why does Upstart like automotive? So good on you guys.

Chris Keene (30:26): Yeah. Hahaha!

Renaldo Leonard (30:26): Cheers.

John Anderson (30:26): Yes.

Renaldo Leonard (30:26): She hopped in and saved us, right?

Chris Keene (30:42): no, Philip, I just said, Philip, missed the assignment. Let's back this up a little. So.

Renaldo Leonard (30:42): Hehehehehe

John Anderson (30:42): Yeah, no, I Philip was, I knew Philip was thinking that cause he busted out left and when I asked the question, so I had a feeling there were some stories circling in your head. Definitely.

Danielle Mills Walden (30:56): Philip's got a lot of stories.

Chris Keene (30:56): Well, I'm going to tell you this, this first segment here, you know, reducing that friction time through the FNI process, gaining trust. You know, these are things right here, listeners and viewers. Okay. These pieces right here, you can do all the right work on the acquisition of the inventory. You can do all the right work.

Renaldo Leonard (30:56): Yeah.

Chris Keene (31:22): in getting that inventory to your lot, driving opportunities to it. You can do all the right work with sales management, salespeople, BDC. You can do all the right work selling the vehicle. But then you could completely blow the whole process up with your F &I, where consumers are online. long before you get to these processes of the acquisition, you can blow this opportunity completely up and never see the opportunity. If you are not three to four to five steps ahead and not be antiquated like traditionally our industry has been listen last week, we had max on max was talking about how far behind our industry is when it comes to the technology pieces and rightfully, I'm right there in the category with you, viewers, listeners, our egos start writing a bunch of checks that we can't cash. But if we stay ahead of the curve, listen to what Danielle was just talking about. She said, we are equipping you. We are arming you. You've heard me talk about this a million times over. The power tool didn't get rid of the carpenter. We still needed the carpenter. The power tool essentially just made the carpenter more efficient. He reduced friction. He improved his time. He gained trust because of his more accuracy, because he had the right tools. This AI piece in FNI is doing the same thing. So with that being said, Danielle, Philip, I want to, I want to make sure that we unpack even more. in understanding this leadership process, but for our viewers and listeners, for you to be able to catch up and understand now the next evolution, we kind of, we kind of began with the end in mind. We gave you the Stephen Covey. We told you what upstart can do. Now we need to understand a little bit of why that's important, but the only way you're going to catch that is if you tune in tomorrow, same bat time, same bat channel, dial right back in here to Lot Talk, powered by Lot Pop, and we're gonna have our good friends now, and I'm gonna call y'all good friends, because you're part of the family, but we're gonna have Danielle and Phillip back on, and you're gonna catch part two of this, just here momentarily. Catch you guys soon.

Your hosts

John Anderson, Co-Host of LotTalk and CXO of Lotpop Inc.
John Anderson
CXO, Lotpop Inc.
Renaldo Leonard, Co-Host of LotTalk and Director of Training & Performance at Lotpop Inc.
Renaldo Leonard
Director of Training & Performance
Chris Keene, Co-Host of LotTalk and CRO of Lotpop Inc.
Chris Keene
CRO, Lotpop Inc.

Stop guessing at the slow season

LotWalk pairs the data with a coach who walks your lot every week and holds the plan accountable. That is how a slow summer turns into a strong one.

Frequently Asked Questions

Quick answers to the questions dealers ask most about this episode.

How is AI used in the F&I office at car dealerships?

Three main ways, per Upstart's Danielle Mills Walden: instant loan decisioning that weighs around a thousand data points instead of a FICO grid, soft-pull pre-qualification so dealers see where a customer lands before any hard inquiry, and automated verification of things like income, which cuts the stip list down. Together they compress an F&I process that used to take hours of waiting into minutes.

What is a soft pull and why do dealers like it?

A soft pull checks a customer's credit without damaging their score or leaving a hard inquiry. Dealers like it because they can see exactly where a buyer will land, and even present a real payment, before shotgunning the application across the lender stack. The customer gets an answer instead of a pile of adverse action notices, and the dealer keeps the deal alive.

Can buyers with thin credit files get approved through AI lending?

Yes, that is one of the main points of true-risk decisioning. The episode's example is a nurse fresh out of school with little credit history: a grid-based lender sees a thin file, while the AI sees years of schooling and consistent payments on phone and credit card bills and approves her as a trustworthy borrower. The model reads who the customer actually is, not just the score.

Why do car deals take four hours, and can AI fix it?

The time mostly burns in F&I: waiting on lender callbacks, hard pulls across multiple banks, and stips like bank statements and proof of employment chased over days. Instant AI decisioning removes the waiting, the soft pull removes the shotgun, and automated income verification removes most stips. The episode's example: a customer in the door at 6:59 on a Saturday, offer by 7:03, signed by 7:45, instead of the store closing at 11.

Does AI-powered lending take the customer away from the dealership?

Not in the model discussed here. Phillip Greer is explicit that Upstart empowers the dealer rather than competing with them: the dealer controls which offer gets presented, the digital retailing flow pushes the lead into the dealer's own CRM and DMS, and the lender is not chasing the captives' subsidized rate specials. The play is helping the dealer keep deals that would otherwise die on time and friction.

Who are the guests on this LotTalk episode?

Danielle Mills Walden, head of operations for the automotive side of Upstart and co-host of the Leadership in the Dealership podcast, and Phillip Greer, a manager on Upstart's dealer success side. This is part one of a two-part conversation; part two covers the leadership playbook for getting a dealership to actually adopt AI in F&I.