Latest / Automotive State of The Union / Credit Loosens, Glenn Pasch on How AI and Dealership Staff Work Together
Transcript
- 0:00Good morning. It's Wednesday, October 14th. It's the Automotive State of the Union. I'm Paul Jay Daly. That's Kyle Malzer. We have a special guest today. Hey.
- 0:07The one and only Glenn Patches to talk about an AI research report to help you balance how much should we give the robots and how much should we give the people? All right. We're just divvying things out.
- 0:17All the chips in, fam. All the... [laughs] Push them in.
- 0:21Strangely enough, I read a story this morning how Anthropic reported that agents had started to take over household devices and be able to run models on, like, stoves and smart TVs.
- 0:34If that doesn't freak you out, I don't know what will. [laughs] What? I'm like- I mean, the smart TV makes sense. Yeah, but the stove? The smart TV makes sense, but the stove- No...
- 0:43there's, like, computer, compute enough in a stove to handle- It talked about optimizing itself to run portions of what it needed to do on- Well, yeah, yeah, yeah, like-... multiple devices...
- 0:54local slices or, like, like, smaller- [laughs] That's true... local models. Yeah, I mean, like, you look at, like, Kimmy- Yeah... like Kimmy Three or whatever- Yeah... uh, like, it is, it- it's really small.
- 1:04Like, you can run that on a, a pretty small device. And then if you splice off, like, I only need this, this portion of it, like- [laughs]... it's totally...
- 1:11I mean, I think, I think there's a world in, like, I don't know, probably six months to a year potentially the next time Apple launches a, a new device, that there's gonna be a local, uh, LLM- Yeah...
- 1:22on that thing, that it'll totally change how- Oh, I keep hoping Apple-... how things get processed... I, I keep hoping Apple's gonna do something, but- They, they are- I just want to-...
- 1:29prioritizing hardware- Yeah, yeah... for better or worse, they're prioritizing hardware Yeah, just fix Siri and we'll all be happy. You know what I'm saying? [laughs] I can't do that.
- 1:37Just, just fix Siri- It's tough, bro... and we'll all be happy. Hey, listen, everybody. We can't fix Siri, but what we can do is bring a bunch of people together in LA on November 18th.
- 1:44We hope you will consider coming out to ASOTU Con West, November 18th in Los Angeles- Yes... right near Beverly Hills. It's our first real deal West Coast event. We have an amazing speaker lined up.
- 1:55We have Vic Keller, Jack Hollis, Steve Greenfield, Alex Lawrence, Ed Roberts, and on and on.
- 2:00You can go to asotuconwest.com and see the full agenda, see the speaker list, um, get your ticket there, and, uh, we have an awesome, uh, happy hour at the end at the legendary LA Motoring Club.
- 2:11So, um, you know, the, the vibes, California vibes all in one room, all in one day.
- 2:17We just, uh, made the time so that if you're, you're close, if basically if you're on the, the, the western half of the country, you should be able to fly in and fly out if you'd like to- Yeah...
- 2:25uh, because we know it's hard to get away. Also, also, check, check me. If you're gonna be in Florida, since we got Glenn on, if you're gonna be in Florida with me at the Pash Group's MRC- Yes...
- 2:35you can still get- On the flight... to the West Coast and hang out with us. You can see Florida and LA in one week. Lots of palm trees. Bet. Yeah. [laughs] Lots of palm trees. It's gonna be fine.
- 2:44[laughs] Oh, all right, let's talk about our first story, then we're gonna bring Glenn up to talk about this report.
- 2:48Um, auto credit availability improved for the fifth straight month in September, reaching its highest level since April of 2015.
- 2:56But as vehicle prices keep rising, more buyers are relying on longer loans, smaller down payments, and negative equity to make all the numbers work.
- 3:04The dealer track credit availability index rose 1.9% to 107.1, which is up almost 10% year over year. Wow. Subprime borrowers accounted for 17% of all loans, up from 15% of all loans a year ago.
- 3:18A record of 31.8% of loans carried terms longer than 72 months. Wow. That's almost a third now, up from only 26.9% a year ago.
- 3:28Negative equity also up to, uh, 57.8%, and down payments are down 12.8%. Oh wait, negative equity is... Now, is that negity- negative equity- It climbed. It climbed. It climbed. But that, is that the- Yep...
- 3:43the percentage of loans that have negative equity at that time? That, that have negative equity. Okay. Exactly, yes. 57.8%.
- 3:49So, um- Yeah, which means half the time you're, you've got someone with negative equity trading their car- Yeah... in your showroom. Which is where it gets complicated. I mean- Yep...
- 3:55we've been talking about this, now the data is showing, and availability is still not tightening one bit, even with the interest rate rise.
- 4:02Yeah, I mean, I think the interesting part here is, like, the banks are continuing to stay, uh, like, available for different credit tier buyers.
- 4:12You know, we were at, actually a month ago, we were at Beaver when, when Toyota extended their lease program down to, uh, like, basically 500 credit scores.
- 4:24Um, uh, you know, obviously not, like, a super ridiculous subven- uh, subvented rate, but, like, the banks are still allowing lending into subprime.
- 4:33They're trying to find that bifurcated market that is struggling with affordability and trying to get them in cars, because I think there's still sidelined buyers out there.
- 4:42The question is, is how long can we manage this?
- 4:44These long terms, extended negative equity, what's the next trade cycle coming around, and can the market handle the negative equity, the loan terms, the, uh, you know, the advances on, on, um, LTV?
- 4:58Uh, it, it's gonna be interesting over the next couple years. If we don't get an interest rate correction, um, it, it could be dicey.
- 5:06I think one of the things that, you know, I would encourage every dealer or every sales manager to encourage GM, encourage your team to do, as they're, you are sitting across from people every day, how do you make a, help, help make the consumer make a good decision?
- 5:19This is something Patrick Abat was talking about yesterday when we were with him. How do you help the person in front of you make a good decision?
- 5:27And sometimes that's going to go against, like, oh, let's just cram them in a vehicle to get a sale. But how do you make them a good decision?
- 5:35Because the dealers that were helping them do that years ago are now reaping the benefits because the negative equity, you know, if your people, you were putting them in good loans, putting them in good situations, now they're bringing their car back to you without negative equity, which makes it- That's right...
- 5:47easier for you to put them in something new, so- The education, the education in front of a customer on which car to buy, how to buy it, what down payment they need, what the best case scenario for them is, is, is, is, is the difference between just, like, accepting the, the, the quest- the, the request as what it is and being a, a consultative approach to, to vehicle buying.
- 6:06Which is the human thing- The very human thing Speaking of the human thing. Oh, good one. Segway time. [upbeat music] Good one. I had one, but that was better.
- 6:16Uh- Hey, we are really excited today to invite our good friend Glenn Pasch to the podcast.
- 6:22And, uh, the Pasch Group has just put out a, a research report on AI and dealership staff, where they compare what is ta- being taken over, what is not being taken over, and where each place sits.
- 6:34Glenn, welcome to the show. Uh-oh. Is he here or no? Thank you very much, guys. Oh, good. We got him. Oh, there he is. We got him. Good. I am. We got him. So let- let's- We got him... do this.
- 6:45Tell- tell us about- You know, Wi-Fi. Yeah, I know, I know. Tell us about the, the synthesis of the report. Why is this report important right now, and what does it cover? Let's dig in a little bit. Yeah, great.
- 6:57Thanks again for having me on. Uh, you know, I was approached to, uh, uh, write this from a perspective of the inside the dealership, what's really going on, uh, more hands-on.
- 7:11We have a lot of people out there writing and talking about AI around data, high-level tech, but no one's really broken it down to say, "Let's go talk to dealers.
- 7:22Let's go talk to vendors, industry people," and say, "How did you implement it? What were your hurdles? What were the things that worked?" Mm. "What didn't work?"
- 7:30Um, and I think this is sort of the first of its kind, where
- 7:35I talk to, you know, a host of dealers that probably oversee 200-plus dealerships, and it was fascinating to hear all of their stories, and then talk to the tech side to say, "Well, what did you promise?
- 7:48What did you tell them? How did you implement?" And I think it was, that was the biggest thing that stood out was, uh, how important planning around the implementation is- Mm...
- 7:59and also the length and ongoing attention that needs in order for AI to be successful in your dealership.
- 8:08So without giving too much of the, uh, report, 'cause I know you, uh, we've got a link in the show notes, people can go download it. But what were- Mm-hmm...
- 8:14some of the biggest gaps between kind of, like, provider promise to dealer implementation, and, and where those two things rubbed or complemented each other? Ah, great question.
- 8:26So, you know, in, in the report, there's sort of an arc that talks about the pre- preparation and wor- your workflows and cadence and connecting everything, and then what do you do with your people, and the pros and cons.
- 8:38The, there was two or three things that really stood out. Number one was the lack of planning ahead of time to really understand- Mm... why am I getting this tool? What, what problem is it going to solve?
- 8:51And plan for it. There were dealers that said, "We had to shut it off, regroup, because we looked at it as just a, a, a software update," you know- Mm... like anything else.
- 9:02And so that stood out about how much work and planning. The second piece was making sure everything connects. Uh, because you could have your AI, uh, in, let's say, in service setting appointments, but if it's not
- 9:21pushing into your service scheduler, people are showing up for appointments that the AI set, but the team doesn't have them- Mm... because that wasn't there.
- 9:28Or, you know, simple things of, um, your CRM would truncate leads after a certain period of time, but your AI was continuing on.
- 9:38And so there was a lot more connectivity and following the path to make all of these tools connect with each other. And lastly, the biggest one that stood out, probably maybe 1% of them actually had an AI policy. Ooh.
- 9:54Because they- Wow. You know, there were so many people- Yeah.
- 9:57Yeah, and there were people in the beginning that said, you know, "I looked at Cox's studies and things like that," and they're like, "85% of dealers say AI's im- important, and we're looking at it," and it's probably 100%, let's be real.
- 10:07Right. Yeah. But only 15% really have it locked in. Mm-hmm. And it... One dealer said to me, "I, you know, um, I was testing it. A year ago, I was testing it, and we were, we're planning about it."
- 10:21But he didn't realize that his team was using it to communicate. Yes. Absolutely. So he thought nobody was using it. Everybody's using it.
- 10:28[laughs] And he said, "Oh, my God, I got my hands, I gotta wrap my arms around this because we have to really decide what we put into it, what we don't put into it." You know, not just all- It's, it's a massive gap...
- 10:40but information. Yeah. And it- Mm-hmm...
- 10:41it's, I don't think it's just, like, it's employees that have registered their own accounts that didn't tell anyone that are starting to leverage AI differently or not based on a set of guidelines.
- 10:50So that level of, like, just, just the baseline of training, getting some ground rules, putting a document together, making sure everybody reads, maybe even signs it.
- 11:01Like, "Hey, look, I'm not gonna leverage my own personal accounts. I'm not gonna do these certain, uh, tasks- Mm-hmm... within AI because they, they expose critical dealership data."
- 11:11And having a paper trail for that because, you know, I, I said this a, a while back. You know, dealers have never had to deal with, like, SOC 2, ISO.
- 11:19It's never been something that dealers have even thought about from a data protection side of stuff.
- 11:23But my, my bet over the next three to four years is that's gonna be a request from a, a lot of, like, maybe outside of industry, maybe political pressure- Mm-hmm...
- 11:34and maybe FTC-type stuff that put these guardrails that software companies have had for a long time. But now with- Mm-hmm...
- 11:41with a dealer that's consuming s- and, and storing so much of this data, even though it's, it's with providers, they may ha- come under the same, uh, strictness.
- 11:49So preparing for that now and documenting those policies I think is super important. I wanna, I wanna back up to one thing you said, uh, in the report.
- 11:58You know, one of the findings, the, the first finding actually, is that you've found that the gap in effectiveness is management and not the tech. Highlighting situation- Mm-hmm...
- 12:07like, uh, some, like, one store redesigned its workflows, I'm reading from the report, redesigned its workflows around AI, and they lifted contact rates 19%. Stores in the same group on the same exact system only saw 3%.
- 12:21And I think that, that- Mm-hmm... right there is something the industry has seen before there was AI. Whenever- Mm... you're implementing a new system.
- 12:30So the stores that-- Do you have any insight in the stores that are implementing it the best way and getting that 19%? What are some of the differentiators they're doing at the outset?
- 12:40Yeah, I, I, I think it goes back to the planning. You know, we've-- al- bo- all of us have been in automotive long enough, and we can go back five years, 10 years and see whatever the new trend is.
- 12:51And everybody runs to this side of the boat, right? Everybody's gonna run... You know. [laughs] A few years ago, digital retailing. Everybody's gonna buy their car online and dealers are gonna be out of business.
- 13:00And lo and behold, 6% of people do that. Okay, great. And I think it's a little like this. I think it's the misunderstanding of how much planning and management and involvement this takes.
- 13:11The, the ones who did it the best, once they planned, but they, and they connected everything, they did two things. One is they made sure they explained to their team proactively what the tool is, how do we use it.
- 13:23To what Kyle was saying, like mapping it out. Because the initial fear is, "Oh, this is gonna-- I'm gonna fi-- I'm getting-- My job's gone." Mm-hmm. Uh- Mm. And that was something that was interesting.
- 13:34Not one person fired anyone. Now, people left. Maybe they didn't, you know, they- Wow... downsized through attrition. But nobody came in and said, "AI, you're fired." Right. Maybe reallocate. Wow.
- 13:44So that was interesting. And you have to explain how it affects their pay plan, because again, everybody's going, "Well, what if the AI sets an appointment? Do I get my money?" And so, so it was that proactive.
- 13:55And then the second thing they did is everyone who was successful said they had to have a champion in the store. Someone had to own the AI and watch it and coach it- Mm-hmm...
- 14:07and look at it, just like you would a new employee. You-- So they said, "I would..." There was a dealer who said, "I still do it after four years.
- 14:14Every day, I look at five to 10 interactions what the AI is having with my customers, and is that the way I want them to respond?
- 14:23And if not, we go into the knowledge bank," or whatever, you know, tool they have, the hub, "and we update their answers." He said, "Too often you just think, 'Oh, it'll learn on itself.'"
- 14:33He's like, "We don't do that with people. We don't just don't go, 'Uh, go talk to customers. You'll figure it out.'" No.
- 14:37Look, the everything that you're talking about, Glenn, is a universal problem that we have in implementing anything that we sign up for at NADA.
- 14:44[laughs] It's like it's always been this way, and it's just when, when you're interacting with customers at the rapid rate that AI is doing, especially in conversational AI, the gaps are so much more exposed.
- 14:56But if you get the process right here, and you implement that process, no matter what tech, what process, what thing you're injecting- Mm-hmm...
- 15:04into the business, it becomes a structural overhaul for how you integrate technology or net new process.
- 15:10And I think that this is just, uh, what I'm hearing from you is we're just exposing a gap in a very customer-facing way that's forcing any dealership to recognize what it looks like to do any sort of change management, right?
- 15:22And, and it's so much more critical when it comes to AI.
- 15:24The, the thing that caught my mind, though, is actually the fact that you said nobody released anyone due to, like, AI, which we're seeing all these larger companies, "Oh, we're restructuring, laying off 3,000 people because of AI," and we're just not seeing that in retail auto.
- 15:38Mm-mm. No. Um, y- what's your thesis on that? Is it just like, "Hey, look, there's so much opportunity for efficiency that we, we've needed it all along," uh, or are people getting placed in new roles?
- 15:48Well, think about... Yeah, well, some might. There was a larger group that, you know, reallocated some people because they, uh, sen- could centralize their service BDC and, you know, we, people could watch more stores.
- 16:02Um, but we all know, we've been in the business, like, and Kyle, you were, you know, in the store at times where you just never got to all the leads that you got.
- 16:11We all wanted more leads, but we never got to all of them.
- 16:14So now what's happening is the AI is able, and that's where that delineation of what should be human, what should, uh, be AI, they can take care of those mundane things.
- 16:23So your people now are be- spending more time talking to people who are asking questions or raising their hands or are engaging versus the mundane, "Hello, could you... Uh, hi, I'm just gonna leave my message. Hello."
- 16:36Sending emails that nobody wants. So I think they became more efficient, which meant they make more money, and the machine goes and does all of, uh, uh, all of that work. So I think that,
- 16:50to your point, is that's why I've created a s- 30, 60, 90-day implementation sort of outline, and then just even narrowed a, a pre-install checklist- Mm... because dealers need things like that.
- 17:02And as I said, it's, it's... I, I just think what we're seeing is, is that w- we, there's always been so much w- not... Yeah, waste. I'm gonna say waste, right? Uh, just dump into- [laughs]...
- 17:13the strainer and stuff just comes out the bottom, right? Our friend, uh, uh, you know, uh, uh, Brad down at, at Ferman in Florida, you know, he always says, "It's not who sells the most, it's who spills the least."
- 17:26And I think what AI does- [laughs]... is allow you not to spill, right? I can clog up because... And, and your metrics are gonna change.
- 17:34One other fascinating metric, a dealer said, you know, everybody used to always say, "How quickly can we get to somebody," right? Yes. Time, lead, respo- He goes, "AI takes care of it.
- 17:44It's five seconds now," or whatever it is. There you go. It's like- What he now measures is, he measures, what he's measuring is how quickly do we humans get to the customer when AI says, "I need a hand." Yes.
- 17:57So that's the new metric, right? H- how, that handoff metric. That's, he goes, that's the real important one, 'cause the AI's gonna have a conversation. Yep.
- 18:05But once it gets to a point where it's like, Paul's gotta jump in here, is Paul there? [laughs] Or is he wandering around? Yeah. Right? So, so these are all real fascinating things.
- 18:14But to your point, Kyle, I think this can also be a roadmap or a plan on any technology install. Yep. Glenn, thank you for joining us today. Where can people get this report?
- 18:24We're gonna link it up in the show notes so they can get it there, but for people just listening, maybe you could just say it. Sure. They can go to Pa- Yeah, just go pashgroup.com and under all of our releases.
- 18:35And then, uh, and, and, uh, Kyle earlier on mentioned our conference. We're gonna be talking about the report, uh, at the Modern Retailing Conference, which is next month in Florida. So- Yes...
- 18:45hope audience will be there. I'll see Kyle there as well. Yeah, you gotta make it. It's a great, great event. And congratulations with SoDeCon West. Congratulations with that.
- 18:52Very excited for you guys on the- Oh, thank you. It's gonna be a ton of fun... on the West Coast as well. We'll, we'll see. I think we're- Absolutely. I think, I think we can do the West Coast thing okay.
- 18:58We're about to find out, aren't we? [laughs] Oh, as usual, Glenn, it's always great to spend a few minutes with you.
- 19:04Thank you for doing the work to provide this to the industry, especially around pertinent topics like AI. Go to pashgroup.com or check the show note links below. Get that report.
- 19:14[upbeat music]