"Money's not in the CRM, It's in the DMS!”– Tracking the Pulse of AI in Dealerships | Art Dessein, Vice President of Sales at Impel

"Money's not in the CRM, It's in the DMS!”– Tracking the Pulse of AI in Dealerships | Art Dessein, Vice President of Sales at Impel Car Dealership Guy Podcast · Sep 9, 2025

Annotations will appear as you listen

0:00

43:31

Car

Ford Explorer

LIVE

If you look at somebody that gets 350 leads a month,

they probably have 20,000 records in the DMS.

So that is 20,000 people in which the average RO is 460 bucks,

in which, by the way, 55% of a dealership's DMS

are people who haven't been in in 12 months plus.

And if you think about service,

I'm gonna simplify it, some groups, and I say,

that's not the truth, but if you think about service,

the reason they went elsewhere is not because of price.

You go to the average consumer

and you're asking what the price of an old changes,

a Jiffy Lou versus the dealership,

they would not be able to tell you.

The reason they go somewhere else is out of convenience.

There's 190,000 independent repair shops in the US.

There's like 36,000 franchise dealerships,

so like they're already at disadvantage,

so you've got to provide a better experience.

Today I'm joined by Art Dessin,

Vice President of Sales at Impel.

Car shoppers are drowning in dealership messages,

sales events, service reminders,

and endless blasts that are blurring altogether,

and that's why Art's message on this podcast

cuts through the noise.

Send fewer messages and sell more cars.

Sounds counter-intuitive, but it's not.

We also dive into the real pulse on AI and auto retail,

what dealers are actually asking

and where the conversation is heading.

A big thank you to our sponsors

for making today's episode possible.

Widewell, Experian, and of course, Impel.

And now let's get into the show.

Arthur Dessin on the CDG podcast, Arthur, welcome.

Yeah, thank you.

Thanks for having me, Yossi, appreciate it.

You're all traveled out, huh?

Namad, you've been all across the coast

the past couple of days.

Past couple of weeks, yeah, for a straight two weeks,

so hopefully I got enough energy

to get to this 30, 40 minutes.

How efficient are you in the conferences?

Like how do you handle that?

I'm pretty efficient.

I mean, I got a good process in place, yeah,

from checking in, getting all the suits and shirts pressed

and making it on time and flights.

One thing you can't control are flights,

but thankfully, actually an interesting story,

my first time ever I missed a flight on the way to Namad

was because the flight, the airline's called Breeze,

it's kind of a new airline,

I don't know if you've ever flown it,

but it does direct flights

from secondary and tertiary cities,

but they left 30 minutes early.

Early.

Early, and I get to the airport

exactly at the same time,

which is 10 minutes before my boarding starts.

That's how I get to the airport.

Yeah.

And by the time I got to the gate,

no flight even on the jet bridge.

So I'm thinking I'm too early,

but apparently decided to leave 30 minutes early

because they get rewarded by landing early

at the secondary airport.

So they're like, oh, we'll leave Art behind,

he doesn't need to come on the flight.

We'll leave him to Delta,

but didn't miss any part of Namad, so it was good.

That's incredible.

Have you figured out how to travel with a suit

without it getting wrinkled?

Have you figured that out?

Yeah, I mean, no, not really.

I mean, unless you like hang your jacket, right?

And sometimes like,

they won't let you hang it and stuff like that,

but hopefully you can wear the suit,

you know, after you land and not necessarily on the plane.

I try not to wear a suit on the plane

because you never know, you know,

planes sometimes are like 90 degrees

or you're in like the middle seat.

You just, you just never know.

Well, dude, when I thought about this podcast,

it reminded me of my friend.

I have a friend that calls me

like a human RSS speed for the auto industry.

And I sort of see people in your positions, right?

So you lead revenue at Impel, right?

Leading AI company in auto retail.

But when I think about someone in your shoes,

I think about someone who is interacting

with lots and lots of dealers

face to face on a daily basis,

which to me is your form of a signal.

I mean, you have the pulse on the market

and that's what I wanna dig into here

because I think there's a lot of discussion

on AI in our industry.

We cover it a lot

and we're talking about what's changing,

what's working, what's not working,

you know, what are people embracing, yada, yada.

And that's what I wanna lean into.

And so I wanna just jump right in first to start with,

give us an overview of, from your perspective,

the state of AI within dealerships.

You can tell us a little bit about what Impel is doing,

which I'm sure most people listening

have, you know, heard of Impel,

but I wanna understand, you know,

what does Impel work on

and how is that working within dealerships today?

Yeah, so I'll give you a distilled version

of not only my perspective, but our perspective,

which is, you know, like seven, eight years ago,

nine years ago, there was a couple of AI players,

not generative AI,

but certainly AI players that were in automotive.

I think a lot of dealerships got burned.

I think a lot of dealerships realized it wasn't true AI.

I mean, it was AI,

but in a different way that people now expect AI to be,

which is, you know, these large language models

and this agentic AI.

But so I think a lot of them had this bad taste

in their mouth and similar, you know,

in automotive, there seems to be every two to three years

an acronym that everybody blatches onto.

So you remember digital retailing, you know, nobody.

I love that acronym.

CDP kind of had that.

Like now you're still kind of seeing a lot of groups

trying to push into establishing a CDP,

but that's kind of like this kind of thing.

And AI had that.

So now, I think now we're at this kind of period,

this threshold where there's some groups

that think AI is that,

this kind of fad, this kind of acronym, it's overblown.

We think that actually they're probably

undervaluing the power of AI.

I mean, if you look at even the publicly traded companies,

right?

Navidia is worth $4.5 trillion or something like that.

Open AI just raised funds at $500 billion.

So like this technology is really, really helping

like these big fortune 500 organizations,

probably not as much as one would think.

It does have some huge implications

in terms of operational efficiencies

and making people a lot more efficient.

But we now are seeing kind of the shift

in dealership groups actually seeing, you know what?

I actually do need AI and not only for communication,

but also for data management, processing, synthesizing

and also like predicting next steps, predicting behavior.

And so we're now seeing probably over the last course

of the last five and a half months,

a lot more dealership groups leaning a lot more heavily.

So we saw this kind of five, six years ago,

AI was burnt.

Then about two years ago,

there's a lot of momentum with Navidia

in the stock market and chat GPT coming out.

And then in Wayne, we thought like about a year and a half ago

for like six months and now it's picking up steam again

and mainly driven by these kind of

agentic AI applications.

Okay, why are dealers asking you for AI?

Is it just because they're competitors

using as a product or is there a problem

they're trying to solve?

Like what does that mean to dealers?

Yeah, I mean like the most like superficial view

is that you hear it all the time.

So of course you got to ask for it.

You can't turn on a news channel without hearing about it.

I think so there's a cohort of dealers

who just like hear AI and I need AI.

You need it, right?

Like you don't want like FOMO, right?

You don't want to miss.

Yeah, FOMO, you don't want to fall behind,

which that I would assume that's at least

some cohort of people.

So that makes sense.

That's human nature.

Yeah, it's like saying, why'd you invent,

you know, you ask people,

why'd you invest in Bitcoin?

Most of it's like, well,

you saw a lot of news cycles on it.

So I invested.

I mean, that's the same.

It's a parallel or it's the same.

They don't really know why they don't actually

know the underlying reason why they did it.

But then there's another cohort,

which is like they're looking at their operations

and there's a ton of efficiencies

and inefficiencies both in variable and fixed.

And a lot of it is driven by like human error

and human issues, you know, like the scale

in which like, if you look at the number of leads

they get or number of records they have

in the DMS for service, it's a gargantuan amount.

And it's really hard for them to do stuff

without automation in place

or at least be a lot more successful.

So there's a lot more dealerships now trying

to do a lot more with less.

And, you know, Gen AI and agentic AI applications

are meant to do just that, right?

Like extract all the efficiencies you possibly

can extract and do things that even

if you hired 5,000 people, you can do.

Okay, those dealers that are looking

for specific applications,

are they looking to become more efficient

or are they looking to replace talent and employees?

What are they looking for?

What are they asking you for?

Yeah, so I think there's, again,

there's a cohort that's looking for,

hey, we need to do more with less.

You know, there's some that are like,

hey, we gotta sell more cars

or we have to service more cars

but we can't take the additional expenditures

that it might take to hire like we used to.

It's pretty costly.

And the output of that human laborer

is like diminishing returns, right?

Like even, think about if you had a thousand leads

if you had 500 BDC agents

or a hundred BDC agents handling those leads,

you're probably not getting the same value

for the additional 400 BDC agents.

There's also this timidity around like saying, yes,

like can it replace people?

And so there's two sets of thought,

there's two mindsets, which is, no, no, no,

people are safe, it's just gonna make them better.

I have a different perspective.

I think, yeah, it should certainly replace people, right?

If it's doing like, and look at how far it's come

in the evolution of it just in the last two years

and what it was able to do,

like Chad GPT when it was released was great.

You could plug in something, you could get an output.

Now you can go in and say use Chad GPT, for instance,

use the HNTKI capabilities

and they can actually book reservations,

book flights for you.

You can set it to connect with Gmail or Excel

and it can actually execute on tasks.

So of course it's gonna replace people

but there's also this fallacy

called the lump of labor fallacy

which they think like technology will come in

and completely eradicate that position

in which the person will be left out of a job.

That's not really the case.

A lot of dealership groups will repurpose that person

as something that he's much better suited for.

So think about a BDC agent may not be great

but maybe that dealership lacks

and having true like a labor around equity mining, right?

Somebody doing specific equity mining outreach

or maybe somebody can spend a lot more time

being a lot more personalized

or having a lot more information service

to deliver a great experience

and not be behind a computer doing tasks

just for the sake of doing tasks.

That's kind of our vision is we think

there's dealers that want it for operational efficiencies.

We think there's dealers that want it

to kind of do more with less.

And then we think that dealers

are just trying to potentially use AI to cut as GNA.

Are you seeing or which one are you seeing more of?

Are you seeing GNA

or where are you seeing the most demand ultimately

and impact?

Yeah, right now I think it's still more towards,

hey, we need to drive more efficiencies.

When you look at the variable side,

I'll pick on the variable side and then we'll go to fix

but even on the variable side

there is thing about the introduction of BDC.

So BDCs were introduced

which is essentially an additional layer of people

who almost interfere between the dealership personnel,

the sales rep and the customer

because most BDCs handed off to the sales rep.

So now there's this additional layer

that the consumers got to talk to

before you can purchase a car

but like the closing rates

aren't in any necessarily higher, right?

They're almost the same.

So you've introduced this huge cost center

and the customer has to talk to yet another person

before you can buy the car

but the outcome is the same for the dealership.

And so there's a lot of groups

that are still looking at using AI

to reduce some of the overhead that they have

especially on the variable side.

On the fixed side,

there's a lot more focus on the customer experience.

Like when you look at service

they don't have BDCs that are overstaffed, right?

They can barely keep up with the inbound phone calls

that are coming in

much less doing the outbound work

that's necessary to try to retain a client.

So I think on the service side

more of them are leaning towards using AI

to do personalized outreach

that's actually effective to compete

with the independent repair shops

which is really who they're competing with

they're not necessarily competing with the dealerships.

And on the variable side,

they're competing with the other dealership down the street.

This episode was brought to you by Widewell

making reputation management effortless.

If you've tried things like bonuses or QR codes

to get more reviews at your dealership

and still feel stuck, you're not alone.

These tactics often fall flat

but that's where Widewell comes in.

Their platform automates the review process

by integrating directly with your DMS.

No manual follow-up needed

and every review is handled

by their expert US based response team.

So each review gets a customer response

in less than 24 hours.

One real insights instead of guesswork

with voice of the customer analytics

Widewell uses AI to turn raw customer feedback

into clear operational insights

so you know exactly what to fix and where to focus.

And here's a special offer just for CDG listeners

to get three months of review response services free

when you purchase the Widewell platform

that's three months free.

So go to widewell.com slash CDG to book a demo

that's W-I-D-E-W-A-I-L.com slash CDG

or click the link in the show notes below.

So very specific question, Arthur.

If you were today to acquire a dealership

and you have to implement your first AI tool

what would you implement?

Like what's that first thing you go to

to bring added value immediately?

Easy, service.

Service is probably the single largest opportunity

for a dealership and it's easier.

Like think about the amount of energy, time, money,

people, they throw out variable, right?

Like almost, and most of the GMs come

from the variable side.

So they're of course they're over indexing on variable.

But so if you think on the variable side

like you spend a lot of money to just sell the car once

and the person comes in they give you

so say you own a dealership, you'll see it

and I come in and I give you $70,000

for a whatever a Jeep Cherokee now these days, right?

And I drive away.

If I trusted in giving you $70,000

like you're losing me as a customer for service

if you can't retain me.

And if you look at the number of records in the DMS

it's about 60 times greater.

So the number of customer records in the DMS

is 60 times greater than the number

of internet leads they get.

But the mind share allocated to service is like 10%.

So there's plenty like 90% of their energy on variable

10% on fixed.

And so we think AI is like perfect for service

because the problem with service is the scale

of outreach that's needed isn't doable by people, right?

So like again, I'll go back to the internet leads

if you get 500 internet leads,

most dealerships would say they are doing long term

follow up correctly, right?

They're not responding fast enough.

They're not giving the personalized touch

and that's just with 500 internet leads for service.

Remember they're gonna have 60 X the volume

and that person is gonna own the car

for three, four, five, six years.

And so every year they gotta reach out

to those consumers several times.

So the scale in which the outreach is needed isn't there.

And then you have other things like,

like think about their service scheduling.

Like that's a huge point of friction.

So I don't know if you've ever,

if you've serviced your car like a regular customer.

But if you look at any of like,

there's only two ways to schedule service right now,

right?

You call the dealership and you schedule service

or you go online to their website

and you go through the service schedule.

So I'm gonna ask you a question.

Do you know what the abandonment rate,

so the people that start out trying to schedule

their car on their service scheduler,

what percentage defect or abandon it

before completing it?

On average.

I'm gonna guess like 75%.

Yeah, it's like 81%.

Oh, wow, look at that.

So they're about at 81%.

I swear, by the way, that was a total.

I was gonna say 90, but I was like,

it's probably too high, so let's go 75.

Oh, is that crazy?

Like out of all?

Well, it makes sense,

because I've tested those service schedulers,

you know, the online ones and they're not friendly.

They don't list all the work you need

or it's like, you know, misclassified

and it's coded like within a DMS, right?

Yeah, yeah.

They have like, you know, like lube oil filter,

but it says like LOF, it's like,

the hell, I'm a customer, I don't know what LOF is.

You know what I mean?

So it's just a very, it's not user friendly at all.

Yeah, but in here's the problem, right?

Like most people in the dealer world

in the automotive world, like that's their language

so they can't put themselves in the consumer shoes.

And also the average number of clicks

to schedule service is like 21 clicks.

And if you've ever known the adage of like form fills

is like every single additional field

that you have drops the conversion by 50%.

And so like, you have this huge drop-off rate in service.

So like, do you go in, you schedule service?

All the marketing that goes out there

is to convince them to either call you

or schedule service online.

So even if the piece of marketing

that you're doing works, 80% is ineffective

because the person can't figure out

what to do on the service schedule.

So you know what they do?

Like half of them end up going

to Jiffy Lou down the street, right?

Because it's like 15 minutes

don't get out of your car oil change

or they call you, but then like 45% of calls are not answered.

And if the calls are unanswered

like 60% of the voicemails are on return.

So you have all these points of defection that exist.

And so fixed ops is the quick way

and our fastest ROI I should say.

Yeah.

That makes sense.

Talk to me now if you were to open a new point

in middle of nowhere, you open a new point.

Are you implementing an AI tool or are you not?

And if you are what, which AI tool?

And I think this, this appeals more to the,

maybe call it the younger dealer

or the dealer that doesn't have such a, you know

wide book of business yet or is growing.

What's your thought?

Yeah.

I mean, I mean, it's mainly gonna be variable, right?

Cause they don't have a huge database

of people who have purchased cars for them.

No fix is still important because if you sell a car

you wanna make sure you retain that customer

right off the bat.

And it's not all about winning back abandoned customers

but on the variable side, there's still a huge value.

And AI, not only for responding to leads faster

and doing the long-term followup

and all those types of things

but it's really about making your staff

a lot more efficient and productive

in their day-to-day responsibilities.

So right now, if you sit behind a sales rep

or a BD agent or an internet manager

and you just watch what they do on a computer

most of it is like checking off the box

of the CRM activity tracking requirements, right?

They're not even, they're like painting by number.

And they're not even thinking anymore

cause it's like, hey, OC, here's the email

that you write and you send on day two.

And by the way, we've pre-written it out for you

cause we don't trust you to write your own email, right?

Here's the template that the elite has

or dealer socket or whatever it might be.

And so there's huge value in AI

to deliver amazing experiences on the variable side.

Like that's the only way you compete, right?

A Ford dealership from Ford one to Ford two

they're all selling the same, you know, Ford Explorer.

It's a commodity, it's the same exact thing.

You can compete on price, which is happening right now

but like you need to compete on amazing experiences.

That's why people go to great hotels.

Like you spend an extravagant amount to go to good hotels

because the experience they give you is 20X better

than the standard kind of hotel you might frequent.

And it's the same thing.

So you can use AI like not only to answer questions

but to deliver this amazing experience

cause the customer is submitting leads

at seven to nine dealerships.

They're all answering the same way.

And if you, you have a dealership

and you can answer in this amazing way

where you can reference and be contextual

and understand deeply what the customer is asking

and answer instantly and not only on the first lead submission

but every time the customer is willing to engage with you

you have the ability to engage back in real time

sustain in a sustained way

through that whole kind of purchase cycle.

That's huge.

So customer experience on the variable side

is the number one factor.

So my next question here is

if I took your phone or your email

and I took like the last hundred client communications

or you know, messages you have

and I ran them through AI, right?

And I said summarize this,

give me like a couple of bullet points.

Like what are people asking you?

Like what is the demand?

What is the market demanding that we don't know?

Right?

What are we about to see next year?

Yeah.

I think so all the trailing 90 days, you know

what we're starting to see this momentum

this kind of shift is a lot of groups

are trying to figure out

how to send a lot less messages to their consumers.

Like that's by and large.

Oh, it's your messages.

Yeah, yeah, every cost.

I think every dealership knows

innately that there's over messaging their customer.

I don't think any dealership be like, no, I think I'm

I don't think I'm sending enough messages.

So they're trying to,

what's happening is they're sending so many messages.

Some are good, right?

But a lot are bad messages,

not personalized, not contextual,

not timely, all that type of stuff.

And so what's happening is all the bad messages

are drowning out the good messages.

So now the good messages are becoming ineffective.

Dude.

Consolidation of messages and by extension,

by definition consolidating all the messages

means you're consolidating vendors, right?

Like, which is why some people are going to

are thinking about going to a CDP,

how I want to collect all this data on the OC

so that my marketing is more personalized.

Like that's like a very simple simplified view of the CDP.

But essentially that's what it is.

And so the same thing is happening with vendors

and tech and their tech stack is can we consolidate?

Like, why do I need 29 different vendors

to be messaging my customer?

Can't one do it or can't two do it?

And so there's this shift in a lot of dealership groups

looking for a solution that can,

that can enable them to do just that.

You bring up an extremely important point.

So I had this conversation on the podcast

a couple of months ago and we were talking,

we were discussing, talking with a consultant

that's consulting for a couple of OEMs.

And he, he mystery shopped a dealer, Florida.

And he received like 10 follow-ups from that dealer.

Yeah.

Like something crazy.

And I had a bunch of dealers reach out to me after that

who watched the segment and one specifically said,

he said, Yossi, he said, look, I get it.

But also, you know, the manufacturer is mandating

that I use, you know, all these different products.

And so there's this friction.

And I'm not completely taking the blame

and accountability away from the dealer, right?

Because you still own that product,

you can control the message at a certain extent.

But there still has to be,

it needs to be a, you know, symbiotic relationship

between the manufacturer, the vendor, the dealer.

Have you run into that type of issue

where it's just manufacturer mandated tools

that are, you know, causing too much noise

and without much signal?

I fully agree by the way with your idea of fewer messaging

because you're right.

The more you say, the less people hear.

And that's a problem, right?

Yeah.

Yeah, that probably happens more so on the fixed side.

There's a lot of OEM programs

that you have to market, you know, consumers are pretty,

much less so on the variable side.

Now on the variable side, you know,

you have the first quality response

and things like that, that the OEMs mandate

that you have to, you know, stuff things in their response.

That may not actually necessarily be great

for the consumer experience,

but we find that much less so being in impact

on the variable side.

So we still think on the variable side,

you can send a lot less message.

Now we hear some dealerships say,

oh, I have a six month follow up process

in place of my CRM.

I'm like, tell me some, it's a lead for you today.

You're still following up in six months.

Are you still in the market?

That's probably not a great experience.

If all sorts of stuff happened

that the dealership doesn't realize,

the open rates from all the CRMs is like 19%.

So like the emails don't even get delivered, right?

So there's huge issues, there's huge implications

outside of what the dealership realizes.

But yeah, it's in terms of there's some overlap on fix.

I don't know if that's ever gonna be solved.

I think you're gonna have to be able to live with yes.

Well, it has to be solved though, right?

Like I see it as, if it's not solved,

this industry will continue losing market share

to, you know, the Carvanas and Carmax

who are working with a unified system.

Like these points, I just think that it's gonna be challenging

for point solutions to survive over time in the age of AI.

And what I mean by point solutions,

I mean like bolt-ons, right?

Where you have one product that does one thing.

I just think that the industry will have to consolidate

and one system doing, you know, many, many different things

because no matter how inefficient it is at the dealership,

if it doesn't impact the consumer,

you could live with it.

You're just running a more inefficient operation.

But the second that you just said,

the customer is getting all these different messages,

now you're harming the customer experience.

Why will I want to do business with you?

Yeah.

Yeah, I mean the difficult parts on it,

at least for dealership groups,

and you mentioned Tesla and I think Carvanas or Carmax

is like, that's the brand.

I mean, they sell used cars,

but that's the brand is like they can do that.

But like, if you own Nissan and then, you know,

Mercedes-Benz dealership, a CDJR dealership

and a VW dealership, like you're,

it's essentially like owning five totally different brands.

It's like owning a Burger King, a Wendy's

and like a Dunkin' Donuts.

So I think it's gonna be hard in the,

at least in a near or medium term,

to have everything be homogenized

in terms of systems and processes.

So that's where I'm saying it's gonna be really difficult,

but it doesn't mean that you can reduce,

you know, like reduce stuff

to allow it to be a lot more effective.

And that's where AI is helping a lot of dealerships,

is it's helping them send a lot less messages.

And by the way, the messages end up being a lot more tailored,

a lot more personalized, a lot more contextual.

But really like, I mentioned service

is where the overlap is.

Like the problem with service too,

is like all the communications right now

is marketing driven.

So it's all one way.

Like you could have the best CDP in the world.

Now you could have like,

you could spend millions of dollars on a CDP

and I could have you, you know,

see exactly what you want,

when you want it,

send you the right message at the right time, et cetera.

But you can't even respond to that message.

Right, because it's all marketing driven, right?

It's an out on the Google display network.

It's an out on Facebook.

It's an OTT advertisement, right?

It's a piece of direct mail.

It's again, super tailored, super personalized,

but the onus is on you to have to take action.

You can't just get a message and reply to it

and say, yeah, when can you bring my car in?

So really that's a huge missing element right now

is like this need for two way.

Consumers are used to now like chat GPT.

They're used to like instant gratification,

instant communication.

They can get the information like that.

They got like dealerships have to provide

that type of experience for service.

They're also going to keep losing out to Jiffy Loubs

and Pet Boys and whoever else might be in the market.

Oh, and by the way, you see like everything

we just mentioned, these like point solutions

and over messaging customers

and all these different programs

that that dealerships are using and investing in a CDP.

That's exactly how we help dealerships.

And that's why we have so many groups

using the AI platforms.

It allows them to unify the customer experience

and unify and coordinate all messaging

across variable and fix.

And I think that's been a huge missing component

is that everything along that customer lifecycle

from when he enters the market,

to when he buys a car, to when he services a car

is way too Balkanized, right?

There's way too many vendors

with their hands in the cookie jar.

The experience is subpar.

It's extremely costly for the dealership

and nothing is really tether and unified.

And so that's how we help dealer

unify all the messaging and unify all the communication.

So the experience is great.

The cost is lower and the outcome is much better.

This episode is brought to you by Experian Automotive.

Like most car dealership guy listeners,

you're constantly looking for the inside edge

on the auto industry.

So if you're ready to step up your game

to the next level outpacing the competition

and building customer loyalty,

there's only one place to go from here.

Enter Experian Automotive.

Whether you're tracking market share,

boosting service revenue,

targeting the right buyers

or making smarter inventory decisions

Experian has you covered.

They're the only ones

with exclusive data across vehicles,

consumers and credit,

plus expert data scientists

who connect the dots to uncover the insights you need.

And their insights aren't just for dealer operators.

OEM, lender, agency, whatever your role,

Experian's powerful tools help you win.

Get industry leading insights you need

from Experian Automotive today.

Learn more by visiting www.experian.com

or clicking the link in the show notes below.

How well do you work cross brands?

In terms of cross brands,

I mean it's pretty fluid.

Now we have the platform as this agentic capability,

which means for instance, on the variable side,

the dealership has control around

how the AI might answer specific lead sources

or even more importantly, sub sources.

So for instance, for the first quality response,

if CDJR wants something

or needs something inculcated in the first response

and Nissan has something else,

the AI is complete or the dealership has complete control

on how the AI is gonna answer

that first quality response.

So they have a lot of flexibility and malleability

in terms of how they want the AI to respond,

but it's also gonna meet all the OEM requirements

which I think is really important

if you're a group that owns multiple brands.

Okay, so slightly shifting topic,

but I always like to think about like second

and third order effects with these things

because this is, there's a lot of investment going into it.

Dealers are adopting it, dealers are asking about it.

I see in our CDG20 group,

it's a conversation every single week,

whether it be this tool, that tool,

this new product.

And then I think about,

so I mentioned second order effects,

I think about like,

think about Osempic, right?

Bear with me here,

but Osempic didn't come out as a weight loss drug, right?

It was a diabetes, like type two diabetes drug.

And today it's like a serious,

like you know, massively popular weight loss drug.

I was reading about this over the weekend,

hence why it's top of mind.

So then I think about

what are the second order effects here, right?

Again, as someone who has their pulse

on the impact of this technology in our industry,

are you witnessing yet any second order effects?

Are you seeing anything there

that the market should know?

What should we know?

Yeah, I mean, I think, yeah, I think we are.

I think we'll send some a lot more.

By the way, my light went off, of course,

but it'll turn back on.

Okay, we like it when your room's intimate.

It's nice, you know?

That's right.

I think we're gonna start seeing

kind of those second slash, you know,

tertiary like third order effects

probably in the next like year and 18 months.

But yeah, I think there's a lot of it around,

especially as a gentakai kind of rolls out,

I think there's gonna be a lot of applications

that people can't even think about it now.

Cause these models, right?

Think about these large language models,

they're not improving it like 10% or 20%, right?

There's no incremental improvement.

The next model they roll out

has like a 10, 20 X increase in capabilities

on what that model is released for.

And so right now, even with this agentikai,

I mean, think about the application used

for the AI being able to mimic your screen

and do actions on your behalf, right?

So think about like inputting information in the CRM.

Sales reps will never be good

at inputting information in the CRM.

Our sales rep, we think we have a great sales team

we're obviously subjective,

like there's still not that great

in putting information in sales force, right?

But imagine AI being able

to grab all the information and plug it in

exactly where it needs to be in real time, right?

Now that's a very, very simple use case,

but like you look down 12 months,

the capabilities that AI will enable you

is probably right now unfathomable.

And so it's, there's gonna have a lot of huge applications

in terms of real time usage and bookings,

cleanliness of the CRM, usage of the DMS,

usage of service scheduling,

vehicle acquisition that I think is gonna change

the way that dealerships do business

even more so than it has in the last two years.

Okay, so two follow-up questions here

on customers and information.

Do you see AI enhancing CDPs or replacing CDPs

or something else?

For anyone who's not familiar, right?

We have people with all different experiences listening.

When I say CDP, I'm just referring to a platform

that houses customer data, your customer data

that is not your CRM and cleans that data, right?

Makes it better so you can send more targeted messaging

to your customer.

So that Sally who has three kids,

that your system knows that as opposed to

the system thinking that Sally's name is Jim

and that she has no kids

and she may need a convertible for the summer.

That's like the basics of what I'm saying.

Does AI kill that?

Does it enhance that or something else?

I think right now it doesn't kill it, right?

Cause like, I mean, AI could go in,

fetch a lot of information very quickly

from a lot of disparate systems, right?

Which is essentially what the CDP does.

I think what the CDP does is you're right,

is it just collects all the information on you,

the consumer, so that you can make

better, faster decisions.

I think right now what AI will do

and the way that we think about AI,

so think about the other CDP platform,

but AI is really gonna be a communication platform,

not just communication,

the way you think of chat GPT,

it can do all the synthesizing,

it can understand data,

so it can grab all that data that's in the CDP

and make a lot better decisions.

And then the last mile being the communication,

it can actually communicate a lot more effectively.

And then really like the last kind of like link in the chain

is the execution of the action.

So meaning you have your CDP,

which is all the data on a consumer,

which allows you to be a lot smarter.

And then you have AI that's grabbing that data,

figuring out, you know,

predictively figure out what to do next with that data,

communicating with you, you'll see the consumer

and then say if it's service scheduling,

going in and booking your service

at the dealership completely frictionless.

So right now I think it's additive, right?

So I think about the layer of CDP

and then you bolt on this massive AI application

that can use all that data to actually drive an outcome.

But who knows, maybe over time,

the next two, three, four, five years,

you know, this AI engine essentially becomes the brain,

essentially becomes the CDP.

Okay, so that brings me to my next question now,

which is today, right, September 2025,

what dealers are CDPs for every dealer?

This is another conversation I had,

like some dealers, you know, text, I'm like,

hey, what do you think?

Like, hey, I don't have a CDP.

Another one is like, I love it.

I use my CDP, it's like my, you know, top of mind.

It's the best thing since sliced bread.

Is a CDP for every dealer in September 2025?

Or if not, who is it for?

What's your take on that?

No, yeah, I mean, look,

you need to be of enough size for it to make sense.

It's not a light investment.

It's not easy.

And when they implement a CDP,

it takes like a year to do it right.

Maybe you could do a little bit faster,

but generally it's gonna take a year

and it's super costly.

So if you're a single point store,

even if you have two, three, four, five stores,

like is it worth it to invest in this CDP?

Like, you know, if you got to spend,

let's say a million bucks, like that's just cash,

like how many cars do you have to sell

to recoup that million dollars?

That's a lot of cars sold, right?

On the triple net basis.

And so I don't, yeah,

I think you need to be of enough size.

I mean, if you said, you know, art,

like what's the, what's the dealership group size that,

what's the acceptable threshold?

I don't know.

It's also based on the total of customer database

that you have.

But you know, I would say if you're a 15 plus store group,

that's probably worth looking into.

But you could still do a lot,

you could still do a lot better with all the data

that already exists in your CRM and in your DMS

without investing heavily into a CDP, right?

You can extract a lot more efficiencies,

a lot better outcomes with all the data

in those two systems to make, to move the needle.

And just to clarify,

when you say a million bucks, 15 plus store,

you're referring to a custom CDP

or an off the shelf CDP, you're referring to custom only.

Yeah, custom.

I mean, there's still expensive

if it's off the shelf, right?

I mean, you look at like the monthly recurring fees,

the set up fees, then you factor in the amount

of mind share and human capital

that you need to throw at it to make it work.

And that's if then you make it.

Oh, okay, got it.

So you're looking at it kind of fully loaded,

I understand.

Yeah, yeah, yeah.

I mean, it's generally expensive.

Now it works if you have a ton of records, right?

Cause that means like the more records you have,

you know, the last call for record it ends up being,

that's the unit economics workout.

But look, like if you think about even with a CRM

and a DMS, say you just had a CRM,

the dealership doesn't use all the data

that's at their fingertips in the CRM, right?

So the same issues exist if you have a CDP

and you don't use it correctly

or you don't have the staff or the infrastructure,

which is you have more data that's more organized,

but like you're still not using it

to the degree that you should.

So there's a lot of groups that are setting aside a CDP

and using AI in lieu of a CDP, right?

Cause it's faster, it's cleaner

and you don't need to make that capital investment.

So that's a very good insight.

And I'm sure some of the people I spoke with

will listen to this

and they will, you know, that'll be helpful for them.

So before we wrap up, voice AI

feels like it's the next frontier in the AI game.

And don't get me wrong, it's already very, you know,

well deployed in many dealerships,

but nonetheless, it's grown quickly

from my anecdotal feel.

Yeah.

How do you feel about voice AI?

Where are we at?

I think that's gonna be a super important part

of how dealerships think about deploying AI

and like a use case for an AI application.

I think right now, for at least for the first nine months

like we don't feel right now

like these large language models, right?

Cause everybody uses the same thing, right?

So think about open AI, think about Anthropic,

think about Bard, et cetera.

Like all those large language models

are powering almost everything.

The problem with voice right now

is we feel it's some dealerships might get burned right now

is the way that they're thinking about voice.

Like true voice, right?

You call like when people say voice AI,

they're thinking, I'm calling

and then I'm gonna talk to an AI

and they're gonna like help me out, right?

And it's gonna be a great experience.

And I always tell dealerships is like,

yes, I think that's gonna come.

I don't think it's ready now

cause that's really difficult.

There's two things.

One is, for instance, if you're talking to an AI you'll

see it's really hard for the AI to understand

when you're done speaking, right?

That small like imperceptible second that happens

where the AI is processing,

because when they ingest voice,

they're taking in voice,

they're transcribing it in the text,

they then retranscribe it in the voice

and then push it back out.

There's latency to that.

But that is, like even if it's 0.3 seconds,

the consumer picks it up that it's a bot, right?

And then they're circumvented, it's not a great experience.

Or oftentimes the AI because of the integration,

the ecosystem that don't exist,

can't answer your question.

So I call you for service and I go,

hey, is my car ready for service?

Or is my car ready?

I dropped it off two days ago.

It's still gotta hand you off to a person, right?

So it's like, again, another layer

that doesn't get me to the outcome I want,

I still have to speak to the person,

I still have to be on hold.

But then in addition, I have to talk to an AI agent.

So the way that we're thinking about it right now,

a much better way is using that transcription.

So like for instance, let a phone call come in,

let AI transcribe it, ideally move it to text, right?

And I'm speaking fixed right now,

but ideally the phone call comes in,

try to push that conversation to text.

So you call in, bring your car in.

It says, hey, Josie, do you want to just text you

and just carry this conversation over text?

That's how Delta does it.

You call Delta and they say, you know,

press one or say yes and we'll text you.

And they're moving like 38% of calls into text.

And with text you have a lot more control.

Yeah, and it solves this whole problem

because with text you can control it.

You can book them, you don't have this thing

that exists right now with latency.

And then you also have this secondary problem

of like ambient noise.

So you're calling, you're in a mall,

you're at the airport, like there's noise around,

it picks up the noise and it says,

I'm sorry, I couldn't pick up what you were saying.

Like that's super frustrating.

Like I don't care how good the AI is,

that's gonna be an issue that you're running to.

So you move them the text and then you can book them

via text, you can communicate via text for instance,

but two things happen.

One, they don't have to call you back

because they've created a habit

that they can get business done via text.

And two, your staff doesn't have to call the consumer back

because you have a texting dialogue going.

Like for instance, would you rather call a restaurant

to book a reservation or text them

if they had a texting service to book it?

You would never call them in your life.

Makes sense.

So anyway, that's a great insight.

All right, so as we're wrapping up here,

for the dealers listening,

whether it be the 30 store group, the two store dealer,

they're thinking about AI,

they're thinking about their group,

they want to grow sales,

they want to be more efficient,

they're thinking about the next year,

just like some of this up for us,

give us your couple pieces of advice,

what should they be doing?

What is the right thing in your opinion

that they should be looking into deploying,

thinking about in again, September 2025?

What's your take?

Yeah, I think there's three easy things

is they always look at the consumer experience.

So for instance, go to your website,

does your chat provider give you

a great customer experience, right?

Like go in and ask some of the most simple questions

in the chat that you have,

which is generally the first experience

the customer has at the dealership, this chat, right?

Unless they call you or walk in

and see if you're delivering great customer experience.

On the variable side, like think about,

do you have a BDC at internet team?

Like, do they have the close rates that you want?

Most will say no.

Has it changed in the last two to three years

that all the changes you've implemented,

probably negligibly?

So it's like, I would think as a dealer,

are there AI applications that can help me

solve some of the stuff on the variable

that hasn't changed, right?

Response time, followup, customer experience,

appointment booking, you know,

have the leads come at night,

so are consumers being neglected for 12, 13 hours?

And then really I think the most,

like again, I said at the beginning of this podcast,

but I think the biggest opportunity is service.

And I think if dealerships are looking,

am I, most do not want to spend any more money

than they're currently spending.

I think they're trying to tighten their belts.

So it's, can you do more with less, right?

And where is the money?

The money's not in the CRM, the money's in the DMS.

And if you look at it, it's like,

how are you going to get more people to service from you?

Do the exact same thing you've been doing the last two years.

Why didn't you say that?

The money's not in the CRM, it's in the DMS.

Why did you say that?

Yeah, I mean, if you, again,

like if you look at somebody at sales that gets,

if you look at somebody that gets 350 leads a month, okay,

they probably have 20,000 records in the DMS, right?

So that is 20,000 people

on which the average RO is 460 bucks,

in which by the way, 55% of a dealership's DMS

are people who haven't been in in 12 months plus.

And if you think about service,

I'm going to simplify some groups and I say,

that's not the truth.

Like if you think about service,

the reason they went elsewhere is not because of price.

You go to the average consumer

and you're asking what the price of an oil changes

that Jiffy Lu versus the dealership,

they would not be able to tell you.

The reason they go somewhere else is out of convenience.

There's 190,000 independent repair shops in the US.

There's like 36,000 franchise dealerships.

So like they're already at disadvantage.

So you've got to provide a better experience

and AI can do just that for fix.

It can simplify the outreach, the communication

and more importantly,

it can simplify the service scheduling

so that you could be a lot more convenient

than Jiffy Lu-Biz.

Art, Dessen and Pal.

All right, this was incredible.

And by the way, the last thing you just said,

I was looking to the side

because I was writing it down.

I love that.

That's a very good quote.

I might have to use that one.

Money's not a CRM, it's a DMS.

And it makes total sense when you think about it.

So you're so focused on the new,

but what about the old, which is just as new?

We just, you know, it's just not as shiny

but it's there and it's valid.

If it sucks here to sell a new thing

then it is to retain something.

So it's, you know, it's human behavior,

but you know, AI doesn't have human tendencies.

Art, Dessen and Pal, Art, thanks so much for joining.

This was amazing.

Thank you, thank you so much for having me.

Appreciate it.

All right, hope you enjoyed that episode.

Please give the podcast a rating,

consider subscribing to the show

and check the show notes for links

to what we talked about.

Thanks for tuning in.

I'll see you guys next time.

About this episode

Art Dessin, Vice President of Sales at Impel, discusses the evolving role of AI in automotive dealerships, emphasizing the importance of effective communication and customer experience. He highlights that many dealerships are overwhelmed with messages, leading to a dilution of effective outreach. Dessin argues that the real opportunity lies in leveraging the data within dealership management systems (DMS) to enhance service operations and retain customers. The conversation also touches on the challenges of integrating AI into existing processes and the need for a more unified approach to customer interactions.

Ford Explorer Jeep Cherokee

Original notes

Today I’m joined by Art Dessein, VP of Sales at Impel. We dig into how AI is reshaping inventory management, the playbook big dealer groups use to roll it out without losing culture, and why less customer communication is actually leading to more sales—plus plenty more along the way. This episode is brought to you by: 1. Widewail - Want real insights instead of guesswork? With Voice of the Customer Analytics, Widewail uses AI to turn raw customer feedback into clear operational insights so you know exactly what to fix and where to focus. Go to @ https://www.widewail.com/cdg to book a demo. 2. Experian Automotive - Like most Car Dealership Guy Listeners, you’re constantly looking for the inside edge on the auto industry. So if you’re ready to step up your game to the next level – outpacing the competition and building customer loyalty – there’s only one place to go from here: Experian Automotive. They’re the only ones with exclusive data across vehicles, consumers, and credit—plus expert data scientists who connect the dots to uncover the insights you need. Get the industry-leading insights from Experian Automotive today! Learn more by visiting @ https://carguymedia.com/4cfcLjZ 3. Impel - Impel AI helps you drive loyalty with proactive, personalized service outreach that scales seamlessly. From service milestones to recall alerts and tailored incentives, Impel makes service outreach feel more like a concierge service than marketing. With AI-powered service scheduling via text and email—your team can stop managing calendars and focus on delivering exceptional customer experiences. Future-proof your service drive today. Visit @ https://carguymedia.com/40nWpRj Check out Car Dealership Guy’s stuff: For dealers: Industry job board ➤ ⁠⁠⁠⁠http://jobs.dealershipguy.com⁠⁠⁠⁠ Dealership recruiting ➤ ⁠⁠⁠⁠http://www.cdgrecruiting.com⁠⁠⁠⁠ Fix your dealership’s social media ➤ ⁠⁠⁠⁠http://www.trynomad.co⁠⁠⁠⁠ Request to be a podcast guest ➤ ⁠⁠⁠⁠http://www.cdgguest.com⁠⁠⁠⁠ For industry vendors: Advertise with Car Dealership Guy ➤ ⁠⁠⁠⁠http://www.cdgpartner.com⁠⁠⁠⁠ Industry job board ➤ ⁠⁠⁠⁠http://jobs.dealershipguy.com⁠⁠⁠⁠ Request to be a podcast guest ➤ ⁠⁠⁠⁠http://www.cdgguest.com⁠⁠⁠⁠ Topics: 00:34 How did missing a flight help? 02:53 How is AI used in dealerships? 04:37 Why renewed AI interest now? 10:54 AI's biggest service opportunity 16:32 Why customer experience drives sales? 22:21 How AI reduces message overload? 27:50 Can AI transform dealership operations? 36:39 Final AI advice for dealers Car Dealership Guy Socials: X ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠x.com/GuyDealership⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Instagram ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠instagram.com/cardealershipguy/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ TikTok ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠tiktok.com/@guydealership⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ LinkedIn ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠linkedin.com/company/cardealershipguy⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Threads ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠threads.net/@cardealershipguy⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Facebook ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠facebook.com/profile.php?id=100077402857683⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Everything else ➤ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠dealershipguy.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠

Car Dealership Guy Podcast

Car Dealership Guy

View classic page →

0:00

43:31