Welcome to another edition of the Car Dealership Guy Podcast Recap—a rundown of key lessons from top operators, founders, and execs shaping the future of auto retail.
Today’s guest is Matt Muilenburg, Chief Product Officer at Impel.
Muilenburg has spent thirty years in dealer technology, starting at the Cobalt Group in the early internet era, and has focused on how artificial intelligence affects dealership operations over the last several years.
The smallest dealers have the most to lose by waiting on AI.
The one- and two-point dealers are actually the most at risk, not the least.
"The smallest dealers are the ones who aren't investing and the argument I hear is ‘We can't afford it and the return isn't there for us.’ The reality is not only can you afford it, but a dealership of the smaller size, it's probably way more important for you to be using it now because it will make you more profitable and in the absence of it, your customers are going to erode out from underneath you and you're not going to see it happening."
His point: A rural dealer with a tight market and loyal customers is actually in a great position to use AI to expand their audience into surrounding towns, because those buyers are already driving 25 or 75 miles to buy a car anyway.
Your team is already using ChatGPT on your customer data, and you probably have no idea.
Matt made a point here that a lot of dealer principals are going to find uncomfortable.
When he walks into any dealership and talks to people in the BDC, service, or sales, they casually mention they used Claude or ChatGPT that morning on a customer list.
"I sit there and cringe because the exposure and legal risk that's being created for that dealer in that moment and in that conversation is huge. Dealers aren't aware that it's going on. So the right answer to your question is, ‘Hey, every dealer, pop your head up, go talk to your people and see what they're doing,’ because they're probably putting you at legal risk or at the very least exposing things that you wouldn't want exposed to the general world."
Getting licensed versions where you control your own data and building a clear usage policy for your team is how you get ahead of this before something goes wrong.
Fixed ops is where the first AI dollar should go.
Every time Impel brings on a new customer and runs campaigns against their existing DMS database, the volume of disengaged customers who come back in is striking.
"The fixed ops department is probably where dealers should be investing first. It's astonishing to me every time we bring a new customer on and launch campaigns to their existing customer database coming out of that DMS system, the volume of disengaged customers, people who haven't been in in the last 12 months, that re-engage."
Layer service lane car buying on top of that and you're doubling down, because every car coming into your service drive is an inventory acquisition opportunity.
The trade-in lead is completely misunderstood, and dealers are torching the relationship from the first message.
Someone goes to KBB to get a value on their car. Dealerships treat it like a hot buy signal and immediately start blowing up the customer's phone.
"The typical person that goes to get a value from a KBB or somewhere else is most likely trying to sell that car private party and will for three or four weeks before they're actually ready to sell that to a dealer. But how do dealers react? It goes to a dealer or two or three or four and suddenly that consumer's phone starts to explode with people saying, I want to buy your car. I want to buy your car. And we've created a terrible experience there."
The smarter move is to just ask which of the three categories they fall into: buying, selling private, or trading, and then have a completely different conversation based on the answer.
Presented by:
1. Digital Dealer - Join us September 22–23 in Detroit and register today at DigitalDealer.com. Use code CDG25OFF at checkout and bring your team for a generous discount. Visit @ here for more info.
2. Uber for Business - The true cost of moving your service customers is bigger than it looks. Visit @ here to get started.
3. Impel - Meet the AI Operating System built for a new era of automotive retailing. Visit @ here for more info.
Stop thinking of AI as software you turn on and treat it like an employee you hired.
Most dealers buy an AI tool, flip a switch, and expect it to go. Matt's pushback on that was this:
"When you buy an AI solution, the first question you should ask of your team is, who's managing this employee? And notice I called it an employee. You're not managing this software. You're managing an employee. The second question should be, what are the success metrics for this employee? What do we need to do to empower that employee to achieve those success metrics? And who's going to be responsible for that level of training?"
A dealer group that treats AI like a hired employee with a manager, success metrics, and a training cycle will get dramatically different results than one that treats it like a piece of software they're paying for and ignoring.
The BDC SDR role is probably the first to go.
Matt was pretty direct on this when asked which dealership positions disappear in the next 12 months.
"SDRs in a BDC, for example, I'm not sure that there's a lot of that role that exists going forward because that is sort of the first skill that the AI agents took on. How do I take a lead? How do I nurture that lead? That capability will get stronger across the AI."
He also flagged marketing managers at big dealer groups and warranty claim processing teams as areas where headcount gets rationalized, not eliminated completely, but fewer people doing the same work at a higher level.
Forty percent of AI-scheduled appointments were leads that came in overnight.
There's a temptation to think of overnight AI as a nice-to-have add-on.
"40% of the opportunity that our AI schedules into an appointment for a dealer came in overnight."
That said, dealers who only deploy AI for overnight coverage and don't invest in the full process change will eventually write it off as a failed experiment, which is exactly why Matt says overnight alone should be the floor, not the strategy.
Consolidating around one brand in a region is a smarter play than owning 20 brands across a market.
This came up because I mentioned Eddie Stivers and his strategy of stacking Ford stores in the same region. Matt co-signed it and added his own reasoning.
"Specializing is important. There are really great reasons to have a diversified portfolio of franchises that you own because there are cycles for different brands and different distribution networks. But some of the most successful operators I know are the ones who consolidate their efforts around a brand that becomes their trusted brand. They end up partnering at a deeper level with that manufacturer, or they have relationships in the building that pay off in different ways."
On top of the manufacturer relationship benefits, having multiple stores of the same brand in the same area means AI has more data to draw from, and the results get better the longer you run it.
AI makes the best salespeople in your store more effective, not redundant.
The version of AI Matt is building toward rewards the people who have been in the store 5, 10, 35 years. They get fed better leads, more prepared customers, and real-time coaching on who's walking in.
"I've got six kids, four daughters. I want my girls to be able to walk into a dealership and have a conversation where they feel empowered, where they have confidence, where they know what they're doing is making the right decisions, where they're in control of that process, but both sides are winning. And I think that's finally possible."
The failure in the current system, as he sees it, is that the face of the dealership turns over constantly and has misaligned economic incentives. When AI filters out the noise and routes the best opportunities to the most trusted people in the store, that changes.
The dealership of the future still has people in it, just different people doing different things.
Ten years from now, dealer operations will look completely different. But when Matt gets asked if the showroom goes away and everything converts to fixed ops hubs run by agents, his answer is more nuanced than you'd expect from someone building AI for a living.
"When I think about where the dealer's going to be 10 years from now, the operations will look entirely different, but they'll still be filled with people. The types of people that are in the dealership will be different. The skills they have will be different. There will be a moment where people choose to opt out because it's no longer the type of industry that they want to be part of. I think that's great."
The dealers who win in that world are the ones who use the next few years to figure out which roles AI handles better and which roles only a trusted, experienced human can fill, and then invest accordingly in both.
