Welcome to another edition of the Car Dealership Guy Industry Spotlight Podcast Recap newsletter—the key lessons from top operators, founders, and execs shaping the future of auto retail.
Today’s guests are Parker Boice, General Sales Manager at BMW Utah and Than Hancock, Chief Sales Officer at Podium.
Together, they discuss how to compete in a market dominated by mega-dealers, why customer buying behavior has completely changed, what AI can actually do in a dealership today, and much more.
Competing in the same market with massive dealer groups.
Firmage Group Auto shares a market with Asbury, Ken Garff, and other top-ten national consolidators. To hold its own, the group has centralized its BDC staff, systems, and processes across all rooftops rather than duplicating resources at each location.
"Something that's interesting about the way we operate in order to compete with some of the bigger players in the market, we share a lot of our systems and processes between the stores. So we have a central BDC team that services all three stores." — Boice
This approach deliberately creates internal competition between stores, with the philosophy that customers benefit when dealerships compete for business together, rather than operating as isolated franchises.
Outsourced call centers fail because they lack product expertise.
Dealerships attempting to solve lead response timing through overseas call centers achieve fast reply times, but destroy the customer experience by connecting buyers with agents who cannot answer basic product questions.
"It solved the problem with being able to respond to leads within five minutes, but those groups aren't able to provide any type of information or value. They're offsite. They aren't experts in the product. They aren't trained to do it." — Boice
Customers immediately recognize when they've reached an offshore agent with no authority or knowledge, causing complete loss of dealership credibility before the actual sales process begins.
Half of dealership leads now originate online before any human contact.
Customer buying patterns shifted permanently during COVID, with more leads now generated through internet sources rather than phone-ups or walk-ins at the store.
"About half of our leads are generated from internet leads, the other half is phone up and walk in. So only 25% of the time when a customer comes in is it their first touch." — Boice
Dealerships responding to this shift create walk-around videos and virtual test drives on nearly all inventory to provide information digitally before customers decide whether to visit in person.
Online-sourced customers convert higher than traditional walk-ins.
Leads qualifying themselves through digital research before arriving at the dealership represent fundamentally different sales opportunities than cold walk-in traffic from previous decades.
"When I started selling cars, I didn't get internet leads. I don't think they existed. Now you've got half of the leads coming in strictly through third-party websites or our own website...that customer is going to walk in the front door before they even talk to any one of our staff. That's a really nice sales funnel that creates great conversion." — Boice
This eliminates wasted salesperson time on unqualified prospects, allowing staff to focus exclusively on high-value activities with customers ready to transact.
Presented by:
1. Podium - Don’t miss another lead. With Podium’s AI BDC, dealerships are seeing an 80% increase in after-hours appointments by handling leads 24/7. Instantly respond to inquiries, book test drives, and let your team focus on what matters: closing deals. Learn how Podium can help you sell more cars at podium.com/car-dealership-guy.
Agentic AI completes end-to-end objectives versus surfacing information.
True agentic AI executes complete tasks like scheduling appointments with customer details on the calendar, distinguishing it from basic chatbots that simply retrieve database information.
"The biggest thing you look for is can it actually get you to an end goal? In the case of Podium Jerry, his goal is to actually schedule a test drive with the customer's information on the calendar...versus can I go in and surface an inventory link because you asked about a car? That's not agentic. That's just simply taking quick answers from a database of information." — Hancock
This capability difference matters because basic information retrieval requires human follow-up to convert leads, while agentic systems complete transactions independently without staff intervention.
Customers request AI agents by name like human salespeople.
BMW Utah launched their AI agent without informing all dealership staff, leading to confusion when customers began arriving for appointments asking for a salesperson nobody recognized.
"You kind of are doing performance reviews, right? Jerry is an agent, but there's training agents within the tool that when you provide feedback, they're going to go in and train Jerry on the responses. So it is very similar to training an employee and providing feedback." — Boice
This name recognition demonstrates the empathy and conversational quality that separates agentic AI from basic chatbots, with customers forming relationships strong enough to remember and request specific agents.
AI employees require the same management framework as human staff.
General sales managers now oversee both human personnel and AI agents, applying similar training, feedback, and performance review processes to both types of employees.
"You kind of are doing performance reviews, right? Jerry is an agent, but there's training agents within the tool that when you provide feedback, they're going to go in and train Jerry on the responses. So it is very similar to training an employee and providing feedback." — Boice
The critical difference is AI agents require training only once per issue before applying corrections across all future interactions, unlike humans who need repeated reinforcement of the same concepts.
Customizable AI accepts single-instance training across all future scenarios.
Traditional employee training requires repeated reinforcement of policies and procedures over time with managers addressing the same issues multiple times with different staff members.
"The difference with training Jerry versus another employee is you only have to tell him once. And he's going to take that feedback. He's going to apply it across the board and going to improve over time." — Boice
Managers define if-then policies once - like warranty eligibility rules or trade-in procedures - and the system applies those decisions consistently across thousands of customer interactions without deviation.
Voice AI optimization differs fundamentally from text-based systems.
Early voice AI prototypes attempted to simulate human behavior by including keyboard typing sounds during pauses to signal the agent was working on requests.
"Our first prototype, we wanted to sound like they were actually doing something, right? So the first one was like, 'Hi, Parker.' And you could hear clacking in the background, like keyboard. And it was just so loud. And then it was almost so aggressive." — Hancock
These audio cues created worse experiences than silence, requiring extensive testing to eliminate robocall characteristics that cause customers to immediately hang up.
Proactive AI could replace human-dependent follow-up workflows.
Salespeople consistently fail at long-term customer engagement, focusing instead on immediate opportunities with customers physically present at the dealership.
"Salespeople are bad at engaging with the customer and playing the long sales cycle...proactively getting in front of a customer and saying, 'Hey, you bought your car a year ago. I see that you're in for service next week.'" — Boice
AI analyzing customer purchase dates, service schedules, and lifecycle patterns can initiate engagement at optimal moments without requiring human prompting or CRM reminder systems.
