Presented by:
Hey everyone,
Reminder!
We're hiring an Assistant Editor at Car Dealership Guy News.
This person will be on the front lines of news, analysis, and storytelling that sets the tone for an entire industry—reaching millions every month.
Key details:
Must have 2+ years of editorial experience in the automotive retail industry
Needs strong knowledge of dealer economics, franchise operations, and market fundamentals.
$65k - $75k / year, remote
Healthcare reimbursement plan
Phantom equity program
Think you’d be a good fit, or know someone that would? Share the full job listing with them.
— CDG
First time reading the CDG Newsletter?
Welcome to The Breakdown, an analysis of auto retail’s top trends, moves, and insights—in under 5 minutes.
Some of the sharpest, scrappiest, most useful AI tools in automotive retailing aren’t coming from big vendors.
They’re being bolted together by dealership operators in real time.
From rethinking new car orders, to surfacing missed sales in GA4 to placing used cars across rooftops, these are five real ChatGPT use cases dealers are building behind the scenes…
Master prompts: Engineering new car inventory for precise ordering.
Andy Wright, Managing Partner at Vinart Dealerships built a master prompt that analyzes his sales history against local market supply to determine optimal new car orders.
How it works: His prompt feeds ChatGPT three data sources…
His dealership's sales history (anonymized)
Manufacturer ordering system exports
And current marketplace inventory within a certain radius
The prompt also explains his ordering objectives, then asks specific questions like "What configurations am I selling fastest?" and "What specs are oversupplied locally that I should avoid?"
Andy's still refining the prompts, admitting he's "throwing stuff against the wall" and hoping to connect with dealers using more sophisticated AI ordering systems. And while he hasn’t placed an actual order yet with this method, he's building the framework to automate what's typically a time-consuming manual process for sales managers.
The goal: prevent ordering mistakes that create lot rot while ensuring popular configurations stay in stock, freeing up management time from clerical tasks so they can focus on customers instead of spreadsheets.
Automation: Spinning window sticker data into customized vehicle descriptions.
Michael Wood, general manager at Jaguar Land Rover Virginia Beach is using ChatGPT to create targeted vehicle descriptions by analyzing window stickers and writing for specific buyer demographics.
Michael Wood
How it works: The workflow, built by digital director Ryan Montville, starts with downloading the digital window sticker from vAuto using iPacket.
Chat then reads vehicle specs and features from the sticker
Next, it identifies the target buyer for that make/model/trim
And then, writes descriptions that connect emotionally with that demographic
The goal: Differentiate his VDPs with memorable descriptions that customers will actually read, while removing time-consuming administrative tasks from frontline staff at the same time.
Data synthesis: Analyzing CRM and GA4 data to determine missed opportunities and areas of risk.
Josh Detzel, regional marketing director at Hudson Automotive Group, uses ChatGPT to solve a fundamental problem…
General managers get bombarded with reports that lack actionable insights.
So, he decided to upload certain data from Google Analytics (GA4) and the dealership’s CRM to try and pinpoint any areas for improvement.
How it works: The data sets Josh uses cover five main metrics, including website users, engaged time per session, VDP views (new/used/CPO), internet leads, and closing percentages.
ChatGPT then analyzes the data against NCM & Associates benchmarks and identifies where there are gaps in execution. After that, the system generates executive summaries in email format with recommendations that GMs can actually act on.
The goal: Replace useless data dumps with specific action plans.
“As a marketer, I can drive all the traffic in the world and all the leads and all the opportunities in the world, but if they're not executing at the store level, it doesn't really matter,” he told me.
Josh Detzel
A quick word from our partner
Tired of broker markups and long delays? Auto Hauler Exchange offers a better way.
Auto Hauler Exchange connects shippers directly with carriers.
No brokers, no hidden fees, just an easy to use, transparent marketplace that saves you time and money.
Shippers using Auto Hauler Exchange save an average of 25% compared to brokered loads. With average delivery in just 4 days, vehicles hit the lot faster.
Sign up for your free shipper account at autohaulerexchange.com and start shipping smarter.
Bespoke workflows: Strategically placing used vehicles across a network of dealerships.
Adam Gaedke, the former COO for Van Horn Auto Group, built a tailored GPT workflow that tells Van Horn exactly where to place each used car across their 16 dealerships for the fastest turns and highest profits.
How it works: Adam feeds three data sources to the GPT…
Five years of actual sales data from the DMS
Dealership performance history by make/model/trim
Seasonal patterns and current inventory levels across locations
From there, when a buyer acquires cars or a store gets trades, they input the specs and get the top three placement recommendations.
Boom.
The goal: Eliminate gut-feel inventory decisions and replace opinion-based decisions with data that maximizes velocity and gross profit across the entire dealer network.
Real-time queries: Delivering instant dealership insights to management.
Jarrod Kilway, VP of Digital Operations at Casa Auto Group, decided to go one step further and build a custom GPT that acts as a real-time query system for general managers.
Jarrod Kilway
How it works: Jarrod’s custom GPT connects to live data feeds in CRM systems, Google Analytics, Facebook advertising, and other channels. From there, it can handle basic performance questions like website conversion rates or lead volumes, as well as complex strategic questions that combine historical and current metrics.
The goal: Instant access to dealership performance data without waiting for reports.
But building a custom GPT is more than prompts and files. It’s about creating an experience that feels natural, useful, and aligned with the dealership’s goals.
That means:
Providing clear step-by-step instructions
Refining prompts so users can actually achieve their goals
And giving the model an identity that makes its responses consistent with the intended outcome
While none of these tools are perfect quite yet…
They're still impressive tactical solutions built by operators who understand their specific pain points and have the technical curiosity to experiment.
Whether it’s ordering, recon, placement, or marketing execution, they’re taking core decisions that used to live in spreadsheets (or sales manager brains), and feeding them into machines that work 24/7, don’t get distracted, and scale without complaint.
And once auto retailers taste that leverage, there’s no going back.
What do you think? Which one of these hacks are you most likely to try?
Missed Wednesday’s episode of Daily Dealer Live?
Presented by:
Jeff Miller on selling one-price Subarus, DriveCentric CEO on CRM innovation
Featured guests:
Jeff Miller, CEO of Mark Miller Subaru
Matt Leone, CEO of DriveCentric
This week on the
"There's no technician shortage!"—The mastermind behind Bozard's service lane machine
Shout out to Cox Automotive and DLRdmv for making this episode possible!
Stream now on:
The time bomb in labor rate increases (and how dealers are maximizing profit)
Shout out to Wooden Automotive Consultants for making this episode possible!
Stream now on:
