Every vendor at NADA told you AI would change your store. Most of them were selling a chatbot with a new logo. This is the first issue of Rooftop Insider, and before we go department by department in the weeks ahead, you need the altitude view: where AI adoption in retail automotive actually is, where the money is actually moving, and why the gap between stores that get this right and stores that don't is no longer linear. It's compounding.

Strip out the vendor noise and the numbers tell a consistent story. Roughly nine out of ten dealers are either deploying AI today or have concrete plans to. Kerrigan Advisors' latest dealer survey puts it at 43% already running AI in operations with another 47% planning to, leaving only about one in ten stores with no AI on the roadmap. Reynolds and Reynolds' Q1 2026 State of AI in Automotive Retail found 57% of dealership personnel now using AI in some part of their job.

Here's the number that matters more than any of those: adoption is a mile wide and an inch deep. Cox Automotive data shows that while about half of dealers use AI in some form, most of it is concentrated in website chatbots. Go one layer deeper and the curve falls off a cliff. CRM-embedded AI sits around 35%. AI call scoring, 12%. Automated speed-to-lead, 8%. Predictive AI, 5%.

Read that again. Nearly every dealer says AI is critical. Only a small fraction have taken it past the chat widget. That gap between belief and execution is the single biggest competitive opening in retail automotive right now, and it will not stay open long.

Cox Automotive's sessions at NADA 2026 put a number on it: dealers who have fully adopted AI are about 50% more likely to report revenue growth, efficiency gains, and higher profitability than those who haven't. Early-adopter stores tracked by Spyne reported up to 30% more showroom appointments, BDC operating costs down as much as a third, and 12 to 15 hours a week in recovered operational time.

Those aren't one-time wins. A store that answers leads in under a minute, 24/7, doesn't just close more of this month's leads. It builds a database of cleaner customer data, which makes its predictive tools smarter, which makes next month's marketing spend more efficient, which funds the next layer of the stack. Meanwhile the store down the street is still letting 64% of its traffic arrive after hours to an unanswered chat window. The advantage stacks on itself. That's what compounding means, and it's why "we'll look at AI next year" is a more expensive sentence than most dealer principals realize.

Follow the budgets, not the booths. Dealer investment surveys heading into 2026 show the spend concentrating in pricing and analytics (around 62% of dealers), sales development across variable and fixed ops (54%), and CRM modernization (49%). Notice what that list is not: it's not conversational widgets. The industry's conversation matured fast. Two years ago dealers walked the NADA floor asking what a tool does. This year they asked what it measurably improves and how fast it pays back. AI moved from a department-level experiment to a board-level line item.

The other visible shift is consolidation. The winning stores aren't the ones writing the most vendor checks. They're running fewer vendors, tighter integrations, and harder accountability on every tool. Twelve point solutions that don't talk to each other is not an AI strategy. It's twelve monthly invoices.

At altitude, here's how the map looks across the rooftop. Sales and BDC is where AI landed first: lead response, appointment setting, after-hours coverage, and increasingly call scoring that grades every phone-up before the morning meeting. Fixed ops is where the quiet money is, with AI service scheduling, predictive maintenance outreach, and recall mining keeping bays full without adding headcount. The desk and F&I are next in line, with pricing analytics and deal structuring tools moving from spreadsheets to models. Marketing and inventory is where predictive AI earns its keep: knowing which unit to acquire, how to price it, and which customer in your DMS is about to defect before they ever shop the store across town.

Each of those is a future issue of this publication. The point today is that AI is no longer a front-of-house novelty. It touches every department that generates or protects margin.

The first wave of dealership AI was reactive: a lead comes in, the bot responds. A chat opens, the bot answers. That wave is essentially over as a differentiator, because a chatbot you have and your competitor also has is table stakes, not an edge.

The next wave is predictive, and it's where the real margin lives. Predictive tools identify the service customer drifting toward the independent shop before the first missed appointment. They flag the lease customer whose equity position just turned positive. They tell you which auction lane to be in next Tuesday. Adoption there is 5%. That's not a warning sign. For the dealers reading this, it's a map.

Three moves, in order. First, audit what you already own. Your CRM, your DMS, your website provider, and your phone system have almost certainly shipped AI features you're paying for and not using. Turn those on before you sign anything new. Second, pick the deepest gap, not the shiniest demo. If your store is like most, that's speed-to-lead and call intelligence, the two highest-ROI categories with the lowest adoption. Third, train for it. The stores pulling away treat AI literacy as a baseline competency for every role, not a specialist skill for the one person who "handles the tech."

The dealers who win the next five years won't be the ones who spent the most on AI. They'll be the ones who spent most strategically and started before it was comfortable. By 2027, the store without an AI operating strategy isn't the majority. It's the outlier.

That's the view from 30,000 feet. Starting next issue, we descend, one department at a time.

Rooftop Insider covers how dealership operators actually use technology and AI across every department of the store. If a colleague forwarded this to you, subscribe to get every issue.

Sources: Kerrigan Advisors Dealer Survey; Reynolds and Reynolds, State of AI in Automotive Retail Q1 2026; Cox Automotive NADA 2026 sessions and adoption data; Spyne Dealer AI Spending Survey; CDK Global AI in Automotive Insights; Seez after-hours traffic data.