Why Predictive AI Is The Dealership's New Competitive Edge

Forbes

7 days ago

Zohar Bronfman is the cofounder and CEO of Pecan AI, a predictive analytics platform making advanced AI accessible to business teams.
A customer walks into your dealership. You sold them a car three years ago. You don't know if they'll buy again, but your competitor might.
The real risk? Missed revenue, bloated outreach and lost loyalty.
The auto industry has weathered disruption before, but this is different. Electrification, autonomous tech and direct-to-consumer platforms are only part of it. In 2024, the U.S. imposed a 100% tariff on Chinese EVs, triggering supply shifts and pricing uncertainty across the global market.
This is where AI earns its keep—not in hype but in prediction.
Beyond The Robots: What AI Actually Delivers
There's no shortage of AI headlines promising fully automated showrooms, talking avatars and virtual concierges. When it comes to dealership survival, however, the real opportunity is quieter and more useful.
Predictive AI answers hard questions. Who's likely to buy this quarter? What's the right offer for each segment? How will supply shifts affect local pricing?
These aren't futuristic dreams. They're tactical decisions happening every day, and they're better made with prediction.
The best part? Predictive models aren't just for data scientists anymore. With the right tools, front-line teams can plug in CRM exports and generate guidance on who to call, what to stock and when to act.
From Gut Instinct To Guided Action
Most dealerships are sitting on a mountain of sales records, service histories, financing data and lease trends. However, a February 2025 report published by Lotlinx (via DBusiness) found that less than 5% of dealerships use that data to actually predict behavior, and they're still operating on gut feel and static dashboards.
Meanwhile, OEMs like GM are embedding predictive intelligence into manufacturing and logistics. GM even appointed a chief AI officer to accelerate this shift across operations.
Prediction isn't optional anymore. It's table stakes.
Macro Trends That Demand Prediction
To stay competitive, dealerships must move from reacting to anticipating. Here's where prediction matters most:
According to a Strategy& report, global battery-electric vehicle (BEV) sales rose 42% year over year in Q1 2025, reaching 4 million units when combined with plug-in hybrid electric vehicles (PHEVs), which use both a battery and an engine. BEVs alone hit a record 16% of new registrations. China led with a 55% sales increase and 27% market share, while the U.S. saw 18% growth and 8% share. Inventory strategies must now update in real time.
According to EY's 2024 Mobility Consumer Index, one in four car buyers completes the entire purchase online. Meanwhile, 87% begin with digital research, and 61% still want to collect keys in person. Predicting which leads need human interaction—and which don't—will drive faster, leaner closes.
Chip shortages, tariffs and delays continue to disrupt parts availability. Dealerships using predictive models to forecast demand can avoid service delays, protect revenue and maintain loyalty.
From micromobility to luxury EVs, demand is shifting fast. Predictive insight helps spot what's gaining traction before the market does.
Each trend demands agility. Trade wars, supply disruption and new channels aren't going away. Those who predict and respond will win. Those who guess won't.
A Real-World Example: Predicting Repurchase At Scale
We worked with one of the largest automotive groups in Israel—the exclusive importer for Mercedes, Hyundai and Mitsubishi—when it suspected many customers returned for a second vehicle around the three-year mark but had too many leads and too few reps.
We helped it build a predictive model using dealership data to rank past buyers by likelihood to repurchase. It factored in timing, vehicle type, service history and demographics. Focusing on the top 5% to 10% led to a sixfold increase in conversion over random follow-ups.
The model also uncovered patterns reps hadn't noticed. For example, buyers over 50 were far more likely to return, certain service centers drove higher retention, and vehicle type and price shaped loyalty.
It changed how they operate. The pilot became a standing program. Monthly lead lists are now baked into the sales workflow.
Why This Matters More Now
Tariffs are just one signal. Supply chains are tightening, demand is fragmenting, margins are shrinking, and new players like Tesla and BYD are reshaping expectations.
In this climate, waiting for perfect data or a full transformation is a luxury. Prediction is about prioritization: what to stock, who to call and how to align with demand before it shifts. Manually sorting leads or guessing reorder volumes is playing defense, and defense doesn't scale.
What To Do Next
Prediction doesn't require a digital overhaul. Start by identifying one underperforming decision and replacing guesswork with visibility. Use the data you already have—recent purchases, lease expirations, service records and even call logs—to score who's likely to return or churn.
Focus on one high-impact area like retention, parts forecasting or reengagement and then test it. Compare your current outreach list with a model-driven top 5%. A 30-day A/B test will tell you everything you need to know.
Don't wait for perfect data. Modern AI can handle messy inputs better than expected. If your team isn't working from forward-looking lists, your data's being used to look back, not move forward.
The Road Ahead
In the next 12 months, dealers that lean into prediction should expect to gain a real edge—higher conversion, less waste and faster pivots.
Dealerships that see around corners won't just survive the next disruption. They'll quietly pull ahead.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?