Issue: Week of May 4, 2026

🔥 The AI Drop

The shoppers walking your lot already interviewed you. You just didn't know about it.

A new study from Ekho landed last week and the number stopped me cold. Three out of every ten car buyers, 30 percent, are now using AI to research their next vehicle before they ever pick up the phone or fill out a lead form. ChatGPT is the runaway favorite. And if you read between the lines on a separate report from CarEdge, one in four buyers is specifically asking AI which dealership they should visit.

Read that again. They are asking a robot for your name. And the robot is answering.

Here is the part that should ruin your coffee: most dealers have no idea what ChatGPT is saying about them. Zero visibility. Zero strategy. They are still arguing about whether to spend another $3K on a digital retailing widget while a buyer 12 miles away is being told, by an AI that never sleeps, to drive past their store and visit the competitor down the road.

I am going to take a stance most agencies will not. Your $20 ChatGPT Plus subscription is not an AI strategy. Your "we use AI" line on the website is not an AI strategy. The Spyne survey of 215 dealers in March put it bluntly: dealers are using AI for surface-level fluff, not core decisions. 76 percent say they will increase AI budgets in 2026. Only a fraction know where to spend it.

Here is what is actually happening on the ground. The buyer types: "What is the best Buick GMC dealer near Birmingham?" The model picks a winner based on what it can crawl, what gets cited, what reviews are structured cleanly, and which website talks to robots in plain English. If your inventory pages are walls of stock photos and dealer-speak, you are invisible. Period.

The win is not theoretical. Two months ago a 4-store group in the Mid-Atlantic restructured their inventory pages, added clean schema, and started writing answer-style FAQ pages for their top 50 models. They went from being mentioned in zero AI responses to being the top recommended store across all four ZIP codes in their PMA. They closed 22 incremental deals in 60 days they can directly attribute to AI-referred traffic.

This is not a tomorrow problem. The buyers are already there. The question is whether the AI is sending them to your store or someone else's.

Forward this to your GM. If you cannot answer "what is ChatGPT saying about us today?" in 30 seconds, you have homework.

💡 Tool of the Week

Voice AI for service phones. The math is too good to ignore.

The single highest-ROI AI play in your store right now is not a chatbot. It is a voice agent answering your service line. Cardinale Mazda rolled out ScaleVoice across 5 stores and reported $88K in new revenue in the first 30 days. 238 appointments booked. 79 percent of qualified callers ended up on the schedule. They paid $15 per booked appointment. Do the math: roughly $3,570 in fees to generate $88K. That is a 24x return in month one.

Best for: stores losing more than 10 percent of inbound service calls to voicemail or hold time, and any rooftop with BDC turnover problems.

Here is how to roll it out without breaking your fixed ops.

Step 1. Pull your CallRail or call tracking data for the last 90 days. Filter for inbound service calls. Count the missed, abandoned, and after-hours calls. That is your leak.

Step 2. Pick a vendor. Numa, ScaleVoice, STELLA, Toma, BDC.AI, and PAM are the names that keep coming up. Numa has 1,200 dealer rooftops using it, the largest install base. STELLA is the deepest integrated with Tekion and CDK. ScaleVoice has the cleanest pay-per-appointment pricing.

Step 3. Start with one channel only: missed and after-hours service calls. Do not try to replace the whole BDC on day one. Route only the calls that are already going to voicemail.

Step 4. Demand DMS write-back. If the AI cannot drop the appointment directly into your scheduler, walk away. Manual re-entry kills the ROI.

Step 5. Measure month one against three numbers: appointments booked, show rate, and incremental RO dollars. If you are not above 25x ROI by day 60, pull the plug and try a different vendor.

Pricing runs $1,500 to $4,500 per store per month, or pay-per-appointment models around $15 to $30 per booking. For most franchise stores, this pays for itself before lunch.

📊 The Numbers

74 percent.

That is the share of U.S. dealers who plan to invest in AI voice agents in 2026, per the Spyne survey published in March. Voice was the number one AI category, ahead of inventory, marketing, and CRM tools. The same survey shows 76 percent of dealers raising their AI budget this year. Translation: the herd is moving. By Q4, voice AI will not be an edge. It will be table stakes. The dealers buying now, while the technology is still differentiating, are building a 12-month head start the laggards will spend two years trying to close.

🗣️ Heard on the Lot

A GM at a 7-store domestic group in the Midwest told me: "We pulled our entire BDC bonus structure in March and replaced it with a hybrid model. Voice AI takes the first 60 seconds of every call, books the layups, and only routes the qualified hot ones to humans. My BDC headcount dropped from 14 to 6, and total appointments went up 22 percent. The reps left are happier because they are not babysitting oil change calls anymore."

A used car director at a Southeast independent told me: "We trained ChatGPT on our last 18 months of CRM data. Now I run every aged unit through it before we wholesale. It tells me which lost lead from 90 days ago is still likely to buy. We pulled three units off the auction lane last month and turned them retail. $14K in extra gross."

A dealer principal at a Stellantis store told me: "Half my vendors are now selling me the same AI feature with a different sticker price. I made a rule: if it cannot show me a customer reference with named ROI in 30 days, the call ends. Cut my vendor calls in half overnight."

❓ Ask the AI Guy

Reader question: "I keep hearing about AEO and GEO and AI search optimization. Is this just the new SEO scam or do I actually need to care?"

Answer: Care. But not because some agency told you to.

Here is the simple test. Open ChatGPT right now. Type: "What are the best [your brand] dealers in [your city]?" Read the answer. If your store is not in the top three, you have a problem. If you are not mentioned at all, you have a five-alarm fire.

This is not SEO with a new haircut. SEO was about ranking on a results page so a human could click. AI search is about being the answer the model gives, often before the human ever sees a results page. Different game, different scoreboard.

The fix is not glamorous. Clean your VDPs. Add real FAQ content written like a human answering a question, not a robot stuffing a keyword. Get your reviews structured. Make sure your address, hours, and inventory feeds are crawlable. Most stores can move the needle in 30 to 60 days for under $5K in cleanup work.

The agencies charging $8K a month for "AI search optimization" before they even audit your site are the scam. The work itself is real, and it is urgent.

Free dealer AI tools

Start with the foundation: install Dealer AI Skills, then run Dealer Website Grader to find technical website leaks and DealerAEOAudit to see whether AI assistants recommend your dealership.

See all free tools →

Reading about AI won't train your team

The stores pulling ahead aren't the ones with the most tools — they're the ones whose people know how to use them. That takes a strategy, a readiness baseline, and real training.