Auto Dealership Software Gets AI Built Into Every Layer

AI Just Landed at Every Layer of the Dealership. Now Comes the Hard Part.

The announcement came on May 5 from DealerOn, a company that builds websites for automotive dealers: a new AI suite called Sidekick, covering four distinct functions a dealership website had previously left to staff. A chat tool that answers shopper questions in under three seconds. A natural language analytics layer for reading website data without running reports. A photo enhancement engine that automatically improves lighting, framing, and backgrounds on inventory images. A compliance tool that scans pages for OEM brand guideline violations before an audit finds them.

Four tools, one launch, one company, one day. What makes it worth examining is the context: Cars.com shipped AI-generated, VIN-level video ads the same week. myKaarma made news for an AI tool that grades technician service videos. BizzyCar released campaign-level reporting tied to actual repair orders. Team Velocity published a detailed guide to integrated call tracking — the final accounting on how much dealer data disappears the moment a lead picks up the phone.

Five companies. Five layers of the dealership operation. One week in early May. Taken together, they show where AI for car dealerships actually stands in 2026: past the experimentation phase and into the operational stack, but still short of a connected system.

These companies were not coordinating their product releases. What happened is that AI adoption moved far enough into automotive retail that developers building for dealers found the same problems simultaneously. The dealership has a website layer, an advertising layer, a service lane, a communications system, and a business intelligence stack — and all of them needed to catch up at roughly the same time.

The result is the current situation: more dealership software has AI built into it than at any point in the industry's history, and the majority of it does not talk to any other system in the building.

What Each Tool Is Solving

DealerOn's Sidekick suite is structured around the dealership website as a customer conversation point, not just an inventory catalog. The Quick Connect Chat handles the questions that usually went unanswered after hours or buried in a form submission: is this vehicle still available, how does the trade-in process work, what service recalls apply to a specific VIN? Responses under three seconds is the benchmark because anything slower produces the same result as no answer — the shopper bounces.

The Insights Sidekick addresses a different problem. Website analytics tools generate data that most dealership staff cannot easily act on because pulling useful reports requires a trained user. Accepting natural language questions lowers the access barrier for GMs and marketing directors who know what they want to know but not how to query for it.

Cars.com took the auto dealership digital marketing layer: the advertising inventory that runs alongside the website. Their AI video platform connects a dealer's inventory feed to automatically generated video ads targeted at in-market shoppers on social platforms — TikTok, Instagram, Facebook. The videos generate within minutes of a vehicle entering inventory and disappear when it sells. The system uses first-party behavioral data from Cars.com's marketplace to target buyers already researching similar vehicles.

Lisa Gosselin, Chief Commercial Officer at Cars.com, described the positioning plainly: "Cars.com is the only one in the industry combining real-time marketplace demand signals, premium media distribution, and AI-generated VIN-level video." The combination she's describing is hard to replicate independently — the demand-signal layer comes from operating a marketplace where buyers actively search, not just a software platform that generates video.

On the service side, myKaarma's Tech Video Grader addresses a specific fixed operations problem: technician video quality. Walk-around inspection videos are how service advisors show customers what needs work and build trust before the repair approval conversation. AI grading gives technicians feedback on what they're producing and whether it meets the standard. Ford named myKaarma a Preferred Partner specifically for this video capability — a signal that the OEM sees the inspection video as part of the ownership experience, not an optional nicety.

BizzyCar's updates landed at a similar layer — appointment attribution and mobile service zones — connecting the scheduling system to campaign-level reporting so fixed ops managers can tie service appointments to the marketing that drove them, not just count how many showed up.

The Call Data Gap

The most persistent, and least glamorous, AI problem in dealership retail is the phone call. Dealers receive high volumes of inbound calls from serious buyers — people confirming availability, asking about pricing, booking service, or asking questions they don't want to put in a form. These are decision-stage callers who have already done research.

Only about a quarter of those phone interactions end with a complete entry in the dealer's system of record.

That figure comes from David Boice, CEO of Team Velocity, and it measures a structural problem: call tracking systems that operate outside the CRM, DMS, and analytics platforms cannot reliably connect a caller's identity to their prior behavior. A shopper who browsed four vehicle listings, started a finance inquiry, and then called the dealership is, from the perspective of most stores, a cold call. The person who answers has no visibility into what they were looking at or what stage they're in.

Boice describes what the conversation looks like without integration: "When your team answers without visibility into the exact vehicle, page path, and offer that triggered the call, you force the customer to repeat work they already did."

The fix is not complicated in concept. Phone intelligence tools exist. The data can be embedded in CRM entries and connected to DMS records in real time. But making it work requires the call system to actually share data with those platforms, and most dealer setups still handle phone calls as a separate information silo.

The AI Search Discovery Layer

Below all of these tools — below the AI chat, the video ads, the service video grader, and the call system — is a more fundamental shift in how shoppers find dealerships in the first place. It is happening in the search layer, and most dealers have not updated their websites to account for it.

Sam Vukas, COO at Dealer eProcess, laid out the mechanics in a Digital Dealer column published the same week: AI search tools now answer intent-based questions rather than return lists of links. When someone asks which Toyota dealer near them has a Camry under $30,000, the answer depends on whether the dealership's website has structured data in a format an AI can interpret — schema markup that accurately labels vehicle specs, pricing, mileage, and features, combined with clean URLs that convey relevance before a page even loads.

Vukas describes schema markup as functioning like "a digital window sticker for shoppers, and the machines looking for recommendations." Without it, the dealership exists in the search ecosystem but cannot be recommended by the AI layer that increasingly sits between a shopper's question and the results they receive.

A dealer can invest in a sophisticated AI chat tool on their website and simultaneously be invisible to the AI-powered search tools that would bring shoppers to that website in the first place. It has to work at both ends — discovery and engagement.

Why These Tools Are Still Mostly Disconnected

The week's car dealership AI tools cover five distinct areas of the operation. The honest picture is that they mostly do not share data. A shopper's interaction with a chat widget does not automatically surface in the call tracking system when that same person calls. The appointment record in BizzyCar does not automatically flow into the attribution picture that Cars.com is building.

A unified car dealership workflow — one where a website interaction, an ad click, a phone call, and a service appointment all feed the same customer record — remains a design goal rather than a default. This is the working condition most dealers will enter when they evaluate these products. Each tool has its own integration requirements — typically a connection to the dealership management system, the CRM, or both — and the work of building those connections falls to the dealership, not the software vendor. The question a GM should ask before signing anything: what system does this connect to natively, and who owns the integration work?

One indicator that this is not a solved problem: dealers are increasingly turning to peer networks to share what is actually working, because the guidance coming from vendors is not giving them an integrated implementation picture. Automotive News columnist Mark Hollmer noted on May 5 that sharing AI implementation ideas with other dealers has become one of the more practical ways to test unproven configurations before committing budget to them.

For operations running tools that don't yet communicate with each other, workflow orchestration platforms can help close these gaps — connecting AI outputs from one system to triggers in another without requiring custom code. AgentPMT's workflow builder is built for this: non-engineers can chain actions across connected systems, with an audit trail logging every step. Whether that fits a specific operation depends on which gaps matter most, but the underlying need is real and not going away as dealers add more AI point solutions.

Where to Start

The week's launches can obscure the practical question for anyone evaluating their options: which gap do I fix first?

Call tracking has the clearest ROI case, because the data loss is immediate and measurable. Three-quarters of inbound calls from qualified buyers leave no usable record in the system. Closing that gap does not require a full platform overhaul — it requires a call intelligence tool that connects to the CRM and DMS the dealer already uses, and a workflow that ensures the connection happens in real time.

AI search readiness is a slower-moving issue but harder to reverse once competitors have solved it and you have not. Website infrastructure changes — schema markup, URL structure, page speed — take time to implement and index. Dealers who wait for this to become an obvious competitive gap will be waiting until it has already cost them.

The advertising and service lane tools are the category most dealers already have conversations about. The AI video and inspection grading products are improvements to workflows that already exist, and the adoption case is relatively intuitive. The harder work is ensuring those tools connect back to the same customer record that the chat and call data are building.

The dealership AI stack is no longer hypothetical. This week made clear it is a real, layered problem — and the integration work that follows the individual product decisions is where the actual operational lift lives.

---

Sources

• DealerOn Unveils Sidekick, an AI-Powered Suite to Elevate Dealership Website Management & Performance — PR Newswire
• Cars.com expands AI video ads to boost VIN-level targeting, in-market conversions — CBT News
• How AI-based technology helps technicians' videos make the grade, spur service revenue growth — Automotive News
• Why AI-Ready Websites Are the New Standard for Dealership Performance — Digital Dealer
• BizzyCar Unveils Service Engine Enhancements and New Appointment Attribution Reporting — Digital Dealer
• Integrated Call Tracking: The Dealer's Guide to Capturing More Revenue From Every Phone Lead — Digital Dealer
• Column: The case for dealership communities in an AI-driven market — Automotive News