The classic layer: inventory and trust
Two engines have always driven dealership digital: inventory visibility and local trust. Inventory-level pages, indexed, priced, with real photos and honest descriptions, capture the model-plus-town searches that signal purchase intent this week; feeds into the marketplaces extend the same inventory to where comparison shoppers already are. Local trust runs on reviews: velocity, recency, and responses, especially the responses to bad ones, which buyers read first. Service-department marketing remains the most underused profit lever, service customers defect to independents largely because nobody markets to them between visits.
What burns budget: generic brand spots and third-party lead packages resold to your competitors. The money is in owning your own demand capture.
The 2026 layer: the assistant rides shotgun
Buyers now negotiate with AI before they negotiate with you: what is a fair price for this trim, is this dealer reputable, what should I watch for in financing, which dealers near me actually honor online prices. ChatGPT, Gemini, Claude, and Perplexity answer from reviews, forums, pricing content, and local coverage, and they name dealerships, favorably or not. A store with thin or contradictory public facts gets described with hedges that kill visits; a store with consistent pricing reputation and well-handled reviews gets the benefit of the synthesis, the same verification layer described in where ChatGPT gets its information.
The dealership-specific inputs: publish straight answers to the questions buyers actually ask assistants, fees, financing process, trade-in logic, out-the-door pricing policy; keep every listing and directory consistent; and treat forum threads about your store as public negotiations, because assistants read them as your reputation.
A 90 day plan for a store or group
Days one to fifteen: baseline. Ask the assistants the reputation and pricing questions for your brands and towns, log what gets said, run the free GEO Grader, and audit listing consistency across every platform carrying your name. Days fifteen to fifty: publish the transparency layer, an honest fees page, financing explainer, trade-in guide, delivery policy, the pages assistants cite and buyers screenshot. Days fifty to ninety: review system tuned for specifics, salesperson and model named, respond publicly to the worst threads, and re-baseline. Ninety days is one buying cycle in this category; movement shows fast.
Measuring visits, not clicks
Dealership attribution should end at the desk, not the click: track share of assistant answers naming your store for reputation and pricing moments monthly via AI search tracking; branded search and direct arrivals to inventory and fees pages, the AI-referral signature per AI traffic analytics; and the up-front question every desk should log, how did you pick us, with an AI option. Cost per showroom visit and per sold unit by source is the scoreboard a dealer principal already believes in; these lines just make the invisible channel show up on it.
