AI customer service for car dealerships: answer every lead and service question, leave the deal to your people

What do car dealership customers contact you about?

Dealership conversations split into two halves: shoppers asking about inventory, test drives, trade-ins and financing, and owners asking about service appointments, recalls and repair status. Agent can finish the questions that are answered by your inventory feed, scheduler and repair orders; anything about money a person has to approve stays with your sales, finance or service staff.

The service half is bigger than most sales teams assume. NADA reports that franchised dealers wrote more than 276 million repair orders in 2025 (NADA Data 2025), and every one of them can generate a status call. The examples on this page are written for a franchised rooftop; independent used-car lots can drop the recall and warranty rows.

What does an AI procedure for an inventory lead look like?

The most common dealership conversation is a shopper on a vehicle detail page asking whether a specific car is still there. Below is an Example procedure written in plain language, the way a BDC manager would brief a new hire. Agent turns each step into an action against your inventory feed, sales calendar and CRM.

Example procedure: vehicle availability and test drive

1. Identify the vehicle from the page the shopper is on, or ask for the stock number, VIN or year, make and model. If more than one unit matches, show up to three with trim, color, mileage and advertised price.
2. Check status in the inventory feed. If the unit is on the lot and not marked sold or pending, confirm it is available and give the advertised price exactly as it appears on the website.
3. If the unit is sold, pending or in transit, say so and offer up to three similar units of the same model within 10% of the advertised price. Never promise to hold a car or take a deposit.
4. Offer a test drive. Book only inside showroom hours and at least 2 hours from now. Collect first name, last name and mobile number, and remind the shopper to bring a valid driver's license. Do not ask for the license number in chat.
5. If the shopper mentions a trade-in, record year, make, model, trim, mileage and any known damage, and explain that the used-car manager appraises the car in person. Never state or estimate a trade-in value.
6. If the shopper asks about financing, explain that the finance office reviews options with them and send the secure credit application link. Never quote an APR, monthly payment, term, down payment or approval chance, and never collect a Social Security number, income or date of birth in chat.
7. Write the lead to the CRM with the vehicle, appointment time, trade-in details and the full transcript, and assign it to the salesperson on the floor schedule for that slot.
8. Hand off to the sales manager when the shopper asks for a lower price, an out-the-door quote with fees, a business or fleet purchase, or delivery to their home; when sentiment turns negative; or when the message contains words like lawyer, lemon law or chargeback. Outside showroom hours, tell the shopper when the team opens and flag the lead for the first sales meeting.

Example conversation (10:42 pm, website chat)

Shopper: Is the gray 2024 RAV4 Hybrid XLE on your site still available? Stock 41872.

Agent: Yes, stock 41872 is on the lot and listed at the advertised price shown on the page. The showroom opens at 9 am. Would you like a test drive tomorrow? I have 10:00, 11:30 and 2:15 open.

Shopper: 11:30. Also what would my payment be with $3,000 down? And I have a 2017 Civic to trade.

Agent: You're booked for 11:30 tomorrow; please bring your driver's license. Payments depend on the lender's approval, so our finance manager will go through the options with you. If you want to save time, here is our secure credit application. For the Civic, what trim and mileage is it? Our used-car manager will appraise it while you drive the RAV4.

What should a car dealership AI chatbot never say?

A dealership AI agent should never quote rates or payments, promise a price or trade-in value, diagnose a car, or collect credit application data in a chat window. Set those as hard rules, then add escalation triggers for the conversations a manager needs to see.

• No credit terms. An advertised payment or down payment can oblige you to show the full APR and repayment terms under Regulation Z (12 CFR 1026.24). Agent doesn't quote numbers at all; your finance office does.
• No negotiation. Agent repeats the advertised price and the fee types you publish. Counteroffers, add-on disputes and out-the-door numbers go to the sales manager.
• No mechanical advice. Agent books the appointment and records the symptom. If a customer describes brakes, steering, smoke, a flashing check-engine light or a recall with a do-not-drive notice, it tells them not to drive the car and hands off to a service advisor at once.
• No legal or financial advice. Questions about lemon laws, contract cancellation, negative equity or whether a deal is a good one go to people.

Dealers that arrange financing or long-term leases are financial institutions under the FTC's Safeguards Rule, which covers information from people who apply for credit and expects you to oversee service providers that handle it (FTC Safeguards Rule FAQs for auto dealers). Dealers also fall under the FTC's Privacy Rule. In practice that means credit applications stay in your secure application tool, Agent's permissions are scoped to the systems each procedure needs, and you can revoke them. Aftersales offers a GDPR data processing agreement and hosts all customer data in the US; that supports your program but doesn't make the dealership compliant on its own.

Which channels do car buyers and service customers use?

Website chat and text messages carry most dealership conversations, with email arriving from lead portals and phone still dominant in the service lane.

• Website chat on vehicle detail pages catches shoppers while they are looking at a specific car, often in the evening.
• SMS is how service customers expect to hear that their car is ready. Text marketing needs prior express written consent, and opt-outs such as STOP must be honored, per the FCC's rules on robotexts. Keep Agent's texts to replies and service updates the customer asked for.
• Email brings leads from third-party listing sites and the manufacturer, usually in the ADF format your CRM reads.
• Social DMs and WhatsApp matter for dealers that list on social marketplaces or serve customers who prefer WhatsApp.
• Phone calls land in the same Inbox, with warm and cold transfers and call recording on Growth and Scale, so your service advisors see the call next to the customer's texts.

Which dealership systems should an AI agent connect to?

Connect the inventory feed, the CRM, the service scheduler and the dealer management system first; those four answer most questions. Each one connects through the Aftersales API or webhooks, so confirm the specific connectors for your vendors on a call.

• Your inventory feed or website platform: stock numbers, VINs, status, advertised price and photos.
• Your dealership CRM: Agent writes leads, test-drive appointments and trade-in details back so the sales team works one list. Dealer groups that run their CRM on Salesforce can use Aftersales' Salesforce connection.
• Your service scheduler: open slots by advisor, loaner and shuttle availability, and service menu prices you publish.
• Your dealer management system (DMS): repair order status, promised times and customer records for owners.
• Recall and warranty data: open recalls by VIN from the manufacturer's dealer system, or the public NHTSA recall lookup, plus parts availability.
• Online parts and accessories store: if it runs on Shopify, Agent can look up and verify those orders directly.

For a full overview of actions and permissions, see Aftersales Agent.

How much does AI customer service cost for a car dealership?

For a single-rooftop franchised dealer, expect a few hundred dollars a month in seats plus $0.99 for each conversation Agent resolves. Here is a worked example using list prices on the pricing page.

Assumptions: 4 BDC and service-BDC staff on Growth seats, sales and service managers on free Lite seats, 1,800 chat, text and email conversations a month, and Agent resolving 50% of them in the first months (the median after 90 days across Aftersales customers is 71%).

Copilot is optional at $29 per rep per month billed annually. Track the real figure in Insights, which reports cost per conversation across seats, resolutions and channels, and read how resolution rate is measured before comparing vendors.

How do you roll out an AI agent at a dealership in 30 days?

Start with after-hours sales chat and service booking, where nobody is answering today, then widen to daytime once the numbers hold.

1. Days 1-7: connect the inventory feed, CRM, service scheduler and DMS through the API or webhooks. Import your sales and service policies into Knowledge. Write procedures for availability, test drives, service booking and repair status.
2. Days 8-14: run Agent against last month's chats and texts and have the BDC manager and service manager review every answer. Tighten the finance, pricing and safety rules.
3. Days 15-21: go live on website chat and SMS from closing time to opening time. Have the first sales meeting of the day review overnight appointments.
4. Days 22-30: extend to business hours, add recall checks and repair status, and compare resolution rate, appointment show rate and CSAT with the human team's numbers.

Your human team's queues, templates and staffing are covered in helpdesk software for car dealerships.

Frequently asked questions