Most dealerships don't have a lead problem in fixed operations. They have a phone problem. An AI voice agent for car dealerships is the cheapest way to fix it, provided you scope it around the calls that actually carry revenue.
The business case usually gets pitched as “24/7 coverage” or “BDC efficiency”. Both are true and neither gets a dealer principal or fixed-ops director to sign. What gets them to sign is a number: how many repair orders walk away every month because nobody picked up.
This guide gives you that number on your own data, a call matrix for deciding what the AI resolves and what stays with your advisors, the guardrails we would insist on, and a vendor checklist. It is written for general managers, fixed-ops directors and BDC managers at franchised stores and dealer groups.
The dealership phone is where service revenue leaks
Three data sets frame the problem.
- Service is the steady profit engine. According to NADA's full-year 2025 data, U.S. new-vehicle dealerships wrote more than 276 million repair orders and sold $164.6 billion in service and parts. The average store did $9.7 million in service and parts sales across 16,252 repair orders, and averaged $494 in service and parts sales per customer repair order (NADA Data 2025).
- A large share of service calls never turns into an appointment. Call-tracking vendor Car Wars says it tracked roughly 53 million inbound service calls across its dealer network in 2025. About 19 million were missed opportunities; 53% of those ended in voicemail and 29% of callers hung up after waiting on hold. Only about 4 million service appointments were booked (Car Wars 2026 white paper, via CBT News, March 2026). Car Wars also reports that only 44% of missed-opportunity calls were recovered in 2025 (CBT News, February 2026). These are vendor-network figures, not an industry census, but they are measured on real calls.
- Customers are leaving for convenience, not price. Cox Automotive's 2025 Service Industry Study found dealers handle 12% fewer service visits than in 2018. Only 54% of owners of vehicles two years old or newer returned to their selling dealer for service in 2025, down from 72% in 2023. Nearly half of owners (45%) are dissatisfied with their dealership service experience, primarily over unexpected costs and poor communication, and Cox explicitly recommends after-hours scheduling as a fix (Cox Automotive, November 2025; survey of 1,974 vehicle owners, April–May 2025).
Put those together and the picture is uncomfortable. Service is the department that keeps the store profitable, owners are already drifting to independents, and the first touchpoint — the phone — often fails before a human is ever involved.
Cox adds one more number that turns this from a service issue into a sales issue: owners who service at the dealership are much more likely (74%) to buy their next vehicle from the same place. A missed service call is also a weaker trade-in conversation three years from now.
Every unanswered service call is a repair order you paid to generate and handed to the quick lube down the road.
What a missed service call is worth: a model you can run on your own numbers
Here is the calculation we would bring to a fixed-ops review. Every input below is illustrative; replace each one with your own call-tracking and DMS data.
Two honest caveats. First, this is sales, not gross profit: apply your own fixed-ops margin before you compare it to the cost of any tool. Second, an AI voice agent will not recover every one of those calls. Even if you only win back half, that is still roughly $52,000 a month of work in the illustrative store, and the math scales linearly across a dealer group.
There is also a labour line. Car Wars puts the average service scheduling call in 2025 at about four and a half minutes (CBT News, July 2026). If even half of the illustrative store's 2,500 calls are scheduling calls, that is roughly 95 hours a month of advisor or BDC time spent on the phone instead of in the service drive. We break down the second-order costs of an overloaded queue — repeat calls, abandonment and complaints — in the hidden cost of call center overflow.
The dealership call matrix: resolve, assist, or keep human
Do not automate by volume alone. Rank each call type by how often it arrives and how much damage a wrong answer does. This is the starting point we would use with a franchised store.
| Call type | AI role | Why |
|---|---|---|
| Service appointment booking and rescheduling | Resolve | The highest-value automatable call. Needs live slot availability, not a callback promise. |
| Repair status: “is my car ready?” | Resolve | A read-only lookup on the open repair order. Frees advisors from the most interruptive call of the day. |
| Hours, directions, loaner and shuttle policy | Resolve | Pure knowledge-base answers. Should never reach a human. |
| Open recall questions | Resolve, then book | Check the VIN against the manufacturer's recall lookup, explain that recall repairs are free, and book the visit. |
| Published menu pricing (oil change, tyre rotation, diagnostic fee) | Resolve within policy | Only prices you have approved and published. Anything else is an estimate and belongs to an advisor. |
| Sales: inventory availability, test-drive booking | Assist | Confirm the vehicle is in stock, capture the lead, book the appointment, route hot buyers to a salesperson live. |
| Trade-in and appraisal requests | Assist | Capture year, make, model, mileage and condition; book the appraisal. Never quote a value. |
| Repair approvals and additional work | Keep human | The advisor's conversation, and the one that drives the repair order total. |
| Financing, rates, credit applications | Keep human | Regulated F&I territory. The agent routes, it does not advise. |
| Complaints and warranty disputes | Keep human (AI collects) | Judgement and goodwill decisions. The AI's job is a fast, complete hand-off. |
“Keep human” does not mean “AI not involved”. On every call, the agent can answer on the first ring, identify the customer and vehicle, and pass a complete case to the right person. That alone shortens every human-handled call.
Don't forget the sales line
Service is where the volume is, but sales calls carry the bigger ticket. Car Wars reports an average sales “Connect score” of 62.39% across its network in 2025 (CBT News, February 2026). Whatever the exact definition in your store, a caller asking whether a specific vehicle is on the lot is a buyer, and that call should be answered instantly, qualified, and transferred live to a salesperson with the stock number on screen.
Five guardrails for dealership voice AI
- Live slots or nothing. The agent must write to your real scheduler, with real advisor and bay capacity. An AI that “takes a message for the service team to call back” recreates the voicemail problem with better manners.
- No improvised prices. Only approved, published prices. Estimates, goodwill and discounts stay with people. We explain why this has to be enforced in code rather than in the prompt in Confidently Wrong.
- No financing talk. Rates, approvals and payment estimates go to F&I, every time.
- A warm transfer, not a cold one. When a caller asks for a human, or the agent is unsure, the call transfers with the customer, VIN and request attached, so nobody repeats themselves.
- Outbound needs consent. In February 2024, the FCC ruled that AI-generated voices count as “artificial” voices under the Telephone Consumer Protection Act, so AI calls generally require the called party's prior express consent (FCC Declaratory Ruling 24-17). Recall and service-reminder campaigns deserve their own compliance review.
A 4–6 week rollout that leaves your phone system alone
Unicall deployments typically go live in 4 to 6 weeks without replacing the existing phone system. For a dealership, that sequence looks like this:
- Week 1 — call-flow audit. Pull a month of call recordings or tracking data and classify calls by type, outcome and time of day. After-hours and lunch-hour volume usually surprises people.
- Weeks 2–3 — integrations and design. Connect the scheduler, DMS repair-order status and CRM; load hours, policies, menu pricing and transfer rules.
- Weeks 4–5 — testing. Replay real call scenarios, including the awkward ones: angry customers, wrong department, a caller who switches from service to a trade-in question mid-call.
- Week 6 — go live on overflow and after-hours first. Then widen to first-ring answering once the transcripts look right.
For reference, across Unicall's live deployments to date the averages are about 60% of calls resolved fully autonomously, a 100% answer rate, and cost per ticket down 52% within 90 days. Most of those deployments are in e-commerce customer care, not dealerships, so treat them as a reference point rather than an automotive benchmark. If you are replacing a menu-driven phone tree on the way, read why your IVR is killing your CSAT first.
Vendor checklist: choosing an AI voice agent for car dealerships
Compare pricing models too. Per-seat pricing makes little sense for an agent that replaces hold queues rather than people; Unicall's managed service is billed on ticket volume and starts from $1,000 a month (see pricing). For side-by-side comparisons with other voice AI platforms, see our compare hub.
FAQ
Which dealership calls should an AI voice agent handle first?
Start in fixed operations: service appointment booking and rescheduling, repair status (is my car ready?), hours, directions, loaner and shuttle questions, and open-recall checks that end in a booked visit. These calls are frequent, rules-based and easy to verify against your scheduler and DMS, and they are the ones most often lost to voicemail and hold queues.
Will an AI voice agent replace my BDC or service advisors?
No, and it should not be scoped that way. The agent takes the repetitive first contact so advisors and BDC staff spend their time on repair approvals, upsell conversations, trade-ins and buyers who are ready to deal. Every call it cannot finish is handed over with the customer, the vehicle and the request already captured.
Can an AI voice agent quote repair prices or financing terms?
It should only state prices you have published and approved, such as menu pricing for an oil change or a diagnostic fee. Repair estimates belong to the advisor after inspection, and financing, rates and credit questions belong to your F&I team. The agent captures the request and routes it, rather than improvising a number.
How long does it take to put an AI voice agent on dealership phones?
With a done-for-you provider, a first production scope such as service booking, repair status and after-hours coverage typically goes live in 4 to 6 weeks, without replacing your phone system. Access to your scheduler, DMS and CRM drives most of that timeline.
Can the AI agent make outbound recall and service-reminder calls?
Only with the right consent. In February 2024 the FCC ruled that AI-generated voices count as artificial voices under the Telephone Consumer Protection Act, so such calls generally require the called party's prior express consent. Treat outbound campaigns as a separate compliance review from inbound call handling.
Unicall.ai builds and operates done-for-you AI voice agents that answer every call 24/7, take real actions in your systems, and hand over to your people, with full context, the moment they should.