Best Practices for Routing Inbound Dealership Calls to Service

Numa's AI Receptionist reads intent the moment a call connects and routes a status request, a new appointment, and an angry customer down three different paths instead of one shared queue, a distinction most service departments never make. That single habit, routing every call in the order it arrived instead of by what the call actually is, is responsible for most of the hold time, transfers, and lost appointments dealerships blame on "being too busy." CDK's Service Shopper 5.0 study found 28% of customers struggled to navigate the phone menu and 26% were transferred at least once just to book an appointment. The fix isn't more phone lines or more BDC headcount. It's routing calls by intent instead of availability, so a status request never competes with a new customer for the same queue.

The Problem: Most Service Calls Are Routed by Luck, Not Design

Walk into almost any BDC or service drive and the routing logic is the same: whoever's line is free picks up the next call. There's no distinction between a customer asking "is my car ready?" and a customer trying to book their first appointment, and no distinction between a routine question and a customer who's already frustrated before the call connects. Everyone competes for the same queue, in the same order they called.

That flat structure is expensive. CDK's Service Shopper 5.0 study surveyed over 2,000 vehicle owners and found 58% of service appointments are still booked by phone, and 61% of those callers were existing customers. But the experience getting there was rough: 24% waited nine or more minutes on hold, 28% struggled to navigate the phone menu, and 26% were transferred at least once before reaching someone who could help. Separately, CDK's Fixed Operations research found average hold times climbed to 9.3 minutes in 2025, up from 8.2 minutes the year before.

The call volume side of the problem is just as stark, and it doesn't require a competitor's proprietary telemetry to prove. Dealership service departments routinely miss 300 to 500 calls a week, and roughly three out of four callers who reach voicemail never call back at all. Numa's own analysis of what that costs a typical store puts the number in concrete terms: at an average repair order value of $450, a store missing 400 calls a week is sitting on roughly $6.5 million in annual Fixed Ops revenue exposure. Foureyes' own contact-rate benchmarking across more than 1,150 dealerships found an average contact rate of 69%, meaning nearly a third of inbound opportunities never get reached in the first place. That gap isn't a staffing problem alone. It's a routing problem: too many calls that shouldn't need a live person are competing for the same line as calls that genuinely do.

What Most Service Departments Get Wrong About Call Routing

The instinct when hold times climb is to add headcount or add phone lines. Both treat routing as a capacity problem. It's usually a sorting problem instead.

Numa's own analysis of dealership call patterns found that between 2PM and 5PM, roughly 75% of inbound calls to a service department are existing customers asking for a status update, "is my car ready?" in one form or another. Every one of those calls is answerable from data the DMS already has. None of them require a live conversation. But because there's no separation between a status request and a new appointment call, they land in the same queue, tie up the same advisors, and push new business straight to voicemail during the busiest hours of the day.

This is also where the generalist-queue model breaks down on the customer experience side. J.D. Power's 2024 U.S. Customer Service Index Study found customers are four times more likely to prefer a status update by text (68%) than by phone call (16%), a clear signal that most of the call volume clogging service lines is volume customers didn't actually want to generate as a phone call in the first place. They called because nothing else told them what they needed to know.

Key takeaway: Most missed-call and hold-time problems at a dealership aren't a lack of capacity. They're the absence of any distinction between calls that need a person and calls that don't.

The Playbook: How to Route Inbound Service Calls Correctly

1. Separate status requests from everything else, first

Before building any routing logic, split status-update calls out of the main queue entirely. These calls have a structured, DMS-answerable question behind them, and they shouldn't consume a BDC seat or an advisor's attention. Route them to an automated lookup, voice or text, that pulls the answer directly from the repair order. This one change removes the largest single category of call volume from the line before routing logic for the remaining calls even matters.

2. Route by intent, not by who's free

Once status calls are pulled out, everything left should route based on what the caller needs, not which extension happens to be open. SQM Group's research on skill-based routing found agents matched to a specific call type by skill deliver 5% to 15% higher first-call resolution than a generalist queue where any available agent takes any call. At a dealership, that means a warranty question goes to someone who actually knows the warranty system, a parts question goes to parts, and a new-appointment call goes to whoever has live visibility into the schedule, not whoever answered first.

3. Set a hold-time ceiling and build overflow into the plan

Waiting isn't the problem by itself; waiting with no information is. General call center benchmarking research points to the widely used "80/20 rule" as a service-level target: 80% of calls answered within 20 seconds. Dealerships don't need to hit a generic call-center number exactly, but they do need a defined ceiling, and a real overflow path once it's crossed, whether that's a callback queue or an AI-handled backup, rather than a hold loop that eventually times out to voicemail.

4. Give after-hours and overflow calls a real path, not voicemail

Voicemail is not a routing destination. It's the absence of one. A missed or after-hours call still needs to land somewhere a human can act on it the next morning, with enough context to call back informed rather than starting cold. This matters most for overflow: a customer who couldn't get through during a hold spike faces the same structural gap as an after-hours caller, and should get the same treatment, not silence.

5. Route escalations to a person immediately, every time

An angry or upset customer is the one call type that should never enter a standard queue. Communication AI's job is judgment, not throughput in these situations: the routing decision itself needs to happen instantly, sending the call straight to a manager or a trained advisor rather than the next available seat. This is also where CSI risk concentrates: a delayed or mishandled escalation is far more likely to turn into a lost customer and a public review than a routine call ever is.

6. Keep every routed call on one customer record

Routing logic falls apart the moment the person who picks up doesn't have the context of what already happened. If a customer called about a status update this morning and calls back this afternoon about the same repair order, the second call shouldn't start from zero. Connecting call handling to the same DMS record that scheduling, status updates, and follow-up already run on is what makes routing decisions consistent instead of a fresh guess every time the phone rings.

Call Type to Routing Path: A Quick Reference

Common Failure Modes in Service Call Routing

The one-size queue. Every call, regardless of type, lands in the same line and gets answered in arrival order. Status requests, new business, and escalations all wait the same amount of time, which means the calls that matter most get no priority at all.

The transfer chain. A caller explains their issue, gets transferred, explains it again, and gets transferred again. CDK's research found 26% of service callers were transferred at least once just to book an appointment. Each transfer is another chance for the caller to hang up.

The silent menu. A phone tree with too many options, none of which match what the caller actually wants, forces customers to guess and often dead-ends them at voicemail. Twenty-eight percent of CDK's surveyed customers said the phone menu itself was the point of friction, before they even got into a queue.

The after-hours void. Calls that arrive outside business hours get no routing at all, just a message and a hang-up. Every one of those is a caller who had to decide, in that moment, whether to try again tomorrow or call somewhere else.

Key takeaway: Every one of these failure modes traces back to the same root cause: a queue with no way to tell one call apart from another before someone answers it.

Quick Wins for Your Service Department

Pull last week's call log and tag every call by type. Before changing any routing logic, know what's actually coming in. Most departments are surprised by how much of their volume is status requests rather than new business.

Turn on automated status handling for the 2–5PM window first. This is where the status-call flood concentrates. Removing it from the live queue during peak hours has the highest immediate impact on hold times for everyone else.

Build a same-day escalation path for upset callers. Define, in writing, exactly where an escalated call goes and who owns it, so the decision isn't being improvised in the moment by whoever picks up.

The Numa POV: Routing Is a Data Problem Before It's a Phone Problem

Most "call volume" problems at a dealership are actually routing problems wearing a staffing costume. Adding headcount to a queue that's still sorting every call the same way just means more people doing the same undifferentiated triage, slightly faster.

Numa's AI Receptionist identifies intent the moment a call comes in (a status request, a new appointment, a parts question, or a customer who's already frustrated) and handles each one differently instead of dropping all of them into the same line. Status requests get answered instantly from live DMS data. New appointment calls get routed to booking with real shop-load visibility. Escalations get flagged and routed to a person immediately through LiveCSI, before a bad experience turns into a bad review. Every one of those interactions writes back to the same customer record that the rest of the dealership's communication already runs on, so the next call, text, or visit starts with full context instead of a blank slate.

One GM Buick GMC dealership's own team described the shift in escalation volume directly: pages to a manager for upset service customers after 4pm used to happen several times a day. Since routing those calls straight to a person instead of a shared queue, there hasn't been one.

Numa perspective: Correct routing isn't about answering faster. It's about making sure the calls that need a human never have to compete with the ones that don't.

That's what correct routing looks like in practice: not more capacity, but less noise competing for the capacity a dealership already has.

Frequently Asked Questions

What's the biggest mistake dealerships make when routing service calls?

Treating every inbound call the same way. Most service departments route calls in the order they arrive rather than by what the caller actually needs, which means a simple status request competes for the same queue slot as a new appointment or an upset customer. Separating status requests out first, since they're answerable from DMS data alone, removes the largest source of unnecessary queue volume before any other routing logic matters.

How much of a service department's call volume is just status requests?

A significant share. Analysis of dealership call patterns shows that between 2PM and 5PM specifically, roughly 75% of inbound service calls are customers asking for an update on a vehicle already in the shop. That volume is answerable from existing repair order data and doesn't need to occupy a live phone line.

Should angry or escalated customers go through the same routing queue as everyone else?

No. Escalated calls should route directly to a manager or trained advisor immediately, bypassing the standard queue entirely. This is where CSI risk concentrates: a delayed or mishandled escalation is far more likely to become a lost customer and a negative review than a routine call.

What's a reasonable hold-time target for a service department?

There's no automotive-specific standard, but general call center benchmarking research points to the "80/20 rule" as a common industry target: 80% of calls answered within 20 seconds. The more important practice for dealerships is having a defined overflow path once a hold-time ceiling is crossed, rather than letting calls ring out to voicemail.

Does better call routing actually reduce hold times and transfers?

Yes. Research on skill-based routing shows agents matched to a call type by relevant skill resolve issues 5% to 15% more often on the first call than a generalist queue where any available agent takes the next call. Fewer misrouted calls means fewer transfers, and fewer transfers means shorter overall call handling time across the department.

See how Numa routes every inbound service call by intent, on one DMS-connected customer record. Talk to Numa.