Answered Isn't Booked: What 965,000 Dealership Calls Say About the 2026 Appointment Gap

American dealerships have spent three years fixing the visible half of the phone problem. Answer rate is tracked. Hold time is tracked. After-hours coverage is staffed, outsourced, or automated. Most general managers can quote their store's answer percentage without opening a dashboard.

Two independent 2026 call-analytics datasets suggest the remaining leak is somewhere else entirely — not in whether a human picks up, but in whether anyone asks the caller to come in.

Myth vs. data

• The assumption: dealerships lose phone opportunities because calls go unanswered.
• CallRevu's midyear 2026 analysis found roughly 94% of inbound fixed-operations calls and about 86% of variable-operations calls were answered — yet only 15% of inbound variable-ops calls became an appointment.
• In a separate dataset, Calldrip's scoring of 965,351 dealership sales calls found that 55% of lead-response conversations ended without anyone asking for the appointment.

The timing matters. J.D. Power and GlobalData project August 2026 new-vehicle retail sales at 1,142,700 units, down 6.9% year over year on a selling-day-adjusted basis, with total retail consumer expenditure falling 7.6% to $49.8 billion — roughly $4.1 billion less than August 2025.

Both firms are explicit that the comparison is distorted. Last August was inflated by buyers pulling EV purchases forward ahead of the federal credit expiring, and Labor Day fell inside the August reporting month in 2025 but not in 2026. Thomas King, president of OEM solutions at J.D. Power, went as far as saying year-over-year volume comparisons are not especially useful for assessing demand this month, and described underlying demand as continued strong.

Read carefully, that caveat sharpens the point rather than softening it. Demand is holding. What is shrinking is the margin for error on each individual opportunity a store has already paid for.

The ring is fixed. The conversation is not.

CallRevu, a dealership communications-intelligence vendor, published a midyear 2026 review of dealership call performance, summarized by F&I and Showroom in July. Its own framing is the useful part: answering the phone is now an expectation rather than a differentiator.

The answer-rate row looks like a solved problem. The conversion row does not. A variable-operations department answering 86% of its calls and converting 15% of them into appointments is not suffering from a staffing or telephony failure. It is losing the opportunity somewhere between hello and goodbye.

CallRevu also reported that when a representative genuinely engages in a sales-opportunity conversation, roughly 40% of those convert to appointments. The distance between 15% and 40% is not a technology gap. It is a behavior gap.

What 965,351 calls say actually moves the number

The second dataset is more specific about which behaviors. Calldrip, which sells AI call scoring, scored 965,351 dealership sales calls across 53 rooftops spanning multiple OEM brands, regions, and call types, and identified 148,699 of them as real sales opportunities. An earlier, smaller phase — 1,215 inbound calls across 15 stores, run with Alan Ram's Proactive Training Solutions — pointed the same direction: trade-ins came up on only 14% of calls, the overall appointment set rate was 24%, and when the trade did come up, appointment rates rose to 42%.

At scale, two behaviors separated from the rest.

The skip rates are the story. Across lead-response calls in that dataset, 55% ended without an appointment ask, 75% never included a trade-in inquiry, and 54% skipped needs identification. These are not difficult behaviors. Asking whether someone has a vehicle to trade requires no product knowledge, no manager approval, and no additional software.

A note on attribution, because it matters for anyone chasing the primary source: this dataset is Calldrip's, published on its own site in April 2026. At least one trade summary of the same figures in August attributed them to a different vendor. The numbers cited here come from Calldrip's published study.

The hardest question to ask in 2026 is the one with the most lift

There is a reason the trade-in question gets skipped three times out of four, and it is not laziness.

J.D. Power's August 2026 data puts the average new-vehicle transaction price at $45,563, up 2.0% year over year, with average monthly finance payments at $812 — a record for the month of August, up 3.7%. Nearly three in ten trade-ins, 28.8%, carry negative equity, up 0.6 percentage points from a year ago. Loans of 84 months or longer now account for 13.9% of financing, up 2.1 points, and subprime penetration has risen 2.0 points to 10.8%.

In that environment, “what are you driving now?” is the question most likely to surface bad news. A salesperson who has watched three deals die at the appraisal learns to postpone it. The data says postponing it costs appointments — and the appointment is precisely where a store still has room to solve an equity problem with a different vehicle, a different term, or a different payment structure. Skipping the question does not make the negative equity disappear. It moves the discovery to a competitor's showroom. We looked at that dynamic in more detail in negative equity and dealership process.

Calldrip's own framing of the ask is worth borrowing: the appraisal has to happen in person, which turns an uncomfortable question into a concrete reason to visit.

What the arithmetic looks like on your floor

Take the published rates at face value and apply them to a round number. A store working 1,000 lead-response opportunities a month at the 27% baseline sets roughly 270 appointments. The same 1,000 opportunities at the 46% combined rate sets roughly 460 — about 190 more.

That arithmetic is illustrative, not a promise, and two cautions belong with it. First, appointments are not deliveries. What the additional appointments are worth depends entirely on your show rate and your closing rate on shows, and those are numbers you should supply from your own store rather than borrow from an industry average. Second, the largest individual-store gains in the study — one rooftop moving from a 13% to a 56% set rate — are best-case examples from a vendor's own dataset, not a median outcome. They are evidence that the ceiling is high, not a forecast.

What survives both cautions is the direction. The incremental appointment costs nothing in advertising, nothing in headcount, and nothing in inventory.

Where AI on the phone genuinely helps, and where it does not

The behavior gap is where AI voice agents have a defensible argument, and it is narrower than most vendor decks suggest.

The strongest case is measurement. A manager reviewing calls by hand typically listens to three to five out of fifty or more per day, under 10% of the sample. Automated scoring evaluates every call against the same criteria, which turns coaching from an anecdote into a specific number a representative can move. That holds whether the calls are handled by people or by software, and it is arguably the higher-value half of the category.

The second case is consistency. A script executed identically on the four-hundredth call of the week at 6:40 on a Saturday evening is exactly what humans are worst at and automation is best at. If 55% of lead-response calls skip the appointment ask, the failure is variance, and variance is a machine-solvable problem.

The third is the callback debt. CallRevu found that 22% of promised return calls in fixed operations and 14% in variable operations never happened. Outbound follow-through at volume is unglamorous work that degrades predictably whenever a store gets busy.

The limits deserve equal billing. An automated agent that asks badly is still asking badly; script quality, objection handling, and knowing when to stop pushing remain human design problems. Disclosure and consent obligations apply to automated outreach and vary by state, which is a question to settle before deployment rather than after — we covered the TCPA considerations and AI disclosure rules separately. And setting an appointment is not the same as earning a customer: the handoff to a person is where most of the perceived quality of the interaction is won or lost.

A 30-day audit that costs nothing

None of this requires a purchase to test.

1. Pull last month's lead-response calls and score a real sample — a few hundred, not five — against four criteria: proper introduction, needs identified, trade-in raised, and an explicit appointment request offering two specific times.
2. Publish the ask rate by representative rather than the team average. The average hides the distribution, and the distribution is the opportunity.
3. Separate ask rate, set rate, and show rate. A store tracking only the last one cannot distinguish a conversation problem from a confirmation problem.
4. Audit promised callbacks against completed callbacks for one week. The gap is usually wider than anyone expects.
5. Re-score after 30 days of coaching before evaluating any vendor. A baseline you measured yourself is the only honest benchmark for a vendor's claim.

If the ask rate is already above 80% and appointments still are not moving, the problem sits upstream in lead quality or downstream in the showroom, and no phone technology will fix it. If it looks like the 45% in Calldrip's data, the cheapest improvement available to the store this quarter is a sentence.

Carbuki builds AI voice agents for US dealerships, with call-level scoring on every conversation instead of a sample. If you want to see what your own ask rate looks like, carbuki.com is a reasonable place to start.

Sources

• JD Power and GlobalData U.S. Automotive Forecast, August 2026 — J.D. Power, Aug. 21, 2026
• The Conversation KPIs That Drive Appointment Sets — Calldrip, Apr. 28, 2026
• Auto Dealers Have Some Phone Homework — F&I and Showroom, July 29, 2026
• Analysis: Sales calls require more than sweet talk to make it onto a customer's calendar — Car Dealership Guy, Aug. 20, 2026

Calldrip and CallRevu are vendors in the dealership call-analytics category, and their studies draw on their own customer bases; their figures are attributed by name throughout. J.D. Power and GlobalData figures are forecasts published Aug. 21, 2026.