Dealership lead response time: the benchmarks, the first 15 minutes, and where the handoffs fail

Half of the dealerships in the country now answer a web lead the way the audits define as perfect: fast, through email or text and by phone, with the customer's actual question answered. Five years ago it was a quarter. The average got there mostly by automating the easy inquiry, and the same audits show where the new losses are: the inquiry that needs a person, the handoff between systems that reports activity the customer never saw, and the shopper an AI assistant sent somewhere else before the form was filled in.

VulcanAX's work is search and AI visibility, which is the part of the lead that happens before the response clock starts. The figures on this page come from the independent audits that measure the clock itself, and the argument follows them.

Dealership lead response in current numbers

Two kinds of study measure this. Pied Piper submits real inquiries through dealership websites and scores what comes back over 24 hours, brand by brand and group by group, every year since 2012. The older research, from MIT and InsideSales.com and later Harvard Business Review, measured what response speed does to the odds of contact and qualification across industries. The table keeps the two apart, because one describes dealerships and the other describes the mechanism.

The clock starts at the form, and most of the industry now beats it

The 2026 Pied Piper study is the largest independent measurement of what a dealership does with a web lead. It submitted an inquiry about a vehicle confirmed to be in stock through 3,290 dealership websites during business hours, each with a unique name, email and phone number, and scored the response over the next 24 hours on more than 20 weighted measures. The industry average reached 71 out of 100, six points higher than the year before, and 51% of dealers delivered what the study calls a perfect response: the question answered, fast, by email or text and by phone. Five years earlier the perfect-response rate was half that.

The three behaviours that moved the average are specific. Dealers used texting to answer the customer's question 54% of the time, up from 38%. Three-quarters of customers received a phone call, up from two-thirds. And 62% of inquiries got both a written answer and a call, up from 49%. Of 33 brands, 26 improved, and Chevrolet's dealers gained 14 points in a year to reach 74, the largest move in the study. Infiniti's dealers led at 82, ahead of Cadillac, Honda, Acura, Subaru and Toyota.

The bottom did not move with the top. Fourteen percent of dealerships scored below 40, which the study defines as failing to respond personally to the website customer, and at Audi, Mazda, Alfa Romeo, Mitsubishi and Jaguar more than one inquiry in ten received no response of any kind. The same inquiry, on the same day, at an Infiniti, Acura, Toyota or Chevrolet store, went unanswered less than 3% of the time. The study's own arithmetic for what that gap is worth: dealers who move from under 40 to over 80 sell 50% more units from the same quantity of internet leads.

The first 15 minutes are where the best groups win, on three channels at once

The dealer-group study, published in May, shows what the top looks like when it is a process rather than a brand average. Pied Piper evaluated every dealership in 31 large groups, 2,414 inquiries, and the group average came in at 74, three points above the industry. Napleton Automotive Group scored 93, the highest ever recorded, and the way it got there is a checklist. Ninety-one percent of its web customers received an email or text answering their question and a phone call, all within 30 minutes; the next-best group managed 80% and the average was 50%. Eighty-seven percent received a phone call within 15 minutes. Eighty percent of responses offered an appointment for a specific date and time, against a group average of 33%.

Across the field the same three behaviours separate the top five groups from the bottom seven. Appointments offered for a specific date and time: 60% against 24%. Texting used to answer the question: 69% against 33%. Responses sent through multiple paths, email, text and phone: 87% against 44%. The study's explanation of the multichannel figure is the one a store should keep: customers ignore calls from unknown numbers and emails are filtered to junk, so a single-path response is exposed to a failure the dealership never sees. Sixty-two percent multichannel response, the industry figure, means nearly four in ten inquiries carry that risk.

The appointment is the behaviour with the least automation in it and the largest gap. The study's reason for tracking it is practical: a hard appointment for a date and time lowers no-show rates and raises close rates when the customer arrives. It is also the one line in the response that an AI reply rarely writes on its own, which is why the groups that offer it two and a half times as often as the bottom groups are the ones with a person in the loop.

The AI raised the average and moved the failures to the handoffs

Pied Piper credits much of the 2026 gain to AI-powered automation answering simple inquiries quickly, and it spends more of the study on what that automation hides than on what it fixed. The clearest figure is the split by inquiry type. Dealers answered a typical web inquiry within 24 hours 78% of the time, often by automated message. When the inquiry asked something that required a person to think, the rate fell to 51%. Evaluations that needed human help scored nine points lower on average, and those customers were twice as likely to receive no personal response at all.

The second failure has no customer-facing symptom. The dealer-group study describes breakdowns between systems, when the AI tool, the DMS, the website, the CRM and the email, text and phone platforms pass an inquiry between them and the data does not arrive intact. The response is technically triggered; the system logs activity; the customer's actual question is never answered. The study's phrase is that the CRM may indicate activity even when the customer experience failed, which is a precise description of why a store can read a report showing every lead answered and still be losing the ones that mattered. The systems post covers what each of those platforms holds and where the seams between them sit; this is the seam showing up in a sales number.

Pied Piper's vice president for metrics, Cameron O'Hagan, put the risk in one sentence in the study: teams assume the automation is working and that someone will step in if it fails, and too often no one does. The operational answer is not less automation. It is a named person who owns the escalation, a test lead submitted from outside the building every week, and a measurement that reads what the customer received rather than what the system sent.

The research behind the five-minute rule, and what it actually says

The rule is usually attributed to Harvard Business Review, and the number underneath it is older. In 2007 Dr. James Oldroyd, then at MIT, published the Lead Response Management Study with InsideSales.com, built on more than 100,000 call attempts across six companies over three years. Its two headline findings are the ones still quoted: the odds of contacting a lead when called within five minutes, against 30 minutes, drop 100 times, and the odds of qualifying it drop 21 times. The decay is steep at the start, with contact odds falling five times between five minutes and ten. A survey in the same study found that each tier of delay in first contact cut the share of leads qualified by 4.3% and close rates by nearly 2%.

The Harvard Business Review article from March 2011, by Oldroyd, Kristina McElheran and David Elkington, audited how companies actually responded and found most did not respond fast enough to matter: 37% within an hour, 24% after more than 24 hours, and an average first response of 42 hours among companies that responded within a month. The 2014 follow-up across 14,061 audited companies found 32% could not receive a web lead at all because there was no working site or web form to take it, and that representatives made an average of 1.3 call attempts before giving up.

None of that research is automotive, and that is the point of reading it next to the Pied Piper data. The mechanism is general: a lead is a person who is still looking, and every minute they keep looking is a minute another store can answer. Cox Automotive's 2025 study has buyers visiting 4.6 sites and spending nearly three hours visiting other dealerships and sellers. A dealership's response window is the gap between its reply and the next store's, and the 2026 data says the next store is faster than it was.

The lead before the lead: the shopper an assistant sent elsewhere

The most useful new measurement in the 2026 study is the one that happens before the form. Pied Piper added a ChatGPT step for a representative sample of dealerships: it prompted ChatGPT as a local shopper from the dealer's own zip code, asking about a specific vehicle the dealer was known to have in stock. The dealer was listed first 77% of the time. It appeared but not first 14% of the time. In 9% of cases the dealership was not mentioned at all, despite having the vehicle and being the shopper's closest store. ChatGPT confirmed the vehicle was in inventory 70% of the time, confirmed test-drive availability 36% of the time, and included a direct link to the vehicle's page 30% of the time.

That is the response-time problem no CRM can see, because the lead never arrived. One store in eleven was invisible to a shopper who wanted the exact car it had, and seven in ten answers gave the shopper no link to it. Pied Piper's own recommendation is that dealers present inventory, availability, pricing and next steps in clear, machine-readable website content that AI tools can interpret. That is the same inventory feed the lead generation post describes as the origin of every lead the store owns, read by a different machine.

The two halves multiply. A store that answers in five minutes converts more of the leads it receives, and a store the assistant names first receives more of them. A store that does the first and not the second is running an excellent process on a shrinking share of its own market, and will not see the shrinkage anywhere in its reporting, because the leads it lost were never counted as leads.

What a store should measure, and what the report will not tell it

The audits are useful because they measure from outside. A store can do the same thing. Submit a test lead from a phone the CRM has never seen, during business hours, about a vehicle in stock, and time what comes back on each channel: the written answer, the call, the appointment offer. Do it with a simple question and again with one that needs a person, because the 2026 data says the two rates differ by 27 points. Read the CRM's log of the same lead afterwards, and compare what the system says it did with what the phone received.

The persistence figures belong in the same test. The cross-industry average is 1.3 call attempts before giving up; the top dealer groups reach the customer through three paths on 87% of inquiries. On texting, the compliance question that hung over 2025 resolved before it took effect: the Eleventh Circuit vacated the FCC's one-to-one consent rule on January 24, 2025, so the consent language on the store's own lead form, under the TCPA's prior-express-consent standard, is what governs the text reply. A store should read that language once and make sure the reply the automation sends is one the form permitted.

Where VulcanAX fits

VulcanAX's own work is the half of the lead that happens before the form: whether the store is the named answer when a local shopper asks a search engine or an AI assistant about a vehicle, and whether the inventory, availability and next steps that answer depends on are readable to the assistant giving it. The baseline audit reports where the store stands on that question, including which competitor is named instead.

The response process inside the store, the CRM configuration, the automation and the people who step in when it needs them, is the store's to run, and a store that wants it handled alongside the search work can raise that through VulcanAX. The audits on this page are here because the two halves are measured separately and lost together.

FAQ