Why Dealership Onboarding Breaks Between the HR Office and the Service Drive
A new hire does not experience the onboarding process you designed. They experience the version their direct manager has time to deliver that day. Close that distance and the first 90 days hold. Leave it unmeasured and the two drift apart, store by store, without anyone deciding that they should.
That was the thread running through our recent panel, From Hire to High Performer, where we sat down with Lucinda Underwood, president of Kickstart Consulting and a 20-year automotive leadership coach, and Felicia Barr, a service director overseeing four stores who started in this industry as a service greeter. Two different vantage points, one consistent answer.
Our State of the Automotive Workforce study, built from more than 800 dealerships across North America, is where the 90-day numbers in this piece come from. If you want to see how your stores compare, start there.
Why do so many dealership hires leave before day 90?
The groups that keep people past day 90 are not more careful than everyone else. They have made delivery visible: they can see which parts of onboarding actually happened at each store, their managers coach in five-minute increments instead of waiting for a formal session, follow-through turns agreement into something real, and they hear what an employee is thinking well before it arrives as two weeks' notice.
Across North America, more than 30% of dealership exits happen inside the first 90 days. Another large share goes before the one-year mark.
The takeaway
Those are not recruiting failures. Those are people a store successfully attracted, interviewed, reference-checked and hired, who then left before the investment paid back. The leak is downstream of hiring.
The practical split is that HR surfaces where the gaps are, supplies the tools and keeps the process from being rebuilt store by store. Store leadership executes it, and GMs hold their managers accountable for the delivery. Neither half works alone.
Felicia described what the first two weeks need to cover before anyone talks about performance: logins, DMS access, OEM registration, key tracks, where to park, who the office manager is, and who to ask when something goes sideways. Plus the verification nobody loves doing. A technician can arrive with years of experience and still need someone to confirm they lift a vehicle the way this store lifts a vehicle.
Felicia's answer is more direct: an HR rep cannot train a technician, because they are not the one doing the job. People learn from the people who do the work. Customers buy cars from people rather than from the internet, and new hires learn the same way.
Her framing: "You can't just throw it on the wall and hope it sticks."
Lucinda's answer to a manager who sees development as an HR function starts with a question: do you want a high-performing team? Because development has to live in the everyday culture of the store, and the GM owns that culture. A leader's job is to build a capable team. The day-to-day operations are how you do it, not the point of it.
Where does the gap between the plan and the floor actually open up?
Both, in different halves.
Lucinda named why the gap opens: in a fast-paced store under pressure, without real intention behind it, a new hire quickly becomes another pair of hands. They get shown the tasks. What gets missed is everything around the tasks. Do they know what success looks like here? Are expectations clear? Do they know where to go for help? Are they getting feedback, or are they spending their energy wondering how they are doing?
Whose job is development, HR's or the store's?
Nothing in a normal operating week surfaces that difference. The process is complete at the group level, every store reports it as done, and the part that varies most is the part nobody can see from the outside.
Most of what surfaces in exit conversations is fixable. Pay plan clarity is fixable. Not knowing a path exists is fixable.The problem is almost never the issue itself. It is that the information arrives at the exit interview instead of in week six, and nothing in a normal operating week asks the question sooner.
It helps to think about the process in three layers. There is the HR administration layer, which most groups have gotten good at. There is the "how am I successful in this specific role" layer, which is harder, because every store runs its own tools and processes and past experience may not transfer. And there is the "why is it worth being here" layer, which multi-rooftop groups are uniquely positioned to answer and rarely say out loud. The second and third are where delivery varies most, and they are the two that rarely get checked.
What those numbers cannot tell you is when someone's experience starts to shift. Felicia described a technician who gave notice after two years. In the exit conversation, it emerged he had never understood the store's pay structure. Her reaction says everything about where the real gap sits: she would have wanted that information long before the two weeks' notice, and by the time someone is browsing job postings, the store has already lost them.
Where does a busy manager find the time to develop people?
They are still worth watching early. In our State of the Automotive Workforce study, completion rates on early activities, including HR paperwork and initial training, were among the stronger leading indicators of whether a new hire would stay, or in some cases whether they would start at all. Early disengagement shows up as unfinished tasks before it shows up as a resignation.
The most common objection Lucinda hears is time, and she is quick to say it is usually legitimate. Managers do have a lot on their plate.
Training completion rates confirm participation and compliance. They do not tell you whether anything changed on the floor.
But the diagnosis underneath it is worth sitting with: a large part of the reason a manager has no time is that every decision routes back through them. They are involved in everything because nobody around them has been developed to the point of handling it.
What should you measure instead of training completion rates?
Her line on this was the one most people in the session wrote down. "Coaching might take longer today, but rescuing will take forever."
That distinction matters more than it sounds. "I need the numbers up" is not a task. If a manager knew how to get the numbers up, the numbers would be up. Is it row count? Appointment volume? The work is turning the outcome into something measurable enough to coach against.
The smallest possible starting move is nota program. It is noticing the reflex to jump in and fix something you have handled a thousand times, and asking a question instead. What do you think we should do? That is coaching. Over time you stop getting problems and start getting problems with a proposed solution, and eventually you stop getting the problem at all.
Felicia's version is blunter. Her managers get a 15 to 20 minute call every morning at nine. When the answer is "I did not have time," she goes to the store and does it with them, which tends to surface the real answer, which is often that they did not know how.
Felicia runs the same play on the drive.She sits at the counter with an advisor and watches. When they bring her a question, she asks what they think they should do. Afterward: "Are you open to some feedback?" Four words that change the temperature of the conversation before it starts.
Everyone nods in the meeting. Lucinda's point is that when something then does not get done, genuine resistance is rarely the cause. More often it is one of four things: expectations were not clear, support or feedback was missing, nobody took ownership, or there was nofollow-through. Without follow-up, a process becomes optional.
Development is also not only correction.Handing a tech back their own warranty paperwork and asking what looks wrong inhere is training. So is telling an advisor you noticed how they handled a difficult customer.
No, and treating it as such is how a rollout quietly dies.
Is agreement the same as accountability?
No, and treating it as such is how arollout quietly dies.
Everyone nods in the meeting. Lucinda's point is that when something then does not get done, genuine resistance is rarely the cause. More often it is one of four things: expectations were not clear, support or feedback was missing, nobody took ownership, or there was nofollow-through. Without follow-up, a process becomes optional.
Felicia's version is blunter. Her managers get a 15 to 20 minute call every morning at nine. When the answer is "I did not have time," she goes to the store and does it with them, which tends to surface the real answer, which is often that they did not know how.
That distinction matters more than it sounds. "I need the numbers up" is not a task. If a manager knew how to get the numbers up, the numbers would be up. Is it row count? Appointment volume? The work is turning the outcome into something measurable enough to coach against.
What should you measure instead of training completion rates?
Training completion rates confirm participation and compliance. They do not tell you whether anything changed on the floor.
They are still worth watching early. In ourState of the Automotive Workforce study, completion rates on early activities, including HR paperwork and initial training, were among the stronger leading indicators of whether a new hire would stay, or in some cases whether they would start at all. Early disengagement shows up as unfinished tasks before it shows up as a resignation.
What those numbers cannot tell you is when someone's experience starts to shift. Felicia described a technician who gave notice after two years. In the exit conversation, it emerged he had never understood the store's pay structure. Her reaction says everything about where the real gap sits: she would have wanted that information long before the two weeks' notice, and by the time someone is browsing job postings, the store has already lost them.
Most of what surfaces in exit conversations is fixable. Pay plan clarity is fixable. Not knowing a path exists is fixable.The problem is almost never the issue itself. It is that the information arrives at the exit interview instead of in week six, and nothing in a normal operating week asks the question sooner.
Whose job is development, HR's or the store's?
Both, in different halves.
Lucinda's answer to a manager who sees development as an HR function starts with a question: do you want a high-performing team? Because development has to live in the everyday culture of the store, and the GM owns that culture. A leader's job is to build a capable team. The day-to-day operations are how you do it, not the point of it.
Felicia's answer is more direct: an HR rep cannot train a technician, because they are not the one doing the job. People learn from the people who do the work. Customers buy cars from people rather than from the internet, and new hires learn the same way.
The practical split is that HR surfaces where the gaps are, supplies the tools and keeps the process from being rebuilt store by store. Store leadership executes it, and GMs hold their managers accountable for the delivery. Neither half works alone.
The takeaway
The groups that keep people past day 90 are not more careful than everyone else. They have made delivery visible: they can see which parts of onboarding actually happened at each store, their managers coach in five-minute increments instead of waiting for a formal session, follow-through turns agreement into something real, and they hear what an employee is thinking well before it arrives as two weeks' notice.
Our State of the Automotive Workforce study, built from more than 800 dealerships across North America, is where the 90-day numbers in this piece come from. If you want to see how your stores compare, start there.
Frequently asked questions
What percentage of dealership employees leave in the first 90 days?
In HR4's State of the Automotive Workforce study, built from more than 800 dealerships across North America, 27.6% of allemployee exits happened before day 90 and 58% happened within the first year. Early exits are concentrated in the first three months rather than spreadevenly across the first year.
How long should dealership onboarding last?
Most dealership onboarding plans cover the first one to two weeks, which handles system access, introductions and storeprocesses. The first 90 days are where retention is decided, so the plan should extend through that window with scheduled check-ins rather than ending once thepaperwork is complete.
What should a dealership onboarding checklist include?
Three layers. Administration: payroll setup, policy acknowledgements, DMS and OEM system logins, key access. Role readiness: store-specific processes, tool training and verification that theemployee can perform core tasks to your standard. Belonging: introductions across departments, a named person to ask questions, and a clear picture of what success and advancement look like.
How can a dealership find out why an employee is unhappy before they quit?
Build scheduled check-in points into the first 90 days and beyond, so the question gets asked at a set moment rather than when someone notices a problem. Track early signals such as incomplete onboarding tasks or missed training. Exit interviews surface the sameinformation months after it could have been acted on.
How do you get dealership managers to coach when they say they have no time?
Start smaller than a formal session. Ask a question before offering a solution, give feedback in five-minute increments onthe floor, and open with permission, such as "Are you open to somefeedback?" Managers are often short on time precisely because every decision routes back through them, which coaching reduces over time.
