The buy center fails in the handoffs, not the offers
Most stores that "start a buy center" actually start a job title. They name an acquisition manager, buy a data tool, and put a banner on the service drive. Ninety days later the numbers look like this: a handful of walk-in appraisals, a few auction-avoided units, and a lot of leads nobody worked past day two.
The offers weren't the problem. The handoffs were.
A dealership buy center is a workflow that crosses four departments — service, BDC, sales, and used car management — and every one of those crossings is a place where a car goes back out the driveway. If you want to buy 25 cars a month from the public, you need to design the crossings first and the offer math second.
Here's how to lay it out.
Start with the service lane, because it's the only place the car is already there
Service drive vehicle acquisition is the cheapest source you have. The customer is on site, the keys are in someone's hand, and you already know the mileage and the service history. And most stores waste it because the ask is either never made or made badly.
Two things have to be true for the lane to produce.
First, the trigger is a data pull, not a mood. Every morning, someone runs tomorrow's appointment list against your inventory needs and equity position. You're looking for: vehicle is a model you can retail, customer is in a positive equity or payoff-clear position, and mileage is under your retail ceiling. That produces a short list — usually 6 to 12 cars out of a 60-car day. Those are the only cars you ask about. Asking everybody trains advisors to stop asking anybody.
Second, the advisor's job is a handoff, not a pitch. Advisors are not going to negotiate. Don't ask them to. Give them one sentence and one action.
"While we've got it up, our used car manager wants to put a real number on it — not a trade number, an actual buy number. Takes about ten minutes and you're under no obligation. Want me to have him come out?"
That's it. Advisor sends a text or a page to the acquisition person. The acquisition person owns everything after that.
Then track two numbers per advisor per week: how many of their flagged cars got an ask, and how many asks produced an appraisal. If an advisor is at 2 of 9, that's a coaching conversation, not a policy failure.
Digital acquisition leads need a different clock than sales leads
A "sell us your car" lead is not a sales lead and shouldn't be routed like one. The person filling out that form is usually shopping three offers at once, and two of those offers are instant and algorithmic. If your first human contact lands at hour four, you're bidding on a car that's already been bought.
Set the standard at 10 minutes during open hours, and make the first contact a call — not a text confirmation, not an email with a range.
The call has one goal: get the vehicle in front of someone who can put a firm number on it.
"Thanks for sending that over. I can give you a ballpark right now, but I'd be guessing on condition and I'd rather not waste your time with a number that changes. If you can swing by tomorrow morning, I'll have a real check-ready offer in twenty minutes. Does 9 or 11 work better?"
Notice what that avoids: quoting a soft high number to get the appointment. That's the single fastest way to burn the acquisition channel, because when the number drops $1,800 at the store, the customer tells everyone and never comes back.
Also set the shutoff. If you've made four attempts across three days with no contact, the lead goes to a monthly value-update cadence, not the daily grind. Acquisition leads go cold in a very specific way — the car gets sold — and once it's sold, more calls just make you look desperate.
Appraisal ownership: one person, one number, one clock
The most common breakdown in a dealer appraisal workflow is that nobody actually owns the number. The used car manager is at auction. The desk gives a "should be around" figure. The rep repeats it. Then somebody real looks at the car and it's $1,200 different.
Write it down like this:
- One owner. A named person, plus a named backup, with hours posted. If both are unavailable, there's a rule for who fills in — not a group text into the void.
- A number in 20 minutes while the customer is on site. Not "we'll call you."
- The offer is written, dated, and has an expiration — typically 3 days or 250 miles. Hand the customer a printed copy.
- The number is entered in the CRM before the customer leaves the lot. No exceptions. If it's not written down, you can't audit it later.
Then add the piece almost nobody does: record the reason when you pass. "Frame damage." "Too much money on it." "Wrong segment for us." When you review a month of passes and 40% say "wrong segment," you've learned something about how you're generating leads.
Follow-up is where the buy center actually makes money
Roughly speaking, the customer who declines your offer on Tuesday is not saying no forever. They're saying your number lost to a number they saw online, or they're not ready to be without a car.
So build two distinct follow-up tracks.
Track A — declined offer. Day 3 call, day 10 call, then every 30 days with an updated number. The value proposition on each contact is new information:
"You turned us down three weeks ago and I don't blame you. The market moved on those — I'm at $900 more than I was. Still have it?"
That sentence works because it's true and specific. "Just checking in on your vehicle" does not work and never has.
Track B — no contact. Four attempts over three days, then monthly. Different message, shorter.
Set an owner per track. In most stores the acquisition person owns Track A and the BDC owns Track B, because Track A requires someone who can move the number.
Match to inventory need, or you're just buying cars
Here's the discipline that separates a used car acquisition process from a spending spree: the buy center should be working from a want list, not from whatever comes in.
Build it weekly. Three columns:
- Segments you sold last month and can't replace — with turn rate and average front gross
- Price bands you're empty in — the under-$15K hole is the usual one
- Segments you're overweight in — where you pay less or pass
Then post it. The acquisition person sees it, the advisors see it, the BDC sees it when they build outbound lists. When a service customer's 2019 crossover is on the want list, the appraisal comes in aggressive. When it's a segment you're 14 units deep in, the number reflects that, and everyone in the building knows why.
Review the want list against actual buys every Monday. If you bought 18 cars and 5 were on the list, either the list is wrong or the buying is undisciplined — and you need to know which.
The audit that keeps it honest
Once a month, pull ten acquisition conversations at random — service lane asks, inbound calls, follow-up calls — and check four things:
- Did the ask actually get made, in the words you agreed on?
- Was a firm number given, or did the rep float a range?
- Was the next step scheduled with a day and a time?
- Did the CRM entry match what was said on the call?
You will find gaps. Most stores discover their reps are quoting ranges on the phone, which is exactly the behavior that kills the show rate. That's a five-minute coaching fix — but only if you heard the call.
That's the part that's hard to do by memory. If you're reviewing conversations at any scale, having them scored against the same four checkpoints every time is what turns "I think our lane asks are getting better" into a number you can manage against. Tools like MoreSignal exist for that, but the discipline matters more than the tool: same checkpoints, same ten calls, same week every month.
What good looks like at 90 days
- Advisors flagging 6–12 cars a day and asking on 70%+ of them
- Digital leads contacted by a human inside 10 minutes
- Every appraisal owned, written, dated, and in the CRM
- Declined offers on a live 30-day re-offer cadence
- A weekly want list that visibly changes what you pay
None of that requires new software. It requires deciding who owns each handoff and then checking that the handoff happened.