The report says the process is fine. The floor says otherwise.

Every dealership I've seen has a report that shows a healthy process. Ups logged. Test drives at 60%. Write-ups at 40%. Manager T.O. rate at 80%. Follow-up tasks completed at 92%.

Then you stand on the floor for four hours and watch something else entirely. A customer gets handed to a manager who never actually walks over. A rep does a "test drive" that's a lap around the lot with the customer in the passenger seat. A follow-up task gets marked complete because the rep called, got voicemail, and hung up without leaving one.

Neither the report nor your gut is the truth. The gap between them is the truth, and the only way to find it is to observe the same deals the report is describing.

That's the audit. Pick a window. Watch what happens. Compare it to what got recorded. Write down the delta.

Set up the audit so it can't be gamed

Two rules make a sales process audit worth doing.

Rule one: observe first, pull the report second. If you look at the CRM before you walk the floor, you'll see what you expect to see. Watch the customer, take notes, then go find what the system says about that same customer. The order matters.

Rule two: audit deals, not averages. Take 10 to 15 specific customers over a week. Name them by stock number or last name. Follow each one from arrival to whatever happened next. Averages hide the failures; individual deals don't.

Block four hours across two days. Saturday morning and a weekday evening will show you two different dealerships.

Bring one page with four columns: manager involvement, customer handoffs, process adherence, follow-up ownership. Those are the four places dealership floor management actually breaks, and each one fails in a way you can see with your eyes.

Column one: manager involvement

Your report probably tracks T.O. rate. It says whether a manager was requested, not whether a manager showed up in a way the customer noticed.

Watch for the difference:

  • Rep walks to the desk, says something, comes back with a number. Report logs a T.O. Customer never met a manager.
  • Manager walks over, shakes hands, asks two questions, leaves. Ninety seconds. Customer met a manager who added nothing.
  • Manager sits down, asks about the trade, asks what's driving the timeline, repositions the deal. Two to four minutes of actual work.

All three can be recorded identically. Only one is manager involvement.

Score what you observe on a simple scale: no contact / greeting only / substantive. On a 12-deal audit I'd expect at least half to be substantive in a well-run store. If you find nine "greeting only" out of twelve, you don't have a T.O. problem — you have managers who are treating the T.O. as a box, and they learned that somewhere.

Also note when it happened. A manager who appears only after the customer says "let me think about it" is a save attempt, not involvement. Note how many T.O.s happen before the write-up versus after the objection.

Manager, walking up mid-worksheet: "Hey, I'm Dana, I run the floor. Before Kevin puts numbers together — is the Tahoe replacing something, or adding to the driveway?"

That's ten seconds and it changes the deal. Write down how many of your managers actually do it.

Column two: customer handoffs

Customer handoffs are where the audit gets uncomfortable, because everyone thinks their store handles them fine.

Watch for these five moments:

  1. BDC-set appointment to the greeting rep
  2. Greeting rep to the assigned salesperson (if you split those roles)
  3. Salesperson to manager
  4. Salesperson to F&I
  5. Delivery to service intro

At each one, note two things: did the receiving person know anything about the customer before they spoke? And did the customer have to repeat themselves?

The tell is verbal and easy to catch:

"So what brings you in today?" — said to a customer who booked an appointment two days ago and answered that exact question on the phone.

Every time you hear that, mark it. If eight of twelve appointment customers get asked their reason for visiting from scratch, your appointment notes aren't traveling. That's not a training issue; that's a handoff design issue. Someone needs to read the note, and right now nobody owns that.

The F&I handoff deserves its own note. Time it. Write down the minutes between "we have a deal" and "customer sits down in the box." If your report shows an average of 20 minutes and you observe 45, 50, and 38, your average is being pulled down by the easy cash deals and your finance-heavy customers are sitting alone in a chair getting cold.

Column three: process adherence

Don't audit ten steps. Audit three you actually care about, and define each one so precisely that two managers watching the same deal would score it the same way.

Examples of definitions tight enough to audit:

  • Needs assessment: rep asks at least three questions before walking to a vehicle, one of which covers timeline or current payment.
  • Test drive: customer drives, minimum eight minutes, includes a road type they'll actually use.
  • Trade walk: rep physically walks around the trade with the customer present and asks about it before appraisal.

Then just tally: done / partial / skipped.

The value isn't the score. It's the pattern. When you look at your tally sheet and see that trade walks happen on nine of twelve deals but needs assessment happens on four, you now know exactly what to work on Monday. And you can show a rep the specific deal.

Do not let "skipped" be the end of the note. Write down why it looked like it got skipped. Rep was covering three customers. Vehicle was blocked in. Customer said they were in a hurry. Some of those are staffing problems wearing a process costume.

Column four: follow-up ownership

Here's the one that separates a real audit from a report review.

Take the same 12 customers. Three days later, pull each record and ask three questions:

  • Who owns this customer by name? Not "the team." A person.
  • What was the last actual contact — not the last logged activity, but a call connected, a text replied to, an email opened by a human?
  • What's the next step and when? A date and an action, or nothing.

Follow-up ownership fails in a specific way: the customer belongs to whoever last touched them, which means they belong to nobody. A rep goes on a day off, a manager reassigns a deal to cover the phone call, the original rep comes back and assumes it's handled. Nine days later the customer buys down the street.

Listen to two or three actual follow-up calls from your audit set. You'll learn more in fifteen minutes than from a month of activity counts. What you're checking is whether the call had a point:

"Hey Marcus, it's Kevin at the store. Two things — the Highlander you drove is still here, and I got your trade re-appraised this morning, it came up nine hundred. Are you around Thursday or is Saturday better?"

versus

"Hi Marcus, just checking in, give me a call back."

Both count as one outbound call in the report. One of them moves the deal.

Turn the audit into two changes, not twelve

At the end you'll have a page full of gaps. Resist fixing all of them.

Pick the two with the biggest spread between what the report claimed and what you saw. Assign each one an owner and a way to check it in 30 days.

Sample output from a real-feeling audit:

  • Gap: T.O. logged on 11 of 12; substantive on 4. Owner: Dana. Change: manager walks the floor at the worksheet stage, not the objection stage. Check: re-observe 12 deals in 30 days, target 8 substantive.
  • Gap: 7 of 9 appointment customers asked their reason for visiting from scratch. Owner: BDC lead. Change: appointment note read aloud at the morning huddle for that day's appointments. Check: listen to 10 greetings in 30 days.

Then re-run the same audit next quarter with the same four columns. The point isn't to catch anybody. The point is to know whether your reports describe your dealership or just describe your CRM.

If you'd rather not sit on the floor with a clipboard every quarter, scoring recorded calls and handoff conversations against the same four columns gets you a lot of the same picture — that's the kind of thing MoreSignal is built to do. But start with the clipboard. You need to know what you're looking for before you automate looking for it.