Spot-Check Audit: Review 20 Live Listings in 15 Minutes
You do not need a day in a conference room to know if Marketplace is clean. A spot-check audit of 20 live listings in 15 minutes will surface wrong photos, stale prices, and sold units still advertised.
It is 4:45 p.m. on a Friday. The GM asks whether Facebook Marketplace "looks right." Nobody can answer without opening forty listings one by one. The honest status is that the store last looked carefully at posting day, and posting day for the oldest live ads was six weeks ago. Sold units, price cuts that never hit Marketplace, and a lead photo of the wrong color of the same model are all plausible. You will not find all of them tonight. You can find enough of them in fifteen minutes to know whether you have a system problem.
A spot-check audit: review 20 live listings in 15 minutes is a sampling discipline, not a full catalog review. Twenty is enough to see patterns. Fifteen minutes is short enough that a manager will actually do it twice a week instead of promising a quarterly deep dive that never happens.
This is for general managers, internet directors, and used-car managers who need a pulse check on live Marketplace inventory without shutting down the desk.
Why a Spot-Check Audit of 20 Live Listings Beats a Quarterly Deep Dive
Full audits die because they are scheduled like projects. Live listings decay daily: a unit sells, a price changes on the website, a rep restocks photos in the wrong folder. A small, frequent sample matches how the lot actually changes.
What the sample is for
- Detecting system errors (import mapping, sold-unit hygiene, shared photo folders), not nitpicking one clumsy sentence.
- Giving the GM a real answer: "we sampled 20, 4 failed, all four were sold-still-live or price mismatch."
- Feeding coaching. The same fail reason three times is next week's floor meeting.
What the sample is not for
- Proving that every listing is perfect. A clean sample of 20 does not certify the other 80.
- Replacing a pre-live review. Catching errors after they are public is more expensive than catching them in draft.
- Generating fake precision. Do not turn 4 fails out of 20 into a "quality score" you present as a scientific rate. Call it what it is: a spot check.
If you want a scored model to sit beside the sample, use a listing quality rubric on the same twenty, not a different process invented in the moment.
How to Pick the 20 Live Listings
Random-looking samples that are actually "the ones I remember" miss the ugly inventory. Build the twenty on purpose.
Use four buckets of five
Consider a lot with 80 live Marketplace units. Pull:
- Five newest. Tests whether today's process still works.
- Five oldest / highest days-in-stock. Tests decay: stale copy, stale mileage, stale price.
- Five from a different rep than last time. Tests whether quality is a person problem or a store problem.
- Five from a problem-prone segment you already know is messy: twins of the same color, rebuilt titles, auction-grade as-is, or units that just came out of recon.
Rotate the fourth bucket each audit so you are not always beating up the same segment.
Pull from the live source, not from memory
Open the actual Marketplace view or your inventory filtered to posted. Do not audit the spreadsheet you wish were true. If a unit is posted in your system and missing on Marketplace, that is a finding, not a skip.
Timebox ruthlessly
Forty-five seconds per listing is the math (20 × 45 seconds = 15 minutes). You are not rewriting. You are marking pass/fail with a reason. Anything that needs a rewrite goes on a fix list for after the sample.
The 15-Minute Spot-Check Checklist
Same questions, every unit, in the same order. If you improvise, you will spend four minutes on a description you like and skip price on the next one.
The nine looks (under 45 seconds)
- Still for sale? If it sold or is pending in the DMS, fail.
- Price matches the intended ask? Website / buyer’s guide / in-store number.
- Lead photo is this vehicle? Color, body, wheels, not the twin in the next stall.
- Year/make/model/trim in the title match the unit?
- Mileage in the listing is plausible vs. last known odometer. You are not doing a new inspection; you are catching 40,000 vs. 140,000.
- Title status is not hidden on branded/rebuilt/salvage units.
- Opening lines are unit-specific, not a cloned blob.
- No bait pricing or "call for price" in the body.
- No obvious wrong-claim (4WD on a FWD, leather on cloth, certified when you are not).
Scoring during the audit
Keep it binary plus a tag:
| Tag | Example |
|---|---|
| SOLD | Still live after a deal |
| PRICE | Marketplace ≠ intended ask |
| PHOTO | Wrong vehicle or unusable cover |
| ID | Trim/year/drivetrain mismatch |
| CLAIM | Superlative or equipment you cannot support |
| STALE | Intake copy on a long-stock unit, obviously untouched |
| PASS | No critical issue in the nine looks |
Write the stock number and tag on a pad or a simple sheet. That is the entire work product of the fifteen minutes.
Aged units and photos
Long-stock listings fail this audit on STALE and PHOTO more than on ID. If the oldest five are consistently dirty or dark, you do not have a "Marketplace problem." You have a photo-refresh problem. Pair those fails with refreshing photos on aged units before cutting price and with a photo order and cover image strategy so the fix is a standard, not a one-off reshoot.
What to Do With the Findings Before You Leave
An audit with no same-day action is a report nobody asked for. Spend the next twenty minutes (outside the fifteen) on the fails only.
Immediate kills
- Sold-still-live: mark sold and archive. Do this before you rewrite anything. Every hour it stays up is a bad conversation.
- Catastrophic price or identity errors: unpublish or correct now, then handle any open buyer threads.
Same-day assigns
- Photo and ID fails: assign back to the owning rep or associate with the stock number and the tag. Do not "mention it at the meeting."
- Stale long-stock copy: batch them for a description pass, not twenty individual debates.
Pattern rule
If three or more of the twenty share a tag, that tag is a process change, not a person change:
- Many SOLD tags: sold hygiene is broken. The mark-sold step is not part of the deal process.
- Many PRICE tags: Marketplace is not in the price-change workflow.
- Many PHOTO tags: folder naming or picker discipline is broken at intake.
- Many CLAIM tags: templates are inventing equipment.
What to tell the GM
Three sentences:
- Sample size and mix (newest, oldest, by rep, by segment).
- Fail count by tag.
- What was killed or assigned today.
That is a management answer. "We think it's fine" is not.
Cadence
Twice a week is a conservative operating standard for a store that posts at volume. Once a week is a minimum for a smaller lot. Daily full-catalog review is not required and will not happen. Protect the fifteen minutes on the calendar the same way you protect a meeting you cannot skip.
How PostToMarket Automates This
The reason a live-listing audit takes all afternoon is not the nine looks. It is hunting: which units are posted, who owns them, which are already sold in one system and live in another, and where the photos live.
- Inventory statuses (draft, posting, posted, sold, failed) let you filter to what is actually live before you sample.
- Team queue view shows who claimed each unit so a PHOTO fail goes back to a person, not a group chat.
- Mark as sold and archive from inventory or the queue, which is the correct response to every SOLD tag you find.
- Analytics by rep and source help you pick the "different rep" bucket instead of auditing the same high-volume poster every time.
- Facebook photo picker and editable AI copy from specs make the fix after a fail a structured edit, not a rebuild from a DMS dump.
- Import history and row-level errors explain PRICE and ID patterns that started in the file, not on Marketplace.
- Mobile dashboard so a manager can run the sample on the lot instead of only from a back-office PC.
PostToMarket is independent and not endorsed by Meta. A clean inventory view does not guarantee reach. It does make a 15-minute audit of 20 live listings a real habit instead of a myth.
Fifteen Minutes Is Enough to Know If You Are Lying to Shoppers
Spot-check audit: review 20 live listings in 15 minutes. That sentence should be on the used-car manager's weekly calendar. The sample will not make the catalog perfect. It will stop the store from claiming the catalog is fine while sold trucks and wrong photos sit in public.
If you cannot name this week's fail tags, you are not managing Marketplace quality. You are hoping. PostToMarket keeps posted vs. sold, owner, photos, and copy in one place so the fifteen minutes is spent looking at listings, not assembling them.
Starter is $59 per month for a one-login independent lot that still needs a posted filter and a sold button. Dealership is $299 per month for up to 13 logins when the audit has to assign fails to associates and reps by name. Start your 10-day free trial with secure Stripe checkout. Cancel anytime. Run the twenty this week. The tags will tell you what the weekend would have told you the hard way.