Modern Dealer Audit Technology: the Evolution of Dealer Self Audits

dealer audit technology

Between two floorplan audits, a dealer can sell a funded car, use the money for weeks and pay it off just before anyone comes to count. On audit day the book looks clean. Fraud happens between the snapshots, and a funder that only sees the stock on audit day finds out late, or not at all.

That’s the real job of dealer audit technology: to let you see a dealer going wrong before the loss lands, without having to take the dealer’s word for it. This guide covers what a dealer audit has to prove, how the main methods compare (from the clipboard to connected-car data), and the questions to ask before you choose a dealer audit system.

Why funders audit dealer stock

“Car dealership audit” can mean several things: a tax audit, a manufacturer’s warranty or sales-standards review, or a compliance check on how cars are sold and financed. This article is about the one a floorplan funder runs. The funder has paid for the dealer’s stock, it holds security over each funded vehicle until that vehicle is paid off, and it needs to know the cars are still there.

The requirement to allow these checks normally sits in the floorplan agreement and in the funder’s own credit policy. In the US, the Office of the Comptroller of the Currency gives national banks a clear benchmark in its Floor Plan Lending handbook. In summary, it says:

  • floorplan checks should happen at least quarterly, and more often (for example monthly) for pay-as-sold lines or dealers whose finances are weakening;
  • checks should include unannounced visits as well as pre-announced ones;
  • every unit should be verified by its serial number (the VIN) against the funder’s records;
  • missing units, and the dealer’s explanation for each, should be verified and followed up at the next check;
  • a car missing at the check and paid off straight afterwards is a sign the dealer may be using the sale money as a float.

That’s guidance for US national banks, not a law that binds every funder, but it’s a fair description of what a credit committee expects anywhere. Outside the US, check your own regulator’s guidance and your floorplan agreements. If a dealer is selling a car without paying it off, you’re looking at selling out of trust, and the audit is usually where it shows up.

What a dealer audit has to prove

Whatever the method, a dealer audit is only as good as the evidence it leaves behind. For each funded vehicle, your credit team should be able to answer five questions:

  • Which vehicle? The VIN, read from the car itself, not from a key tag or a stock list.
  • Where? At the dealer’s site or one of its known locations, such as a compound or bodyshop.
  • When? At the time of the check, not last week.
  • Who recorded it? A named person, whether that’s dealer staff, your own team or a third-party auditor.
  • Does it match the book? Every VIN checked against what you’ve funded, with anything missing, sold or unexplained flagged.

There’s a sixth thing an audit should capture: whether the dealer did the check at all. A dealer who misses a check or keeps asking for more time is often telling you something before the statements do.

From field audits to dealer self-audits

For decades the standard was a field auditor driving from site to site with a printed stock list, ticking off cars and posting or keying the results back to head office. It works, but it’s expensive, it’s infrequent, and it leaves weeks or months between checks. Dealers also get good at spotting when the auditor is due.

Dealer self-audits grew out of that cost. If dealer staff can walk their own lot with a phone, a funder can check each dealer far more often for the same money, and send its auditors where they’re needed most.

Any credit manager’s first objection is the right one: why would you trust a dealer to audit itself? Making the check easy for the dealer doesn’t make it accurate. A self-audit is only worth relying on when the evidence holds up whoever walks the lot. In practice that means:

  • live capture in the app at the time of the check, with no uploads from the camera roll;
  • a location and a timestamp on every photo, compared with the dealer’s site and the window you set;
  • the VIN read from the photo and matched against your funded book;
  • the funder, not the dealer, deciding when a check happens and who does it;
  • a missed or late check flagged to credit as a risk, not filed as an admin reminder.

With those in place, a self-audit becomes evidence. Without them, it’s a dealer telling you everything’s fine.

Dealer audit technologies compared

Most dealer audit tools borrowed their technology from warehousing and logistics. Each one proves something slightly different, and that difference is what matters to a funder.

Paper and clipboard

An auditor works through a printed list and marks each car as seen. The record is only as good as the auditor’s care in checking the VIN rather than the stock number or the number plate. Results reach the credit team days later, after someone has typed them in. And because the visit is planned, the dealer usually knows it’s coming.

Barcode and QR code scanning

A label with a barcode or QR code goes on each vehicle (or on its keys) when it arrives, and is scanned with a handheld reader or a phone. It’s cheap and quick. The weakness is that a scan proves the label was there, not the car. A label on a key fob can be scanned in the sales office while the car is somewhere else, and a label can be moved to another vehicle.

NFC tags

Near-field communication is the radio technology behind contactless payments. It works at 13.56 MHz over a very short distance (the NFC Forum gives a typical range of up to about 2 cm), so the phone has to be held next to the tag, usually a sticker on the windscreen. That forces someone to stand at each car, which is a real improvement on a key-tag scan. But it’s still the tag that’s being read, every vehicle needs one fitted on arrival, and a tag that’s peeled off and stuck somewhere else reads just the same.

RFID

Radio-frequency identification reads tags from further away. Passive UHF RFID, the kind used in retail and logistics, reads from a few centimetres up to several metres, so a reader can pick up many tags in one sweep. That suits a warehouse of boxed goods. On a car lot it brings the same problem as the other tags, only worse: a reader can’t tell a row of cars from a tray of tags in a drawer. It also needs tags and readers bought and fitted.

Live photo capture with location and time

Here the evidence is the vehicle itself. Every car carries a 17-character VIN, and on most cars it can be read through the base of the windscreen (in the US, federal rules require that for passenger cars). Whoever walks the lot photographs the VIN with a phone, and the photo carries its own location and time. Software reads the VIN from the image and matches it against the funded book. Extra photos can record the car’s condition or any damage.

Close-up of a vehicle identification number (VIN) plate seen through the base of a car windscreen

Nothing has to be fitted to the car, and a phone is the only equipment. The risks are all about the photo: an old picture from the camera roll, a photo of a screen showing someone else’s photo, an image taken off site, or one that’s been edited. A photo system is only as strong as its controls on those. Look for one that forces live capture in the app and checks every image before you rely on it.

Connected-car data

Most new cars now report data from their own on-board systems. There are two ways to use that for stock checks. One is a device plugged into each car’s diagnostic (OBD-II) port, which means buying and fitting a device per car, and it can be unplugged. The other is the vehicle’s own connected systems, with nothing fitted at all.

CheckVentory FunderConnect takes the second route. Once a day by default, it checks whether each funded vehicle is active, whether it has moved, and whether it’s on the lot or off it. Each vehicle is read against the funder’s floorplan data, so the system knows whether it’s new, used or a demo. The dealer’s documented locations, such as compounds and bodyshops, count as on the lot. The funder’s own rules then decide what’s an exception: a demo out on the road is expected, while a brand-new car being driven that isn’t registered as a demo is flagged as likely sold. Exceptions go to the dealer to explain, before anyone plans a visit.

With our first customer, more than 94% of funded vehicles report in within any rolling 30-day period. That still leaves gaps. Some vehicles don’t report, and connected-car data can’t tell you that a car parked on the lot has been sold and is waiting for delivery. Settlement checks and photo-based audits cover those, which is why connected-car data works alongside physical and self-audits rather than replacing them.

At a glance

MethodWhat it provesWhat a dishonest dealer could exploitWhat it needs
Paper and clipboardAn auditor saw a car on that dayAdvance notice, cars moved between sites to be counted twice, keying errorsStaff travelling to every site
Barcode or QR codeA label was scannedLabels on keys scanned without the car; labels moved between carsA label on every vehicle
NFC tagA phone was held next to the tagTags peeled off and re-stuckA tag on every vehicle
RFIDA reader picked up the tag, from up to several metresLoose tags read as if the cars were thereTags and readers
Live photo with location and timeThat VIN, at that place, at that time, recorded by a named personOld or edited photos and off-site capture, unless the app forces live capture and checks every imageA phone
Connected-car dataEach vehicle is active or not, has moved or not, and is on or off the lot, every dayA parked car that’s already sold; vehicles that don’t reportConnected vehicles and the funder’s floorplan data

Our 2021 white paper compared several of these methods in more detail.

Choosing a dealer audit system: what could a dealer exploit?

No single method suits every funder, but one question sorts them quickly: if a dealer wanted to hide something, how would they get past this? Here are the questions to put to any dealer audit software, and the answers you should expect.

Does it prove the vehicle, or a tag? Labels, NFC stickers and RFID tags prove the tag was there. You want evidence taken from the car itself, which in practice means its VIN.

Who decides when a check happens? If the dealer chooses the day, the dealer has time to tidy up. You should be able to start a check without warning, on a schedule or when something looks wrong, and choose who carries it out.

Can the evidence be faked or reused? Ask how the system stops uploads from the camera roll, photos of a screen, edited or AI-generated images, and photos taken away from the site. Each photo should carry its own location, time and integrity checks, so you can defend an exception later.

Is every VIN matched against your funded book? Matching by hand after the event is slow and easy to get wrong. The system should work from your book, show the person on the lot which cars to find, and flag anything missing, sold or unexplained.

What happens when a dealer doesn’t do the check? Silence is a risk signal. A missed deadline should land in the same queue as a missing car, the same day.

What does the report show your credit team? A good dealer audit reporting system shows the exceptions, not a page of green ticks: the dealer’s reason for each missing car, the evidence they’ve added, what your team decided, and the history across past checks, so you see the pattern at a dealer and not just this month’s answer.

Does it work with your physical audits? Self-audits and remote checks narrow the gap between visits. They don’t remove the need for someone on the lot for a new facility, a dealer you’re worried about or a dispute. Everyone should work in the same system, so results compare like with like.

What has to be bought or fitted? Tags and readers cost money per vehicle and per site. Phone-based checks and connected-car data need nothing fitted to the car.

How CheckVentory does it

We built CheckVentory for funders, around those questions. Three parts work together, and none of them replaces your field audit programme.

  • CheckVentory Auditor runs the checks. You choose when a check happens and who does it: the dealer, your own team or a third-party auditor. Each funded vehicle is recorded with a live, geo-tagged, time-stamped photo, and automated checks look at every image for location, VIN, distance, time and signs of AI or editing. Every VIN is matched against your funded book. A typical 50-vehicle check takes minutes. Missing vehicles, failed photos and late checks all go to your credit team as exceptions, and the dealer has to account for each one. See how Auditor works.
  • CheckVentory FunderConnect checks every funded vehicle each day using connected-car data, so audit effort goes on the vehicles that can’t be confirmed.
  • Physical audits, when you want someone on the lot, arranged through our audit partners and recorded in the same system. We work with Bureau Veritas and with local audit firms in each market. CheckVentory isn’t tied to any one audit firm or technology, so we use whichever suits the funder.

Many funders combine them in either order. A dealer self-audits the whole lot and an auditor looks closer at the few cars that raised questions, or an auditor’s physical audit comes first and the cars they couldn’t find go back to the dealer to account for.

Dealer audit FAQ

What is a dealer audit?

In floorplan finance, a dealer audit (or floorplan check) confirms that every vehicle a funder has paid for is still at the dealer and unsold, by checking each vehicle’s VIN against the funder’s records. Any car that’s missing has to be explained, for example as sold and awaiting settlement, out on demo or at a bodyshop.

What are the audit requirements for a car dealership with floorplan finance?

They’re set mainly by the floorplan agreement and the funder’s credit policy. For US national banks, the OCC’s Floor Plan Lending handbook says checks should happen at least quarterly, more often for higher-risk lines, should include unannounced visits, and should verify every unit by serial number, with exceptions followed up.

Can a funder rely on a dealer self-audit?

Only if the evidence holds up. That means live capture rather than uploaded photos, a location and timestamp on every photo, the VIN read and matched against the funded book, the funder deciding when the check happens, and a missed check treated as a risk signal.

How do funders audit a whole dealer network?

Check each dealer as often as its risk warrants rather than on one fixed cycle, and include checks the dealer doesn’t see coming. For dealer groups with several sites, check every site at the same time, so a car can’t be moved from one site to another and counted twice. We’ve written more about group-wide audits.

Does connected-car data replace physical audits?

No. It can confirm daily that most funded vehicles are on the lot, which cuts the number that need checking by hand. But some vehicles don’t report, and it can’t show that a parked car has been sold. Self-audits, settlement checks and physical audits cover those.

Is a dealer audit the same as a dealer performance audit?

No. A dealer performance audit is usually a manufacturer reviewing a dealer against its sales, service or warranty standards. A floorplan dealer audit is a funder confirming its collateral is where it should be. Some dealer networks run both.

If you’d like to see what reaches a credit team after a check, have a look at our sample audit record. To see a check run on your own dealers, book a 30-minute demo.

For floorplan funders

See the dealer going wrong before the loss lands

CheckVentory Auditor and PreCheck give you evidence on every funded vehicle, from the first request to the final settlement. We’ll show you how it works with your own portfolio, in 30 minutes.