Ask a shipper where their freight goes missing and most will point at the road. Ask the loss data and it points somewhere else entirely: at the places freight stops moving. Cargo is rarely stolen in transit. It is stolen while it waits.
That distinction is not academic. It decides where the money for security should go, and most programmes spend it in the wrong place.
The journey, and the six places it pauses
A load between a manufacturer and a receiver stops more often than the bill of lading suggests. Each stop is a different exposure with a different failure mode.
1. The shipper’s dock, before it leaves
This is where the fastest-growing loss category happens, and no lock prevents it. A criminal presenting as a legitimate carrier — using a real motor carrier’s MC number, a lookalike email domain, or a compromised load board account — is handed the freight by the shipper’s own staff. There is no break-in to detect and no alarm to trip. The paperwork is in order because the paperwork was the attack.
Verisk CargoNet notes that criminals also use social engineering to misdirect drivers and redirect shipments to fraudulent addresses without ever taking physical possession at the origin. The load is delivered to the thief by a legitimate driver who believes they are doing their job.
2. The first stop after loading
Surveillance-led theft happens here. A load watched out of a distribution centre is followed to the driver’s first stop, which is frequently a fuel or food stop within the first two hours. The driver is inside; the trailer is not. This is the classic loss and it remains common because the pattern is predictable — drivers stop when they stop.
3. The overnight rest
The largest single block of exposure by hours. Federal hours-of-service rules mean a loaded trailer will be parked, and parking capacity is short enough that the choice is often between a lit truck stop and an unlit shoulder. CargoNet’s Q3 2025 analysis specifically named unattended loaded trailers in Southern California, the Bay Area, Phoenix and Lake Tahoe as targets.
Phoenix appearing on that list is not incidental. It is where loads out of the ports and the Inland Empire naturally reach the end of a driving shift.
4. The drop yard
Trailers dropped for later collection can sit for days. A yard with a gate but no reconciliation — nobody comparing what should be present against what is present — will not notice a missing trailer until someone comes to collect it. Discovery delays of 48 to 72 hours are normal, and recovery odds collapse over that window.
5. The cross-dock
Where pilferage lives. Freight is broken down, re-sorted and re-loaded, often at speed, often by mixed staff. Cartons removed here are not discovered until a receiving count that may be weeks away, and by then the custody chain has too many hands in it to reconstruct.
6. The receiver’s yard, before check-in
The forgotten one. A trailer that arrives after hours waits in the receiver’s yard until the dock opens. The shipper considers the load delivered. The receiver has not signed for it. Nobody is actively responsible for it, which is exactly the condition thieves look for.
Where the loss is recorded versus where it happened
Loss data is reported by discovery location, not by exposure location. A trailer taken overnight in Arizona and found empty in California is often logged in California. A carton removed at a cross-dock in one state and counted short at a warehouse in another is attributed to the warehouse.
This systematically flatters the safe parts of the network and blames the last party to touch the freight. If your internal loss reporting works the same way — and most do — you are directing your controls at the place the problem was noticed rather than the place it was created.
The corrective is dull and effective: record for every loss the last verified point of custody with a timestamp, not just the point of discovery. Six months of that data will point somewhere different from where instinct points.
The numbers behind the strategy shift
In Q3 2025 CargoNet documented 772 cargo theft incidents across the United States and Canada — only 1% above the same quarter a year earlier, and 10% below Q2. Total value taken was $111.88 million. But the average shipment loss doubled, from $168,448 to $336,787.
Flat volume with doubled value is a signature. It means target selection improved. Someone is choosing which trailers to take, which requires knowing what is on them — from the load board, from a compromised system, or from a person.
That is why “more security” is often the wrong prescription. The relevant question is how the adversary is learning what is inside.
What to change, in order
- Map your own dwell. For a representative week, record how many hours each load spent stationary and where. Most shippers have never done this and are startled by the answer.
- Harden the dock handover. Verify the carrier through a number you already hold, not one presented to you. Confirm the driver against dispatch. Photograph the tractor and the driver’s identification at pickup.
- Attack the overnight. Where high-value lanes require an overnight, pre-arrange a secured yard rather than leaving the choice to the driver at the end of a shift.
- Reconcile drop yards daily. A trailer count against a manifest, every day, catches in hours what otherwise surfaces in days.
- Assume the visible tracker will be found. Fit a second, independently powered device.
- Close the receiver gap. Define in the contract who holds responsibility for a trailer waiting in the receiver’s yard before check-in.
None of this is expensive relative to a single $336,000 loss. All of it is procedural, which is why it tends to be skipped in favour of hardware that photographs well.
Honeybadger Solutions covers this ground as transportation and cargo security, including yard control and the process side that hardware does not touch. The wider pattern of what is being taken and why is set out in our companion piece on the cargo theft threat picture, and where a repeating loss suggests information is leaking from inside, that becomes a matter for private investigations.
Building a loss map for your own network
The six pause points are the general pattern. Your own network has a specific one, and finding it is a data exercise rather than a guessing exercise.
Take two years of claims and, for each one, record three things that most claim files omit: the last confirmed location where the freight was verified intact, the next point at which it was checked, and the elapsed time between them. Then plot the losses against that interval rather than against the location where the shortage was discovered.
Almost every operation that does this finds the loss concentrated in one or two intervals that nobody had identified, usually because the discovery happened far downstream and the claim was filed against the wrong stage. Fixing the interval where the loss occurs is cheap. Adding controls at the place where it was noticed is expensive and ineffective.
Seals are evidence, not security
Seal programmes fail in predictable ways, and understanding why changes how they should be run.
A seal does not prevent entry; it makes entry visible. That value depends entirely on the seal number being recorded at every transfer and verified against the record at the next one. In most operations the number is recorded at origin and checked at destination, which means a seal cut and replaced anywhere in between is invisible unless someone compares the numbers — and if the replacement seal’s number is written on the paperwork at an intermediate stop, the record simply follows the theft.
Three practices fix most of it. Record the seal number at every custody transfer, not just the ends. Verify the number against the previous record rather than against the paperwork in the driver’s hand. And treat a broken or mismatched seal as a stop-and-report event rather than something to note and continue past — the most common failure is a mismatch observed, noted, and never escalated.
Bolt seals are meaningfully harder to defeat than plastic strip seals and cost very little more. For high-value freight, the difference is worth it.
The internal component
Losses that repeat at the same interval with no sign of forced entry, in a facility with reasonable access control, are usually not the work of outsiders. That is uncomfortable and it is where the data consistently points.
Internal involvement typically takes one of three forms: someone inside supplying information about which loads are valuable and when they move; someone in the chain of custody altering counts or paperwork; or collusion at a facility where freight is staged. None of these is addressed by more fencing.
What addresses them is segregation of duties — the person who counts should not be the person who records, and neither should be the person who investigates discrepancies — plus rotation of who handles high-value staging, reconciliation performed by someone outside the operation, and an anonymous reporting channel that people believe is actually anonymous. Where the pattern is established, a discreet investigation is more productive than an announced audit, which simply pauses the activity.
Documentation is the control nobody funds
In practice, the difference between a recoverable claim and an unrecoverable one is the quality of the record at each transfer.
What a defensible record contains: piece count verified and signed by both parties at each handover, seal numbers recorded and verified, photographs of the load at origin and at each transfer point where practical, timestamps that come from a system rather than from handwriting, and exception notes made contemporaneously rather than reconstructed afterwards.
Photographs deserve particular mention because they are nearly free and they resolve the most common dispute, which is whether a shortage existed before or after a given stage. A photograph of a sealed, loaded trailer at origin with the seal number legible ends an argument that can otherwise run for months.
Yards, cameras and what coverage should actually watch
Most yard camera systems are aimed at the gate. The gate is where the record is made; the loss usually happens elsewhere in the yard, in the row of trailers nobody can see from the office.
Better coverage priorities: the rows where loaded trailers are staged, the perimeter where a tractor could back in from an adjoining property, the area where seals are applied and checked, and the approach to the building rather than the door itself. Add lighting to the same places — detection and deterrence both depend on someone being visible while deciding.
Retention matters as much as coverage. Shortages are frequently discovered days later, and footage that has already rolled over is footage that does not exist. Thirty days is a reasonable target for a yard handling valuable freight.
What to change, in order, for a typical operation
- Build the loss map from claims data and find your actual interval.
- Record and verify seal numbers at every custody transfer, with bolt seals on valuable loads.
- Photograph loads at origin and at transfer points; make it part of the process, not an exception.
- Separate counting, recording and discrepancy investigation.
- Move camera coverage and lighting from the gate to the staging rows.
- Prohibit stops within the first two hundred miles after loading.
- Where losses repeat without external evidence, treat it as an investigative matter rather than a hardening problem.
Browse by topic
Security guard services · Private investigations · Cybersecurity · Digital forensics · Financial fraud investigation · Executive protection · All articles