A pallet of your product leaves the dock in Carrollton on a Tuesday. It arrives in Reno on Friday. The receiver signs the delivery receipt, drops the pallet in a rack, and goes home. Monday morning somebody cuts the shrink wrap and finds two cases crushed on the bottom tier. Now a customer service rep is asking a question nobody in your building can answer cleanly: how do we file a freight claim on this, and are we already too late?
Most shippers do not lose freight claims because the carrier was not liable. They lose because by the time anyone decides to pursue it, the documentation chain that proves liability has already been broken — usually at the dock, by a receiver who signed clean and had no reason to think it mattered. Knowing how to file a freight claim is half the job. The other half is a shipment record that survives the question “prove it.”
What filing a freight claim actually looks like in most companies
Watch the real workflow, not the one in the SOP binder:
The rep searches her inbox for the BOL, asks the warehouse for the PRO, and finds a carrier portal login a former employee set up. She downloads the POD and discovers a clean signature with no exception notation. She asks the receiver for photos of the packaging; the packaging went into the compactor Friday. She pulls the invoice from the ERP for value, fills out the carrier’s PDF claim form, emails it to a claims inbox, and hears nothing. Six weeks later somebody asks what happened to it. Nobody set a reminder. The file ages out.
Six systems, four people, no system of record. Every data element that claim required — BOL, PRO, class, weight, delivery timestamp, POD image, invoice value, PO — already existed inside your business at the moment of delivery. It was never assembled in one place while it still had evidentiary value. That is an architecture problem, not a diligence problem.
The clock is federal. Know which parts are law and which are tariff.
Freight claim timing is not carrier preference. For interstate motor freight it is codified, and the distinctions matter operationally.
What counts as a claim. Under 49 CFR 370.3(b), a proper claim requires three elements: facts sufficient to identify the shipment, an assertion of carrier liability, and a demand for a specified or determinable amount of money. All three, in writing.
The trap almost nobody knows about. The same regulation, at 370.3(c), states that damage notations on a freight bill or delivery receipt — and even carrier inspection reports — do not, standing alone, constitute a filed claim. Teams routinely believe exception-noting the POD started the clock. It did not; it preserved evidence. Conflating the two is how shippers reach month eight thinking they are protected.
How long you have. The Carmack Amendment at 49 U.S.C. 14706 bars a carrier from contracting for a filing window shorter than nine months from delivery. Check your actual bill of lading — nine months is the floor, not a guarantee.
What the carrier owes you back. Section 370.5 requires written acknowledgment within 30 days, including a statement of what additional documentation the carrier needs. Section 370.9 requires the carrier to pay, decline, or make a firm compromise offer in writing within 120 days, and if it cannot, to report status in writing every 60 days thereafter. Carriers miss these constantly. Shippers rarely hold them to it, because nobody is tracking the dates.
Concealed damage is the exception — and it is tariff, not statute. The NMFC shortened the concealed-damage notification window from 15 days to 5 days effective April 2015, and as transportation counsel George Carl Pezold has documented for the Transportation Loss Prevention Council, that item lives in the classification and carrier rules tariffs — not federal regulation. Your carrier’s tariff governs. Do not assume five days universally, and do not assume you have longer.
The architecture we would build
Here is the architecture we would build for a shipper running meaningful LTL volume. Nothing here is exotic; it is the same connected pattern behind any working freight API and transportation management system.
One shipment record, created at booking. When the order rates and books through the API, the record already holds BOL, PRO, NMFC class, dims, weight, carrier SCAC, invoice value, and PO. No one retypes anything later.
Delivery events write back automatically. The tracking API posts the delivery timestamp and POD image to the shipment record within hours — not when a rep goes hunting in week three.
Exception rules open the case, not people. Any POD returning with an exception notation, any OS and D code, any delivery flagged by the receiving app — the system opens a claim case automatically and timestamps day zero. The default becomes “a case exists and someone must close it,” instead of “a case exists only if someone remembers to open it.”
Photo capture bound to the PRO. Three phone photos tagged to the PRO before the pallet moves are worth more than any downstream process improvement. The highest-leverage item here is not software — it is a dock rule.
The claim assembles itself. The three elements required by 370.3(b) are generated from fields you already hold. A human reviews and sends. Clocks run automatically at 30, 120, and 60-day intervals, and escalate when the carrier goes quiet.
Humans stop being the transport layer for data and start doing what software cannot: negotiating disputed liability, fixing the packaging that caused the damage, and deciding which carriers have stopped earning their volume. That is the logic behind managed transportation programs — the machine handles state, people handle judgment.
The economics, honestly
Illustrative example — run your own numbers. A shipper moving 400 LTL shipments a month at a 1.5% exception rate sees roughly 72 claimable events a year. At a $900 average claim, a manual process recovering 35% — because half the files are abandoned or filed late — returns about $22,700. Connected records and enforced clocks lifting recovery to 60% return about $38,900. The delta is roughly $16,200 a year plus coordinator labor. These are modeled figures, not client results; your exception rate, claim size, and carrier mix move this materially.
Where this argument is weakest
Now the part most vendors skip. For a lot of shippers, building a claims system is the wrong project.
Carrier liability on LTL is frequently capped by released value in the classification or the carrier’s rules tariff, often stated in cents per pound. On a low-class commodity, a damaged 1,200-pound pallet worth $4,000 may cap out in the low hundreds of dollars no matter how flawless your paperwork is. Automation cannot recover money the tariff already says you are not owed. If that describes your freight, the higher-leverage moves happen before the truck loads: negotiating released value terms, correcting classification, and fixing packaging. Claims recovery is a lagging indicator; packaging and terms are leading ones.
There is also a scale threshold. If you file eight claims a year, a shared folder, a calendar reminder, and a written dock rule capture most of the benefit. Do not buy integration work to solve a problem a checklist solves. In our own experience, the build only earns its keep when exception volume is high enough that the manual process fails reliably — not occasionally.
And the context is worth keeping in view: CSCMP’s 36th Annual State of Logistics Report put U.S. business logistics costs at $2.58 trillion in 2024, about 8.8% of GDP. Claims recovery is a rounding error against that number. Rate execution, mode selection, and network design are not. If you only have bandwidth for one project this quarter, it probably is not claims.
The playbook
If you are going to do this, do it in this order:
1. Fix the dock first. Inspect before signing, note every exception on the delivery receipt with specifics, photograph the pallet and packaging before it moves, and never dispose of packaging until the claim closes. Costs nothing, protects everything.
2. Read your actual bill of lading and carrier rules tariff. Find the filing window and the released value terms for the classes you ship. You cannot manage a clock you have not read.
3. Stop treating the POD notation as the claim. File a written claim containing all three elements from 370.3(b). Date-stamp it and keep proof of delivery to the carrier.
4. Calendar the 30, 120, and 60-day dates the day you file. Then enforce them in writing. Carriers respond differently to shippers who cite the regulation.
5. Consolidate the record. Get BOL, PRO, POD, photos, invoice, and claim correspondence into one shipment record — even if version one is a spreadsheet and a folder convention.
6. Automate only what is provably breaking. Usually that is POD retrieval and clock tracking. Start there, not with a claims module nobody asked for.
7. Audit the root cause quarterly. If one lane, one carrier, or one SKU generates most of your claims, that is a packaging, classification, or mode decision — sometimes the answer is moving that freight to full truckload rather than getting better at claims.
If step six looks like a real project rather than a weekend, that is where an outside operator earns their fee. Our logistics API and automation consulting work starts by mapping what your systems already know before anyone proposes new software, and ELM runs FreightPOP TMS with free API setup for qualified shippers across a 60-plus carrier network. The goal is not a prettier claims queue. It is a shipment record complete enough that “prove it” has an answer sitting in a database, not in somebody’s inbox.
Get a Free Freight and Logistics Review
Tell us what you are shipping and we will show you exactly where the savings are. No consultants. No commitments. Just straight answers from operators who have been doing this for 20+ years. Call (866) 854-5341.