If you have not audited your NMFC codes since mid-2025, you are quoting LTL against a classification system that no longer exists. The National Motor Freight Traffic Association's Docket 2025-1 was the most significant classification overhaul in nearly a century: it replaced the legacy 11-tier system with a 13-subprovision density scale and reclassified more than 2,000 commodity codes.

Two follow-up dockets have landed since. Docket 2025-2 took effect on December 6, 2025, refining item language and handling flags. Docket 2026-1 was released on February 6, 2026. The scale stopped being rebuilt after the first docket, but individual commodities keep moving.

The operational consequence is narrow and expensive: a quote generated from a stale class code re-rates on the invoice, and the difference is a dispute you have to work rather than revenue you booked.

What actually changed

The old system sorted freight into 11 density tiers. The new one uses 13, adding class 50 and class 55 at the dense end of the scale. Alongside that, special handling specifications were tightened.

The effect splits cleanly by freight profile:

• Dense freight benefits. Metal components, machinery, anything heavy and compact can drop into a lower class, which lowers the rate per hundredweight.
• Light and irregular freight gets worse. Furniture, signage, foam and similar commodities can move up a class under the new handling specifications, raising the rate.

If your book skews one way, the aggregate impact is directional and predictable. If it is mixed — which most brokerage books are — the impact is per-shipment and invisible in aggregate.

Why this becomes a quoting problem rather than a pricing problem

Freight classification is a data-freshness problem wearing a pricing costume.

The class code is not a judgement call. Given dimensions, weight and commodity, the correct class is determinable. The failure mode is not that anyone decides wrongly — it is that the code sitting in the customer's item master, the code in your quoting tool, and the code the carrier will actually rate against were all set at different times against different dockets.

Nobody notices until the invoice arrives with a reclassification charge attached.

• Customer item master — What goes wrong: Class set before Docket 2025-1, never revisited · What it costs: Every quote on that item is wrong
• Missing dimensions — What goes wrong: Class estimated from commodity name alone · What it costs: Density cannot be computed at all
• Quoting tool defaults — What goes wrong: Legacy 11-tier logic still applied · What it costs: Systematically wrong on new tiers
• Carrier re-rate — What goes wrong: Discovered at invoice, not at tender · What it costs: Dispute cycle plus margin loss
• Handling flags — What goes wrong: New special handling specs not applied · What it costs: Accessorials appear unexpectedly

How do you stop a reclassification charge before it happens?

You stop a reclassification charge by computing density at quote time from actual dimensions and weight, comparing the result against the class code on the order, and flagging the mismatch before the load tenders — not by auditing invoices after the fact.

That is three mechanical steps, and all three are things software should be doing on every shipment rather than things a rep should be doing on the ones they happen to look at:

1. Capture dimensions and weight at intake. If the request arrives as an email with a commodity description and no dims, the agent asks for them before quoting rather than guessing a class.
2. Compute density and derive the class. Pounds per cubic foot against the current 13-tier scale, plus the handling specifications that now apply.
3. Compare and flag. If the derived class differs from the class on the order, surface it to a human with both numbers and the density that produced them.

This is the same intake-and-validate pattern that makes automated quoting work generally. The agent is not being asked to negotiate anything. It is being asked to notice that two numbers disagree — at volume, on every shipment, without getting bored.

The dimensional data problem underneath

Most teams that try to automate this discover their real constraint is not classification logic. It is that a meaningful share of inbound LTL requests arrive without usable dimensions.

That is a fixable intake problem, and it is worth fixing first. An agent handling inbound quote requests can hold the shipment at intake and ask for the missing dimensions in the same reply thread, in seconds, rather than a rep noticing three hours later. The classification logic downstream is trivial once the inputs are complete — which is the general shape of most logistics automation that actually pays off.

Teams working through the broader differences between LTL and truckload automation will find the intake burden is the main structural gap; the mechanics of LTL versus FTL quoting diverge mostly at that step.

What to do this month

Three actions, in order of payback:

Audit the item masters for your top 20 customers. These are the codes generating the most quotes. If they were set before July 2025, assume they are stale until proven otherwise. This is a finite, one-time piece of work with a permanent benefit.

Instrument the gap. Compare derived class against quoted class on the last 60 days of shipments. The rate of disagreement tells you the size of the exposure and makes the case for fixing the rest.

Close the intake hole. Requests arriving without dimensions are the root cause. Automating the follow-up for missing dims is a small, safe, high-volume workflow — and a reasonable first agent project for a team that has not deployed one yet.

Frequently asked questions

Do the NMFC changes raise LTL costs overall? Not uniformly. Dense freight generally benefits from the new class 50 and 55 tiers, while lightweight and irregular freight tends to move up a class under the revised handling specifications. The net effect depends entirely on your commodity mix.

How often do NMFC codes change now? The scale itself is stable after Docket 2025-1, but individual commodities keep moving. Docket 2025-2 took effect in December 2025 and Docket 2026-1 was released in February 2026. Treat classification data as something that needs a refresh process, not a one-time setup.

Can classification be automated if we do not have dimensions on file? No, and that is the point. Density-based classification requires dimensions and weight. The automation worth building first is the one that collects missing dimensions at intake, because without it the classification step has nothing to work with.

Who absorbs a reclassification charge? It depends on the contract, but in practice a broker who quoted a class that could not be supported by the shipment's actual density will usually eat some part of it or spend hours arguing about it. Catching it at tender is cheaper than winning the dispute.

The bottom line

The NMFC overhaul reclassified more than 2,000 commodity codes and replaced an 11-tier scale with 13 tiers. Shippers who have not audited since mid-2025 may be paying incorrect rates on every LTL shipment.

Compute density at quote time, compare it against the class on the order, and flag disagreements before tender. The classification math is easy. Getting complete dimensions at intake is the part worth automating.

Debales deploys AI agents for freight quoting, order processing, ETA updates, and multi-channel customer communication — collecting missing shipment data at intake and validating it before a quote goes out. Book a demo.