ArcBest’s August Tonnage Gain Shows Why LTL Demand Needs a Density Test

ArcBest’s August results look like a strong demand signal at first glance. Tonnage in its asset-based business, which includes ABF Freight, increased 9% year over year. Revenue per day also rose 9%, accelerating from 7.7% growth in July. For shippers watching for an LTL recovery, those numbers are encouraging—but they are not enough to justify a network-wide change in routing rules.

The reason is in the composition. FreightWaves reported that ArcBest’s daily shipments fell 4% year over year while weight per shipment increased 14%. Yield, including fuel surcharges, was unchanged, and yield excluding fuel declined by a low-single-digit percentage. This was a heavier network, not necessarily a busier one.

That distinction matters. A carrier can move more tons with fewer bills when industrial customers tender larger shipments or truckload freight migrates into the LTL network. The added weight can improve trailer utilization and terminal economics without proving that demand has strengthened across the shipper base. Transportation teams should therefore run a density test before interpreting carrier tonnage as permission to change consolidation, mode-conversion, or procurement rules.

Separate four signals that tonnage blends together

Headline tonnage combines several operational effects. A useful reading starts by separating four measures:

1. Shipment count: Are more individual orders entering the network? ArcBest’s 4% decline says no at the aggregate level.
2. Weight per shipment: Are tenders becoming larger? ArcBest’s 14% increase more than offset the shipment decline.
3. Revenue quality: Is revenue growing because of core pricing, fuel, shipment size, or accessorials? ArcBest’s revenue per shipment rose 14%, but fuel and mix contributed materially.
4. Network density: Are compatible shipments concentrating on the same origin-destination pairs and service days? Company-level releases rarely answer this, but it is the variable that determines whether a shipper can consolidate economically.

The cross-carrier picture reinforces the need to disaggregate. Saia reported an 8.7% August tonnage increase, made up of 1.1% growth in daily shipments and a 7.5% increase in weight per shipment. Yet its two-year-stacked tonnage growth slowed from 8.7% in July to 6.5% in August. Meanwhile, XPO’s August metrics moved in the opposite mix direction: tonnage increased 3.7% as daily shipments rose 5.7%, while weight per shipment fell 1.8%.

Three carriers can all report positive tonnage while describing three different networks. ArcBest’s result points to heavier freight, Saia combines modest shipment growth with heavier pieces, and XPO shows more but lighter shipments. That is why a single carrier’s headline should be treated as a hypothesis for lane analysis—not a market verdict.

Build a lane-level LTL density test

The practical test can be run weekly using shipment records already captured in a TMS. Evaluate each origin-destination lane, or a geographic cluster when individual lanes are thin, across a rolling eight- to 13-week window.

Start with shipment frequency and consistency. Count tenders per week, active shipping days, and the coefficient of variation for weekly volume. Rising volume that arrives in one irregular surge does not create the same consolidation opportunity as repeatable freight across several days.

Next, calculate weight and cube compatibility. Track median and 75th-percentile shipment weight, pallet positions, density class, stackability, and handling constraints. Higher weight per shipment is attractive only if cube, dock fit, and commodity rules allow the freight to share capacity efficiently. A dense pallet and an oversized non-stackable unit can have the same weight but very different economics.

Then measure consolidation overlap. For each lane and promised-delivery window, estimate how many orders could be held for 12 or 24 hours and combined without missing customer commitments. Compare the modeled consolidated cost with current LTL spend, including additional handling, storage, accessorials, and service risk.

Finally, test mode-conversion economics. Aggregate compatible shipments by pickup day and destination region. Flag groups approaching a configurable volume threshold—such as pallet positions, linear feet, or weight—then request both volume-LTL and truckload pricing. There should be no universal weight cutoff: freight class, distance, appointment requirements, and market capacity can move the break-even point substantially.

Use a decision rule, not a market hunch

A defensible trigger requires several indicators to agree. Consider changing a lane’s routing logic only when:

• shipment frequency has increased or remained stable for at least four consecutive weeks;
• median weight or cube per shipment supports better equipment utilization;
• at least 70% of candidate shipments share compatible service windows and handling requirements;
• modeled savings remain positive after accessorials and inventory-delay costs; and
• the pattern appears across multiple customers, facilities, or carriers rather than one unusual account.

The 70% threshold is a planning starting point, not an industry benchmark. Each shipper should calibrate it against service failures, freight characteristics, and cost history. The important control is persistence: require the rule to pass for several weeks, and create a reversal trigger if frequency, savings, or on-time performance deteriorates.

Broader indicators can support the test without replacing it. August’s manufacturing PMI was 54.6, marking an eighth month in expansion territory, while the new-orders subindex remained above 50 at 53.7. Those readings suggest an industrial tailwind. But LTL tonnage can lag manufacturing activity by roughly three months, and carrier mix changes can still overwhelm the signal on individual lanes.

What ArcBest’s guidance adds

ArcBest also raised third-quarter adjusted operating-income guidance for its asset-light segment from $6 million–$8 million to $10 million–$12 million. August daily revenue in that segment rose 26% year over year, shipments were flat, and revenue per shipment increased 26%. Purchased transportation declined 60 basis points from July to 85% of revenue.

That improvement points to yield discipline and productivity, not simply more transactions. The company said shipments per person per day were 35% higher year over year in the second quarter. For shippers, the lesson is consistent: volume, price, mix, and productivity must be read separately before operational rules change.

ArcBest’s August tonnage is a meaningful positive data point. It is not proof that every LTL lane is tightening or that every shipper should consolidate more aggressively. The right response is to identify where heavier freight is creating repeatable lane density, validate the service and cost tradeoff, and automate routing changes only where the evidence holds.

Ready to turn shipment data into lane-level routing decisions? Request a CXTMS demo to see how centralized freight visibility, analytics, and configurable workflows can support smarter LTL planning.