AI Playbook for Rising Truckload Linehaul Rates

Freight volumes are still sliding, yet truckload linehaul pricing keeps climbing. That’s not a contradiction—it’s the current truckload market in a nutshell.

The latest data point driving that home: November 2025 shipments were down 7.6% year over year, but the Cass TL Linehaul Index was up 2.2% y/y, and total freight spend was only down 1.2% y/y. Rates are carrying the revenue line while volumes sag. If you run transportation, brokerage, or a shipper’s logistics team, that combination is the kind that quietly destroys budgets—unless you’re forecasting, routing, and buying capacity with more precision than last year.

This post is part of our AI in Transportation & Logistics series, and I’m going to take a clear stance: when pricing rises in a soft-volume environment, manual playbooks fail first. AI-driven forecasting and predictive analytics aren’t “nice to have” anymore—they’re how you avoid paying peak prices for non-peak demand.

What the Cass TL Linehaul Index is really signaling

The simple answer: truckload prices are firming because capacity is getting harder to access at the margin, even while demand remains under pressure.

Cass’ data shows a market where the volume story and the rate story diverge:

• Shipments: -7.6% y/y in November 2025
• Expenditures: -1.2% y/y (nearly flat)
• TL Linehaul Index: +2.2% y/y; third straight sequential increase (September +1.7%, October +1.1%, November +0.1%)

That gap implies higher effective rates. Cass suggested rates were likely up around 7% y/y in November (while also noting it paused inferred rate releases due to mix shifts).

Why this “rates up, volumes down” pattern happens

Three forces tend to create this pattern, and they’re all present right now:

1. Seasonal and weather-driven constraints: Winter storms reduce available capacity quickly. December dislocations can raise spot prices even if the overall year is soft.
2. Regulatory and compliance tightening: Constraints on the driver pool (including enforcement and eligibility rules) can shrink effective capacity without any change in freight demand.
3. Network imbalance: Even in a down market, certain lanes and regions can flip tight when imports, retail replenishment, or weather create pockets of urgency.

Here’s the operational reality: you don’t buy “the national market.” You buy your lanes. And your lanes can behave like a tight market while the headline indices say demand is weak.

Why 2025’s market is punishing outdated planning

Direct answer: the old “annual bid + occasional spot cover” approach can’t keep up when shocks happen monthly—and when the carrier pool itself is shifting.

Cass projected full-year 2025 freight volumes down 6% y/y, with December projected down 4% y/y. The report also highlighted that truckload volumes improved briefly in Q3 ahead of an import tariff deadline, then softened in Q4 as inventories were drawn down.

That kind of demand pattern creates a planning trap:

• You staff and allocate based on Q3 improvements
• Q4 softens, but your service risk increases (weather + compliance + holiday volatility)
• You end up paying spot premiums for what “should have” been easy freight

The hidden budget killer: volatility, not averages

Averages are comforting and dangerous. Your budget doesn’t blow up because the annual rate is up 2%. It blows up because:

• a handful of weeks go sideways,
• a few lanes turn tight,
• and your team buys late.

Predictive analytics for freight pricing is mainly about preventing those late buys.

Where AI actually helps when linehaul rates rise

Direct answer: AI helps by improving decisions you already make—forecasting demand, selecting carriers, routing freight, and timing spot exposure—using better signals and faster refresh cycles.

If you’re trying to connect the Cass TL Linehaul Index trend to practical steps, start here: rising linehaul rates are a signal to tighten your control loop.

1) AI-driven supply chain forecasting: fewer surprises, better buys

A strong forecasting setup doesn’t predict the future perfectly—it reduces error bars enough to buy capacity earlier and more calmly.

What this looks like in transportation planning:

• Forecast shipment counts by lane (not just network totals)
• Detect “mix shifts” (e.g., LTL to TL changes) before they show up as budget variance
• Build scenarios for tariff-driven pull-forward, inventory drawdowns, and retail promotions

Actionable move this month: build a forecast that updates weekly and outputs a lane-level capacity risk score (green/yellow/red) for the next 2–4 weeks.

2) Predictive analytics for the spot market: controlled exposure

When spot rates step up for three straight months, the goal isn’t “avoid spot.” The goal is use spot intentionally.

AI models can support:

• Should-cost estimates by lane and week (what you should pay given current signals)
• Spot-versus-contract recommendations (where spot is cheaper, and where it’s a trap)
• Early warning when rejections and weather imply tightening

Actionable move this month: define a spot policy like an investment policy.

• Maximum spot exposure by lane (e.g., 10–25%)
• Triggers to shift volume (e.g., rejection index moves, weather alerts, tender lead time collapse)
• Approval workflow for “panic buys” above should-cost

3) AI route optimization: defend service without paying peak pricing

When capacity tightens, many teams respond by paying more. Often, you can respond by shipping smarter.

AI route optimization can:

• consolidate orders into fewer, fuller moves,
• recommend mode shifts (TL vs. pool distribution vs. LTL),
• reduce empty miles and deadhead exposure for carriers (which improves acceptance).

The contrarian truth: carriers accept freight they can run efficiently. If you make your freight easier to serve—better appointment flexibility, better packaging, tighter dwell control—you can improve acceptance even in a tighter market.

Actionable move this month: use AI to identify your “dwell-tax” lanes—lanes where detention, loading delays, or appointment constraints correlate with higher paid rates.

4) Warehouse automation + yard visibility: don’t create your own capacity crunch

Rising linehaul rates make internal inefficiency more expensive. If your DC adds 90 minutes of dwell, you’re effectively shrinking capacity on your own freight.

AI in the warehouse and yard can help:

• predict inbound congestion windows,
• prioritize staging for high-risk outbound loads,
• reduce door conflicts and trailer hunting.

This matters because the market is already dealing with capacity constraints from storms and compliance. Don’t stack self-inflicted delays on top.

A practical AI “tight market” checklist for shippers and 3PLs

Direct answer: you don’t need a massive transformation to benefit—you need four measurable controls.

Here’s a field-tested checklist I like because it’s concrete and easy to audit:

1) Build a lane-level forecast you can prove

• Forecast horizon: 2–8 weeks
• Output: volume, expected cost per mile, and confidence bands
• Validation: track forecast error weekly (MAPE or simple % error)

If you can’t measure forecast error, you’re not forecasting—you’re guessing.

2) Create a “should-cost” spine for every key lane

For your top 50–200 lanes:

• establish a baseline expected rate (linehaul-only if possible),
• update weekly using recent wins/losses and market indicators,
• flag loads bought above threshold (e.g., +8% above should-cost).

This is where freight spend management becomes operational, not just financial.

3) Automate exception handling, not everything

Most companies try to automate tendering end-to-end and get burned by edge cases.

Start by automating:

• late tenders,
• repeated rejections,
• high-risk weather corridors,
• loads with low lead time.

Then route those exceptions to a human with a recommended fix (carrier swap, time window adjustment, consolidation option).

4) Turn compliance and eligibility shifts into a carrier strategy

If compliance enforcement is tightening the available driver pool, your carrier strategy must adapt:

• diversify primary carriers on tight lanes,
• expand backup carriers before disruption hits,
• track acceptance by lane/day/time, not just monthly scorecards.

A modern transportation management system with AI should show you where acceptance is decaying and why, not just that it happened.

People keep asking: if volumes are down, why are rates up?

Direct answer: because capacity isn’t a single number—it’s availability at the right place, time, and compliance conditions.

In November, shipments fell sharply year over year, yet:

• winter storms constrained spot capacity,
• holiday demand held up better than expected,
• driver pool regulations and enforcement reduced supply,
• tender rejections signaled tightening.

So even a soft demand environment can produce local tightness and higher linehaul pricing.

The next step: treat indices as signals, then run your network like a control system

The Cass TL Linehaul Index extending its positive run is a message: pricing power is flickering back to carriers in short bursts, even as the broader demand picture remains soft.

If you’re still managing routing, carrier selection, and spot buying mostly through spreadsheets and tribal knowledge, you’ll keep paying for those bursts—especially through winter and into early 2026 when tariff and affordability pressures could keep demand choppy.

If you want a practical place to start, pick one: lane-level forecasting, spot should-cost, or dwell and appointment optimization. Then instrument it, measure it weekly, and expand. That’s how AI becomes a working advantage in transportation logistics, not a slide deck.

What’s one lane in your network that always seems to “mysteriously” get expensive in December—and do you have the data to explain it?