Truckload Rates Are Rising—Use AI to Stay Ahead

Truckload pricing is doing something that’s been easy to forget over the last couple of years: it’s climbing even while shipments fall. In November, the Cass multimodal shipments index was down 7.6% year over year, yet the Cass truckload (TL) linehaul index still posted another month of gains—its third straight sequential increase—and remained up 2.2% year over year.

That mismatch is the signal. When rates strengthen while volume stays soft, the market is telling you capacity is getting tighter, volatility is rising, and your old “set it and forget it” procurement and routing playbook won’t hold.

For this edition of our “AI in Trucking & Freight: Fleet Intelligence” series, I’m using the Cass TL Linehaul Index’s positive run as a case study. Not to repeat the report—but to translate it into practical moves: how AI-driven freight forecasting, routing, and capacity planning help shippers, carriers, and 3PLs protect margin when the market shifts under their feet.

What the Cass TL Linehaul Index is really saying

Answer first: The index’s steady increase says the truckload market is tightening in pockets—even if headline volumes look weak.

Cass reported that shipment counts remained under pressure in November. Meanwhile, total freight expenditures were close to flat versus a year ago (only slightly down), which implies **rates are doing more of the “work” in total spend than volume is. Cass also noted that when you net volume changes from expenditure changes, rates were likely up around 7% year over year in November (with the usual caveat that freight mix is shifting).

Three drivers matter most for fleet and transportation leaders:

1) Peak season didn’t fizzle

Holiday spending surpassed low expectations, which supports demand for truck capacity in the near term. Even if overall industrial demand is sluggish, consumer-driven surges still create short, sharp capacity crunches.

2) Weather is a capacity event, not just a delay

Cass pointed to winter storms consistent with a La Niña pattern affecting spot capacity and pushing spot rates higher. The operational lesson is blunt: weather is effectively a temporary reduction in supply, and pricing reacts accordingly.

3) Driver-pool constraints are showing up in spot behavior

The report referenced tighter conditions tied to increased regulation and enforcement (e.g., English-language proficiency requirements, non-domiciled CDL restrictions, ELD and driver school crackdowns). Whether you agree with any individual policy or not, the business reality is straightforward: if the eligible driver pool shrinks or gets harder to activate, spot capacity tightens first.

If you manage transportation, this is where “market intelligence” stops being a nice-to-have. It becomes a daily control system.

The hidden risk: soft volumes can still produce expensive freight

Answer first: The most expensive freight environment is often one where demand is uneven, not uniformly strong.

A common mistake I see is teams treating “volumes down” as shorthand for “rates down.” November’s Cass data is a clean rebuttal: shipments fell, but linehaul pricing kept improving. That’s not unusual late in the cycle because:

• Volume declines aren’t uniform by lane or region. Some networks drop 10%, others spike due to promos, port dynamics, or replenishment.
• Carrier behavior changes faster than shipper contracts. If carriers sense a spot rebound, they protect trucks for higher-yield loads.
• Disruptions (storms, enforcement, holidays) create micro-markets. You don’t buy “the national market.” You buy your lanes.

This matters because your cost outcome is driven by variance: missed appointments, last-minute mode changes, unexpected dwell, and broker premium loads.

If you’re running quarterly bid cycles and static routing guides, you’re basically driving using the rearview mirror.

Where AI forecasting actually fits (and where it doesn’t)

Answer first: AI forecasting wins when it predicts lane-level volatility and converts that into actions—tenders, prebooks, and smarter routing—not when it produces a pretty dashboard.

The Cass report gives you directionally useful indicators: shipments down 7.6% y/y, the TL linehaul index up 2.2% y/y, and a market briefly tipping toward fleets in December due to holiday spending and weather capacity constraints. That’s the macro story. But your transportation budget is won and lost in the micro story.

Here’s how AI-driven freight forecasting turns macro signals into lane decisions.

Build a “rate + capacity + disruption” forecast, not a demand forecast alone

Most demand planning feeds transportation a single number: “shipments next week.” That’s not enough.

A practical AI forecasting stack for trucking and freight should combine:

• Internal order signals: booked orders, promo calendars, DC inventory position, customer OTIF penalties
• Transportation execution data: tender acceptance/rejection history by lane, dwell time, lead time to pickup
• Market indicators: spot linehaul trend, rejection index trend, load-to-truck balance proxies
• Disruption signals: weather forecasts, holiday calendars, major event schedules

The output shouldn’t be a single forecast. It should be a set of probabilities:

• Probability of needing spot coverage by lane
• Probability of service failure at current lead time
• Expected rate range (not a single point)

Use AI to identify “rate inflection lanes” early

When spot turns, it rarely turns everywhere at once. AI models can flag lanes where:

• tender rejections are rising faster than your network average
• lead times are compressing
• dwell and appointment slip is trending up

Those are the lanes where you should prebook capacity, shift pickup windows, or temporarily rebalance to alternate carriers.

Don’t overfit: keep humans in the loop for policy shocks

Regulatory enforcement and driver-pool constraints can create structural breaks. Models trained on “normal” years often underreact.

The fix is not more complexity—it’s governance:

• add manual override rules (e.g., “storm declared,” “holiday surge,” “policy enforcement spike”)
• retrain frequently (weekly or biweekly during peak volatility)
• measure forecast quality by lane, not just network averages

Fleet intelligence: turning price momentum into margin

Answer first: When truckload rates rise, AI helps you defend margin by reducing empty miles, improving tender outcomes, and pricing volatility into decisions.

Whether you’re a shipper, a carrier, or a 3PL, the operational moves look different—but the objective is the same: make fewer reactive decisions.

For shippers: make routing guides adaptive

Static routing guides assume stable acceptance behavior. During tightening conditions, that breaks.

An AI-assisted routing guide can:

• recommend dynamic carrier splits by lane (e.g., 60/30/10 becomes 45/35/20)
• adjust tender lead times based on carrier responsiveness
• trigger spot bids earlier when the probability of rejection crosses a threshold

A simple rule that works well: if a lane’s rejection probability rises for three consecutive days, switch from “post-tender salvage” to “pre-tender coverage.” It’s cheaper and less chaotic.

For carriers: forecast where your network will break first

Carriers feel tightening markets as a chance to improve yield—but only if execution stays clean.

AI can support:

• empty-mile reduction via load sequencing recommendations
• network balance alerts (where inbound/outbound mismatch is growing)
• driver hours utilization predictions to prevent late-week failures

If winter storms are shrinking capacity, the carrier who wins isn’t the one who says “yes” most often. It’s the one who says “yes” on the loads they can actually execute.

For 3PLs and brokers: price volatility explicitly

When spot is climbing, underpricing one bad commitment can erase a week of margin.

AI-guided pricing should:

• estimate a rate distribution (P10/P50/P90)
• apply a volatility buffer when disruption signals are active
• incorporate carrier response likelihood (price is meaningless if you can’t cover)

A broker with strong models doesn’t just quote faster. They quote more honestly—and protect service.

A practical 30-day playbook for a tightening TL market

Answer first: The fastest wins come from better inputs, faster feedback loops, and automated exception handling.

If you want to act on what the Cass TL Linehaul Index is indicating—without boiling the ocean—here’s a focused month-long plan.

1. Pick 25 lanes that drive 60–80% of your spend. Don’t start with the long tail.
2. Create a daily “lane health” score using five signals:
   • tender acceptance rate
   • lead time to tender
   • dwell time
   • on-time pickup
   • spot rate trend for that lane or region
3. Automate two actions:
   • early spot bid request when lane health falls below threshold
   • appointment window expansion recommendation when storms hit a region
4. Run a weekly forecast review with transportation + ops:
   • what lanes were wrong?
   • what signals predicted failure earliest?
   • what rule would have prevented the premium load?
5. Measure outcomes in dollars, not vibes:
   • cost per mile (linehaul vs fuel separated if possible)
   • % loads covered at first tender
   • premium freight as % of total spend

If you do only one thing: shorten the time between signal → decision → result. AI is valuable because it compresses that loop.

People also ask: what should I watch besides the Cass index?

Answer first: Watch indicators that move before your freight bill moves.

• Tender rejections by equipment type (van, reefer, flatbed) to anticipate tightening
• Lead time to pickup as a proxy for shipper leverage
• Dwell time by facility because detention becomes rate pressure
• Spot linehaul trend (excluding fuel) to avoid confusing fuel swings with true price moves

Rates are the outcome. The signals above are the causes.

What to do next if you want AI-driven freight decisions in 2026

The Cass data hints at a market where tariffs, affordability pressure, and uneven demand can coexist with periodic capacity shocks (weather, regulation, holiday surges). That combination produces one predictable result: more variance.

If you’re serious about fleet intelligence, focus on systems that answer three questions every day:

1. Where will capacity break first?
2. What’s the cheapest action that prevents a service failure?
3. How do we learn from last week fast enough to matter next week?

Truckload rates rising isn’t the story by itself. The story is whether your operation can see the inflection early and act while options are still cheap.

If your freight network had to absorb another storm-driven capacity crunch next week, which 10 lanes would cost you the most—and would you know that before the tenders start failing?