TL;DR
- A TMS is rate-shop, tender, track, audit. Anything beyond that is convenience features.
- Parcel makes sense up to about 150 pounds. LTL takes over from there to roughly 12 pallets. Past that, look at partial truckload.
- Multi-modal routing only saves money when you measure landed cost, not just linehaul.
- Real-time tracking is worth the investment when it changes operational behavior. If nobody acts on the alert, the alert is decoration.
What a transportation management system actually does.
A transportation management system is the operating layer between an order and a moving truck. It rate-shops carriers, tenders the load, generates documents, tracks the shipment, audits the invoice, and feeds the data back into planning for the next cycle.[1] Every TMS in the Gartner Magic Quadrant does some version of those five things. The differences come down to how well each one is executed and how cleanly the system integrates with the warehouse and the order management layer.
The TMS market is mature. The 2025 Gartner Magic Quadrant covers more than a dozen vendors, with Manhattan named a Leader for the seventh time and Oracle holding more than 19% market share.[1] Vendors increasingly differentiate on AI-driven decision support, with generative and agentic AI features showing up in execution, exception handling, and procurement workflows.[2]
For a 3PL like Warpspeed the calculation is different from an enterprise shipper buying its own TMS. The right system is one that runs the warehouse-to-carrier handoff cleanly, keeps lane data clean enough to negotiate annually, and exposes performance metrics that customers can act on. ShipOS, the Warpspeed transportation layer, was built for exactly that. It is integrated with the WMS, so a tender uses real shipping weights and dimensions rather than catalog estimates, and exception alerts route to operators with the context to do something about them.
“The best TMS is the one your operators trust enough to follow. A perfect rate-shop nobody uses is worth less than a 90% rate-shop everybody runs.”
Mode selection: getting the answer right before the carrier shop.
Mode selection is the most consequential decision in transportation. A 600 pound shipment that ships as four parcels because nobody checked the math will cost roughly twice what the same shipment palletized and tendered as LTL would cost. A 14,000 pound shipment that tenders as standard LTL gets refused or surcharged by most carriers, who treat anything past 12 pallets as a partial truckload routing problem.
The math itself is straightforward. Below 150 pounds, parcel almost always wins. Between 150 and 500 pounds, run both quotes and let the system pick.[4] From 500 pounds up to about 12 standard pallets, LTL is usually right. Past that, partial truckload, volume LTL, or full truckload depending on lane and timing.
Mode break points
| Weight band | Default mode | Notes |
|---|---|---|
| Under 150 lb | Parcel | Watch dim weight on bulky cartons |
| 150 to 500 lb | Parcel or LTL | Always rate-shop both |
| 500 to 6,000 lb | LTL | Class and accessorials drive cost |
| 6,000 to 12,000 lb | Volume LTL or partial TL | Volume LTL often cheaper than 12 pallets standard LTL |
| 12,000+ lb or 12+ pallets | Full truckload | Spot or contract depending on lane |
Multi-modal routing when distance and time line up.
Multi-modal routing means combining modes inside one move. A container from Long Beach to a Pennsylvania warehouse might run drayage to a rail ramp, intermodal rail to a Chicago ramp, and LTL or truckload from there. A 2,500 pound shipment moving from California to Florida might run regional LTL to a consolidation center, then ride a long-haul partial truckload, and split back out for last-mile delivery. The pattern only saves money when transit time tolerance allows it.
Intermodal rail typically saves 20% to 30% per loaded mile against over-the-road truckload on long lanes, with a transit penalty of two to four extra days. The savings grow as fuel prices rise. Drayage on each end runs $250 to $800 per container depending on origin port, distance, and accessorials.[5] For freight where two extra days of transit does not break the customer promise, intermodal is the default for cross-country moves.
The tricky part of multi-modal is landed cost. The line haul rate sheet looks great. The chassis fees, terminal congestion charges, and per-diem on a container that did not get unloaded fast enough do not. A TMS that exposes total landed cost (linehaul plus all accessorials plus dwell costs) gives operators a real number to compare. A TMS that only shows linehaul invites bad decisions.
Real-time tracking that changes behavior.
Real-time tracking is in every TMS pitch. The honest version is more complicated. Carrier EDI 214 status messages have always existed and have always been somewhat unreliable. ELD-integrated GPS tracking is more accurate but only available on lanes where carriers agree to share the data. Some carriers do; some do not. A working tracking program triages which loads are worth tracking minute by minute (high-value, time-sensitive, critical retailer appointment) and which loads are fine on standard EDI updates.
The value of tracking is not visibility, it is action. An alert that nobody acts on is decoration. A program that flags a missed pickup window, routes the alert to a dispatcher who can call the carrier, and triggers a backup tender if the original carrier cannot recover, that is operationally useful. Building those workflows is the harder half of the job.
Procurement, contract bids, and the spot market.
Carrier procurement runs on two parallel tracks. Contract rates are negotiated annually (sometimes more often) for predictable lane volume. Spot rates fill in the gaps for one-off moves and surge demand. Spot moves first, because it reflects today's capacity. Contract lags, because it is locked until the next bid cycle.[8] The gap between the two is real money in either direction.
Cass Information Systems data showed freight market tightening through Q1 2025 as shipments and expenditures both rose, with carriers more selective about which loads to accept.[6] When the market tightens, contract rates fall behind spot. When the market softens, contract rates run above spot. The discipline is to keep enough lane data to renegotiate when the gap gets meaningful, and to avoid bidding lanes you cannot deliver on.
The 2025 NMFC reform on the LTL side complicates procurement. Most commodities now classify on a 13-tier density scale, so historical rate cards may no longer apply.[10] Shippers re-bidding LTL contracts after July 2025 should re-rate their top SKUs against the new scale before signing.
How a tender becomes a delivery, step by step.
- 1
Order routed
The WMS releases an order. The TMS pulls real shipping weight and dimensions, applies any customer routing rules, and rate-shops carriers across modes.
- 2
Carrier tendered
The chosen carrier receives an electronic tender. They accept, decline, or counter. A backup carrier is on standby for declines.
- 3
Documents generated
Bill of lading, shipping label, and any customer-required paperwork (PO, packing list, ASN) generate automatically and stage with the load.
- 4
Pickup executed
Driver arrives in the window, freight loads, BOL signs. EDI 204 (tender) and 990 (acceptance) confirm the move is in motion.
- 5
In-transit tracking
Status updates from the carrier (EDI 214) and ELD-integrated GPS where available feed the tracking layer. Exceptions route to a dispatcher.
- 6
Delivery and POD
Carrier delivers, POD captures, and the appointment is closed. Any damage or shortage gets noted on the delivery receipt.
- 7
Invoice audit
Invoice arrives, the TMS audits against the original quote, and discrepancies (reweigh, reclass, accessorials) get reviewed before payment.
How Warpspeed runs ShipOS inside the network.
ShipOS is the transportation layer Warpspeed built on top of its WMS. Because the two systems share data, the rate shop uses real cartonization output instead of catalog assumptions. Mode selection runs against current carrier rate cards by lane. The tender workflow handles parcel, LTL, FTL, drayage, and intermodal in the same operating model, so an order that should ship LTL does not get split into parcels by accident, and a container that should ride intermodal does not get tendered as cross-country truckload.
The other half of ShipOS is exception handling. Missed pickups, late arrivals, damaged freight, and reweigh charges all flow into one work queue. Operators see the whole picture, not a fragment, and customers see a single status feed instead of a tab for each carrier. That integration is what lets the transportation function run on the same cadence as the warehouse instead of running parallel and out of sync.
“The point of a TMS is not to track freight, it is to make better decisions every time an order moves.”
The cost levers worth pulling first.
Most transportation savings come from a small set of moves. Mode optimization at tender time. Cartonization that reduces dim weight on parcel. Accessorial discipline at the BOL stage so freight does not get reclassed in transit. Carrier selection by lane density instead of headline rate. Annual bid cycles informed by the prior year's actual lane data, not estimates.
Less obvious but equally valuable: facility location. Kansas City is one of the largest inland freight hubs in the country, with six of seven Class I railroads serving the metro and 85% of the US population reachable by truck within two days.[9] Shippers running national distribution from a Kansas City node tap into cheaper, denser carrier networks than shippers running the same volume from a coastal warehouse. The TMS records the savings; the warehouse footprint earns them.
Want a TMS that runs the whole tender to delivery loop.
ShipOS is integrated with the warehouse, the carriers, and the customer-facing tracking. Talk to our team about how it fits your network.