What a should-cost is — and what it isn't
A should-cost is the bottom-up rebuilt cost of moving a load on a specific lane: how much diesel the truck burns on that terrain at that country's price, what the driver costs for the real hours of driving and waiting, which tolls apply plaza by plaza, how much maintenance and depreciation weigh per kilometer, which insurance applies, how many hours are lost at the port or the border and — the line almost nobody quotes — what happens to the truck when it comes back empty.
It's important to be honest about what that number is: a cost floor, with no carrier margin. A market rate is cost + margin + lane imbalance + risk. So the should-cost doesn't tell you what price to pay; it tells you where to negotiate from, and how much of a quote is legitimate margin versus padding. When we audited a global ocean carrier's 10 Central American corridors against that floor, the average gap between what was paid and what it cost was ~12% — spread across lines no flat quote shows.
The 7 lines of any full-truckload quote
A per-kilometer rate hides the structure. These are the lines that make it up, with the weight they carry on real corridors:
1. Fuel. The largest line on most corridors — 20% to 30% of the rate, and #1 in Guatemala (28%), Nicaragua (29%), El Salvador (27%) and Colombia (29%). It depends on real consumption by terrain (a loaded truck climbing doesn't get flat-road mileage) and on the country's diesel price, which varies more than 2× across the region.
2. Driver. Hourly wage for driving and waiting hours, plus per-diem and overnight. This is where things get double-counted most when no one is looking (see below).
3. Tolls. Plaza by plaza, by axle class, outbound and on the empty return. A per-kilometer average is wrong in both directions.
4. Maintenance and depreciation. Per kilometer, by truck class and country. Usually fine; rarely where margin hides.
5. Insurance. Unit and cargo, by class.
6. Dwell and detention. Hours at the port or at loading and unloading. The most volatile line on a port corridor, and the one a flat price reflects worst.
7. Empty return. If the truck can't find a return load, that trip is amortized into your outbound rate. On Bogotá → Medellín the difference between no backhaul and a secured one is US$985 vs US$550: 44% of the rate depends on an assumption almost no quote states. We break it down here.
The 7 places a quote hides margin
This isn't theory. These are mechanisms we found auditing a cost engine against real corridors — each with the amount it moved. They're exactly what a buyer should check in any quote:
Per-diem billed twice. A driver per-diem and a statutory per-diem, both added to the subtotal with no lodging/meals split. In Honduras: US$25 + US$17 for the same day.
Overnight rest paid at full driving wage. A 16-hour mandatory rest billed at the driving rate: on a 10-hour trip, roughly US$98 of overnight, more than the leg's driving wage. Rest is standby, not driving.
Border idle fuel counted twice. Waiting-time idle charged on the fuel line and on the border-delay line. On a Guatemala → Panama lane: ~US$56 duplicated.
The empty return as the single largest line. With no explicit assumption, a model assumes a full-corridor reposition. On 200 km in Costa Rica, the empty return (US$178) exceeded fuel (US$103). It isn't wrong — it has to be declared.
Tolls over-summed. Every gantry of one concession charged separately instead of one entry and one exit. Panama → Colón: US$19.16 quoted against US$13.90 documented, +38%.
Empty-return tolls charged but never itemized. Bogotá → Medellín: US$113.72 — about 12% of the rate — inside the total without appearing on any line. You pay for something you can't see.
A line that appears but isn't charged. A US$20 deadhead shown in the breakdown and excluded from the total. Less serious, but the breakdown no longer reconciles — and if the breakdown doesn't reconcile to the total, it isn't a breakdown.
The practical rule that falls out of all this: demand that the breakdown sums exactly to the total, that every line says where the data comes from, and that the return assumption is written down.
The legal floor: when cost doesn't set the price
In some markets the price has a regulated floor. Honduras sets minimum cargo-transport tariffs (through the IHTT): on those corridors the rate is the greater of your operating cost and the legal floor. When operating costs fall — with drop-and-hook, say — the gap to the floor widens (on Puerto Cortés → San Pedro Sula, over 50%) and the rate stays up. That legal minimum includes a temporary fuel surcharge of US$0.22/km on top of the base tariff (about US$55 on 250 km); it is not a charge added on top of the operating cost.
For procurement this is a compliance guardrail, not just a price one: a should-cost that shows the floor next to the cost tells you whether a bid is illegal, not just whether it's cheap. No generic per-kilometer rate gives you that.
How procurement teams use it
What a procurement team needs from a should-cost is that it be auditable line by line, with the source of each figure, its confidence level and its refresh date. That is the negotiation-grade standard: with it you can walk into a carrier meeting and defend every peso. That's the primary use: not to replace the quote, but to have the number to read it against.
The concrete uses we see: (a) compare every bid in an RFP against the lane's floor, line by line, instead of against the second-lowest — here's how to evaluate an RFP; (b) defend the freight budget with a rebuilt cost, not last year's history; (c) win the cross-border or cold-chain argument, where generic rates don't reach; (d) annualize overpayment per lane to prioritize what to renegotiate first; (e) report CO₂ per shipment for Scope 3.
One idea that came out of validation with a global ocean carrier: use the should-cost as an "invited quoter" in a live spot process. Because the floor carries no margin, every market quote instantly reads as % above cost. It's the fastest way to see which carrier is charging padding.
How the numbers are validated
A should-cost is worth exactly what its data is worth. Ours refreshes weekly — diesel by country, exchange rates, tolls plaza by plaza and port dwell times — and every line cites its source with a confidence level. The corridors were validated one by one with the operations teams of a global ocean carrier: where the kilometers matched, the amounts came out "very accurate," and the Honduran corridors "well grounded in reality." When a team found an incorrect distance, we fixed it and documented it. That's the standard: the number has to survive the people who move the freight every day.
Where to start
You don't need to audit 500 lanes to know whether you're overpaying. Start with your 10 highest-spend lanes: that's where most of the budget and most of the hidden margin sit. Rebuild their should-cost, compare each current quote line by line, and annualize the gap. That gives you the prioritized list of what to renegotiate first — and the argument to do it. For large tenders, the same engine processes up to 100 lanes per batch from a file.
Before you approve a quote
- Does the breakdown sum exactly to the total? If not, there's a hidden line.
- Does every line say where the figure comes from — diesel price, toll plaza, wage? A number without a source can't be audited.
- Is the return assumption written down — no backhaul, partial, or secured? It's 44% of the rate.
- Are dwell and border separated from the linehaul? They're the most volatile lines; blended, they hide the total.
- Is the quote above the country's legal floor? If not, price isn't the problem.
The should-cost doesn't tell you what price to pay. It tells you where to negotiate from — and how much of the quote is legitimate margin versus padding.
Frequently asked questions
Keep reading
See the method applied to your own lanes
Upload up to 10 lanes with your current quotes. We rebuild each one's should-cost with the engine and send you an overpayment report with real numbers — at no cost.
We audited a global ocean carrier's 10 Central American corridors and found ~12% average overpayment.