How to Quote Regional Carriers That Have No API

The usual explanation — "small carriers are not digital" — is condescending and wrong. Regional hauliers run telematics, digital tachographs, ePOD apps and an ERP. They are not afraid of software. They decline to build a rating API because, priced honestly, it is a bad investment for them.

Price the project the way that haulier's managing director would. The figures below are an illustrative build, not a market survey — rebuild it with your own local day rate and the shape will hold.

That is 43 developer-days to launch. At €600 per day, €25,800 one-off, plus roughly 14 days a year, or €8,400, to keep it alive. A haulier moving 30 loads a day over 230 working days handles 6,900 loads a year, so year one costs about €4.96 per load — and only a fraction of those loads come from brokers who would ever call the endpoint, so the real cost per API-priced load is several times higher.

Three non-financial reasons finish the argument. There is no product owner: nobody in a 30-truck company owns an API roadmap, so a breaking change has no scheduled fix. Tariffs are customer-specific and confidential — the commercial director negotiates a different grid with each broker, and an endpoint that returns a price is an endpoint that leaks the negotiation. And the binding constraint is capacity, not price: the honest answer to "what does this cost?" is often "€395 if I have a trailer going that way Thursday", which no rating schema expresses. Cargavo, a TMS for freight brokers and transport commissionaires, is built on the assumption that this carrier never changes its mind — see the carrier onboarding category.

This is the failure nobody writes about, because it only shows up on the invoice. A rating call is authenticated by an account number, and the price it returns is read from whatever rate table the carrier's ERP has attached to that account. Three outcomes are common:

• Published tariff. The endpoint answers, fast and correctly, with the carrier's list price. Your negotiated grid is 12 % to 30 % below it and lives in a different table entirely.
• Partially provisioned account. A blanket discount was applied to the account, but the bracket-level exceptions you negotiated — the two lanes where you got a better rate for committing volume — were never loaded.
• True contract table. It matches your invoices to the cent. This happens, and it is worth having; it is just not the default.

There is a second layer. Rating responses frequently carry line haul only. The fuel surcharge is applied at invoicing from the month's index, the tail lift (liftgate) and waiting time are added by the depot, and the ADR supplement comes from a separate schedule. So the "price" the API gave you is a floor, and your invoice grows after you have already committed a sell price to your customer.

The test costs an afternoon. Pull three shipments you have already been invoiced for, covering a light one, a heavy one and one with at least two accessorials. Call the endpoint with the exact same inputs. Compare each response to the invoice line by line. If they do not match to the cent, the endpoint is not your contract — it is a quotation from a stranger. A stored rate grid has the opposite property: it contains the numbers you negotiated because you are the one who put them there, and you can read every bracket at once instead of one call at a time. The three integration methods are compared in full in API vs EDI vs rate grid.

Fleet size is the wrong predictor and it sends brokers looking in the wrong place. A 400-vehicle bulk tanker operator has no rating API; a 60-van same-day courier network may well have one. The variable that decides it is whether the carrier sells to strangers online. If a person who has never signed anything can get a price and pay by card, the carrier needs machine-readable pricing. If every price is the outcome of an annual negotiation, it does not.

Read the profiles, not the row count — this table describes categories of carrier, not the composition of any particular panel. Build the same table for your own carrier list and you will find the API column empty exactly where your best home-lane prices are, because the regional haulier that is cheapest between its two depots is quoted by fewer brokers than the national network everyone can reach.

That is the commercial reason to stop treating no-API carriers as a second tier. A panel where 2 carriers are integrated and 12 are priced by hand quotes slowly on the 12 lanes where the margin is best. Cargavo, a TMS for freight brokers of 2 to 25 people, makes the last five rows of that table the default case: each one becomes one of four grid shapes, described in the four types of road freight rate grids.

Brokers digitise the rate table and stop, because the rate table looks like the tariff. It is roughly half of it. The two parts that get skipped are where quotes go wrong six weeks later, on an invoice.

• The rate table — brackets down one axis, zones across the other. This is the part that photographs well and the part every spreadsheet already contains.
• The conditions page — minimum charge, chargeable-weight rule, rounding, validity dates, currency, cut-off time, maximum consignment. Usually prose, often a footnote, almost never a table.
• The surcharge annex — fuel indexation basis, tail lift, ADR, timed delivery, waiting time, ad valorem insurance, redelivery.

A PDF tariff is retyped and an Excel tariff is copied, and either way the work is bounded, because a tariff is a document of a few pages rather than a system with a release schedule. Structured entry has a side effect worth paying for: a rate that does not fit any of the four grid shapes surfaces at entry time, while the carrier's sales rep is still reachable, instead of on an invoice. Cargavo, a TMS for freight brokers, supports weight_zone, pallet, mpl_bracket and hybrid_weight grids with currency and validity dates on each one. For the field-by-field capture list see the carrier onboarding checklist, for the spreadsheet route see rebuilding an Excel rate sheet, and for the LDM arithmetic use the free loading metre calculator.

Here is the whole calculation, from a PDF tariff to a price a customer can accept, with no integration anywhere in it. Every figure below is an example built from the tariff described in the previous section.

The carrier. A regional part-load haulier, no rating API, tariff received as a PDF valid 1 January 2026 to 31 December 2026, digitised as an mpl_bracket grid in EUR.

The shipment. Six non-stackable EUR pallets, 1.20 × 0.80 m each, 2,100 kg total, zone 3, tail lift required at delivery, one pallet carrying ADR limited-quantity goods.

1. Floor space: 6 × (1.20 × 0.80) = 5.76 m². Divided by the 2.40 m billing width, that is 2.40 LDM.
2. Bracket lookup: 2.40 LDM falls in the 2.01–2.50 LDM row, zone 3 — €395.00. Above the €145.00 minimum charge, so the minimum does not bite.
3. Chargeable weight cross-check: 2.40 LDM × 1,850 kg = 4,440 kg against an actual 2,100 kg. The grid prices by LDM here, so the floor changes nothing — but on a weight_zone carrier the same pallets would be charged as 4,440 kg, not 2,100 kg.
4. Fuel surcharge: index €1.71 per litre against the €1.42 base peg, fuel weighted at 32 %. (1.71 − 1.42) ÷ 1.42 × 0.32 = 6.54 %, applied to line haul: €395.00 × 0.0654 = €25.83.
5. Options: tail lift €35.00, ADR limited quantity €48.00.
6. Buy price: 395.00 + 25.83 + 35.00 + 48.00 = €503.83.
7. Sell price at a 16 % margin, using p ÷ (1 − m): 503.83 ÷ 0.84 = €599.80. Margin earned: €95.97, which is 16.00 % of the sell price.

Two things are worth noticing. The margin formula is p ÷ (1 − m), not buy × 1.16 — that would have produced €584.44 and a 13.79 % margin, €15.36 short on one consignment. And the entire calculation depends on numbers that live in the tariff document: the bracket, the minimum, the LDM floor, the peg and the weighting. Cargavo, a TMS for freight brokers, recomputes this server-side on every quote so the sell price cannot be edited into a loss, and the buy price and the margin stay in a private record the customer's portal never reads. Related reading: chargeable weight and showing a price without revealing your buy rate.

Specialised carriers are further from an API than regional hauliers are, and for a reason that is structural rather than commercial: their price depends on conditions a generic rating request has no field for. No standard rating schema carries an ADR class and packing group, a tunnel restriction code, a set-point temperature, whether a trailer runs mono- or multi-temperature, whether empty pallets return, or the crane reach a delivery site needs. A rating endpoint that cannot receive the question cannot answer it.

The practical split is between what is tabulated and what is assessed:

• Tabulated — supplements the carrier has already written down: ADR limited quantity at €48.00, full ADR at 12 % of line haul, a frozen band at €0.09 per kg, a timed delivery window at €55.00, waiting time at €45.00 per hour after 60 minutes. These are options on the grid, and they belong in the quote automatically.
• Assessed — prices that require a human to look at the job: abnormal loads needing route surveys, permits and escorts; crane-equipped deliveries; anything where the answer depends on what else is on the trailer that week.

The mistake is letting the second category fall out of the system. A quote request that arrives for an oversize load and gets answered from someone's mailbox leaves no trace: no reference, no record of what was quoted, no margin captured, nothing to reuse when the same customer asks again in March. In Cargavo, a TMS for freight brokers and transport commissionaires, a request that no grid covers is still accepted, still numbered Q-YYYY-NNNN, and still priced by a human inside the same pipeline — then converts to a tracked shipment S-YYYY-NNNN on acceptance. The full off-grid path is covered in handling a quote request with no rate for the lane.

One habit pays for itself with specialists: record their operating envelope next to the grid — zones actually served, maximum consignment weight, ADR classes carried, cut-off time, whether they take non-stackable freight. Half the manual quotes a broker sends are sent to carriers that were never going to accept the job.

An honest guide states the cost of its own recommendation. Quoting from stored grids gives up real things, and a broker should know which ones before choosing.

• No live capacity check. The grid tells you the price, not whether the carrier has a trailer on Thursday. That call still gets made — but you make it after the customer has accepted, not before you can quote.
• No carrier-committed transit time. Transit is a figure you stored, not one the carrier confirmed for that consignment.
• No automatic consignment number or label. Booking with the carrier stays a message or a portal entry on their side.
• No automated status feed. Milestones are recorded by you, from what the carrier tells you.
• No automated invoice reconciliation against a carrier system. You compare the invoice to the stored buy price, which is a real check, but it is yours to run.

What you keep is the pricing, and pricing is what a quote is. You also get two properties an endpoint cannot give you: every bracket is inspectable, so you can sanity-check a whole lane from 100 kg to 3,000 kg in one screen, and the quote is offline-proof, because no third-party outage sits between a customer's request and a price.

State the boundary plainly. Cargavo, a TMS for freight brokers, has no carrier API and no EDI connectivity, and is not a freight exchange, not a load board, not a fleet or dispatch system, not a procurement and tendering marketplace, and not accounting software. It prices from grids you rebuild and own, in EUR, USD, GBP, CHF, CAD, MAD or PLN, with an interface in English, French, Spanish and German. Carrier onboarding is included in every plan: Starter €49/month, Growth €149/month, Scale €399/month, annual billing worth about two months free, 14-day free trial and no credit card — limits on the pricing page. Next step: onboarding a carrier that has no API.