How to Run a Freight RFQ and Tender Process

Most companies discover they need a freight tender the same way: volume has grown to the point where occasional spot bookings have turned into a steady flow across several lanes, and someone finally asks why nobody has compared forwarders side by side. The instinct at that point is almost always the same — send the current forwarder's rate sheet, and two or three others, a request for pricing on the same lanes, and pick whichever number comes back lowest.

That instinct is what produces a tender that fails before it starts. A freight tender is fundamentally a specification exercise: the quality of what comes back is set entirely by the precision of what was asked for. Ask five forwarders for "a rate to Bangkok" without defining container mix, service scope, or what "all-in" is supposed to include, and five different assumptions come back dressed up as five comparable numbers. None of them are actually comparable, because none of them priced the same thing.

A buyer running a tender for the first time usually has the discipline for this in every other category they purchase — nobody asks five printing vendors for "a quote for some brochures" and expects the numbers to mean anything. Freight gets treated differently, partly because terms like FOB, all-in, and per-container sound precise even when the request behind them isn't, and partly because forwarders are used to filling gaps with their own assumptions rather than pushing back to ask for the missing detail.

This article works through what a tender needs to contain to produce prices that actually mean something: the difference between an RFI, an RFQ, and an RFP; the volume and scope data that has to exist before a single price is requested; how to structure rate requests so responses are genuinely comparable; how to evaluate what comes back on more than price alone; and the award, implementation, and ongoing-management work that decides whether an agreed rate ever actually gets used.

RFI, RFQ, and RFP are three different questions

The three acronyms get used almost interchangeably in freight, and that looseness is where a lot of tenders go wrong before the first bid ever comes in.

An RFI — request for information — is a screening step. It's sent to a wider list of forwarders than you intend to actually use, asking broad questions about network coverage, systems capability, relevant experience, and financial standing, with no expectation of a firm price. Its job is to narrow a long list down to a shortlist worth spending real evaluation time on.

An RFQ — request for quotation — prices a scope that's already been defined. It works when the buyer already knows exactly what's being bought: specific lanes, specific volumes, specific service inclusions, and the only open question is the number. An RFQ sent without that definition doesn't function as an RFQ at all — it functions as an unstructured request for a number, which is exactly what produces five incomparable answers.

An RFP — request for proposal — asks a forwarder to design part of the solution, not just price it. This is the right instrument when the buyer knows the outcome it wants (consolidate shipments from multiple China suppliers into fewer, fuller containers, for example) but not the specific mechanics of how to get there. An RFP invites forwarders to propose a routing, a consolidation model, or a service structure, and to price that proposal — which means responses will legitimately differ from each other in substance, not just in number, and have to be evaluated as such.

The common failure is sending an RFQ when the real need was an RFP: asking for a price against a scope the buyer hasn't actually worked out yet, then being surprised that the numbers that come back don't compare, because each forwarder quietly filled the gaps with a different assumption about what was actually being asked for.

The volume data to produce before asking anyone for a price

Every input in this section has to come from the buyer's own shipment history and forecast — not from what "feels about right," because a forwarder pricing against guessed volume produces a rate that nobody ends up honoring once real volume shows up different.

Start with lanes: every origin-destination pair actually shipped, not just the two or three biggest ones — a lane quoted once at tender and then never revisited becomes a gap the winning forwarder has no obligation to hold a rate on. For each lane, break down annual volume by mode (sea FCL, sea LCL, air), and within sea freight, the container mix — how much moves as 20ft, 40ft, and 40ft high-cube, and what share is genuinely full containers versus LCL that gets consolidated. For air and LCL volume, chargeable weight profile matters as much as actual weight, since that's what's actually priced.

Seasonality needs to be shown, not just mentioned: if a quarter runs at double the average volume, a forwarder pricing against an annual average alone may not have accounted for whether it can actually cover that peak. State the Incoterms in force today, lane by lane, since that determines which costs already sit with the buyer versus the supplier and shapes what the tender is even asking forwarders to price.

Cargo characteristics and special handling — hazardous classification, temperature control, oversized or high-value cargo, anything requiring particular equipment or documentation — has to be disclosed upfront, because a rate priced against general cargo and then applied to cargo that needs special handling is not the rate that will actually be charged. Finally, state current pain points explicitly: is this tender about cost, about service reliability, about coverage on a lane the current forwarder handles poorly, or some combination — because that context shapes how the rest of the tender should be weighted.

Defining scope: what's included, and who does what

A rate is meaningless without a matching scope, and scope has to be defined at three points: at origin, at destination, and everything that happens in between.

At origin, specify whether the quote covers pickup from the supplier's factory or only from the forwarder's own consolidation point, whether export customs clearance is included, and who arranges and pays for origin trucking. At destination, specify whether the quote covers delivery to the buyer's own warehouse or only to a port or airport, whether import customs clearance and duty payment are included, and who handles destination trucking and any last-mile delivery. In between, specify the routing basis — direct or transshipment — and whether the quoted transit covers door-to-door or only port-to-port.

Customs responsibility deserves its own explicit line, separate from the general origin/destination split: which party's name goes on the customs declaration, who holds the import licence or permits where required, and who is liable if a declaration is found incorrect after the fact. This isn't a detail to leave implied by the Incoterm alone — buyers and forwarders both default to assuming the other side has it covered, and that assumption is exactly what surfaces as a dispute mid-shipment.

"All-in" is the phrase most likely to hide a gap, because every forwarder means something slightly different by it. The only way to make it mean something specific is to spell out, as a line item, every charge the all-in price is meant to include — origin handling, documentation, customs brokerage, terminal handling at both ends, destination delivery — and then require each forwarder to state explicitly, also as a line item, anything not included. A rate that looks lowest on the summary line and highest once local charges get added afterward is the single most common way a tender's winning number turns out not to be the number actually paid.

Structuring rate requests so responses are actually comparable

A tender produces comparable answers only when every forwarder is pricing against identical terms, so the request itself has to fix several variables before asking for a number.

Specify the Incoterm the rate is quoted against — a rate under FOB and a rate under DAP are not the same rate for the same cargo, and comparing them side by side without adjusting for that difference produces a false comparison. Specify the exact service scope from the definition above, so "all-in" means the same bundle of inclusions across every response. Specify the currency the quote must be denominated in, since a mix of currencies across responses buries the actual price difference under an exchange-rate assumption nobody stated.

State the validity period being requested and ask each forwarder to state theirs in return — a rate valid for thirty days and one valid for a year are not comparable on the number alone, because one carries far more pricing risk for the forwarder and is priced accordingly. Ask each forwarder to state explicitly what can move during that validity period and under what mechanism, and what is fixed regardless.

This is where surcharges need to be addressed structurally rather than left as a footnote: ask each forwarder to name which surcharge categories apply to the quoted rate, and how each one is triggered and adjusted, without asking for — or accepting — specific amounts or percentages baked into a public tender document, since those figures move with market conditions the forwarder doesn't control any more than the buyer does. What the tender should extract is the mechanism, not a number: which categories exist, what triggers a change, and how the buyer will be notified. A response that states "all-in, no surcharges apply" for an ocean or air rate over any meaningful validity period should be treated as a red flag rather than good news, because it usually means the surcharge exposure was left unstated rather than genuinely eliminated.

The surcharge categories worth naming explicitly

Even without asking for a single figure, a tender should require each forwarder to confirm which categories of surcharge could apply to the quoted rate, because the categories themselves are structural and worth understanding regardless of current market conditions.

Fuel-related surcharges move with the underlying cost of bunker fuel or jet fuel and are typically the most frequently adjusted component of an ocean or air rate. Currency-related adjustments apply when the base rate is set in one currency but costs on the ground are incurred in another, and exchange-rate movement between quotation and shipment shifts the forwarder's actual cost. Peak-season charges apply during periods of unusually high demand relative to available capacity, and tend to appear and disappear on a calendar the forwarder doesn't fully control either, since it follows market-wide capacity, not any single forwarder's schedule. Congestion-related charges apply when a port, terminal, or border crossing is running at reduced throughput, adding cost and dwell time that sits outside the forwarder's control. Equipment-related charges apply when specific container types are in short supply in a given location, or when a shipment needs equipment beyond a standard dry container.

None of these categories is exotic, and every established forwarder should be able to name them without hesitation. What separates a well-structured agreement from a fragile one isn't the absence of these categories — they're a structural feature of freight pricing, not a sign of a bad forwarder — but whether the mechanism for triggering and communicating them was agreed in advance. A tender that asks "which of these categories could apply to this rate, and how would we be notified" produces a far more useful, and far more honest, answer than one that just asks for the lowest all-in number and hopes surcharges never come up.

Service levels: what belongs in the agreement, and what doesn't

A tender should also define the service commitments expected alongside the rate, and the discipline here is being honest about which parts of the shipment a forwarder actually controls.

A forwarder does not control a vessel's schedule, a carrier's transit time, or a port's congestion — those sit with the carrier and the terminal, not the forwarder booking the space. What a forwarder does control is how it handles the parts of the process inside its own operation: how quickly it confirms a booking once a request comes in, how quickly it turns around shipping documents once cargo is loaded, whether it sends a pre-alert ahead of arrival so the buyer's own customs broker or warehouse can prepare, how it reports an exception the moment something goes off-plan rather than waiting to be asked, how it handles a claim when cargo is damaged or lost, and what its escalation path looks like when a shipment needs attention above the day-to-day contact.

Ask each forwarder in the tender to describe these processes specifically — not "we provide excellent service," but the actual workflow: who confirms a booking and by when relative to the request, what a pre-alert contains and when it's sent, how an exception gets flagged and to whom, what the claims process actually involves. This is the material that belongs in a service-level section of the eventual agreement, and it's evaluable in a tender the same way a price is, because two forwarders describing the same process in specific, concrete terms are easier to compare than two forwarders each claiming to be responsive.

What doesn't belong in this section is a fabricated performance number pulled out of the air to sound impressive — a specific figure only means something if it's one the forwarder actually tracks and reports against, and that's a conversation to have directly with the shortlisted forwarders once the tender narrows, not something to write into a comparison spreadsheet.

Evaluating bids: a scorecard, not a spreadsheet sorted by price

Once responses come back, resist the pull toward a spreadsheet sorted lowest to highest. Price should be one line in a weighted scorecard, not the sole basis for a decision, because a forwarder with the lowest number and no actual capability on the buyer's specific lanes typically ends up costing more once real shipments start moving.

Coverage on the buyer's actual lanes matters more than general network size — a forwarder with an extensive global presence but no established agent on the specific origin lane in the tender is starting from the same position as one with a smaller but more targeted network. Origin network strength speaks to how much of the process described earlier in this article a forwarder can actually control directly versus subcontract to a partner it doesn't fully oversee. Customs capability matters concretely: licensing, experience with the buyer's product category, and a track record of clean declarations rather than a general claim of "customs expertise." Systems and visibility determine whether the buyer will actually see shipment status and documentation in a usable form, or will be chasing updates by phone and email. Financial stability is worth checking directly rather than assuming — a forwarder in financial difficulty is a forwarder that may struggle to pay carriers on time, which becomes the buyer's problem when cargo gets held for an unpaid freight bill. Team responsiveness during the tender itself is a genuine data point: a forwarder that's slow or vague answering tender questions is unlikely to become faster once it has the business. References, where available, are worth calling rather than just collecting — a specific question about how the forwarder handled a real disruption reveals more than a general endorsement.

Weight these criteria before responses arrive, not after, so the evaluation isn't quietly reverse-engineered to justify whichever number already looks attractive.

Award structure: one forwarder, or more than one

A tender doesn't have to end in a single winner, and for many buyers it shouldn't. Awarding the entire volume to a single forwarder is administratively simplest and can produce the strongest pricing leverage on volume, but it also means every lane and every mode depends on one company's capacity, systems, and continuity.

A primary-plus-backup structure keeps a second, qualified forwarder live on the same or overlapping lanes, even if it carries only a small share of actual volume day to day. The value of the backup isn't the volume it moves under normal conditions — it's that during a disruption on the primary's side, whether a capacity crunch, a service failure, or something as basic as a system outage, the buyer already has a working relationship and an agreed rate structure to shift volume to, rather than needing to source and negotiate cold in the middle of a problem.

Splitting the award by lane is a related but distinct approach: rather than one primary across everything, different forwarders win different lanes based on where each genuinely has the strongest network and pricing. This tends to produce the best result on any single lane, at the cost of more relationships and more coordination overhead for the buyer to manage.

Which structure fits depends on the buyer's own volume concentration and risk tolerance. A buyer whose freight spend is dominated by one or two lanes has less to gain from splitting and more to gain from a strong primary with a backup on those specific lanes. A buyer with volume spread evenly across many lanes and modes may get more value from splitting by lane than from forcing one forwarder to cover territory it doesn't naturally serve well.

Implementation: the part most tenders skip

An agreed rate that never gets implemented properly saves nothing, and implementation is the stage most tenders under-invest in, because the tender itself feels finished the moment an award decision is made.

Implementation starts with a standard operating procedure document that spells out, in writing, the exact process agreed during the tender: who books, how, and by when; what documents each side sends and when; how an exception is reported and to whom; what the escalation path is when something needs attention above the day-to-day contact. This document should be specific enough that a new staff member on either side could follow it without needing to ask what was verbally agreed six months earlier.

A contact matrix — naming actual people, not just departments, for booking, documentation, operations, and escalation on both sides — prevents the common failure mode of a shipment stalling because nobody knew who to call. Systems setup covers whatever integration or portal access the agreement calls for: EDI connections, a booking portal, shared tracking, and confirming these actually work before the first real shipment moves through them, not after.

A go-live date should be set explicitly, ideally with a transition period where the outgoing and incoming forwarders (if there's a change) both have visibility into shipments crossing the changeover, so nothing falls into a gap between one arrangement ending and the next starting.

Finally, set a review cadence from day one — a fixed schedule, not an ad hoc one, for checking actual performance against what was agreed, addressed in the next section. A tender that ends at the award letter, with no SOP, no named contacts, and no review date, hands the winning forwarder a rate and a lane list and hopes the rest sorts itself out — which is exactly how a carefully negotiated agreement quietly reverts to being handled the same way ad-hoc spot bookings always were.

Ongoing management: measuring what actually happened

An agreement is a starting point, not a finished task, and the review cadence set during implementation is where a tender either keeps producing value or quietly stops mattering.

Measuring what actually happened means comparing the specifics agreed during the tender against what was delivered: were bookings confirmed within the process described, were documents turned around at the pace described, were pre-alerts sent, were exceptions actually reported when they occurred rather than only when asked about, was the claims process followed on the occasions it was needed. This is a different exercise from asking generally "how's it going" — it's checking the specific commitments made during the tender against specific instances of what happened.

Rate increase requests should be expected, not treated as a breach of the original agreement — cost inputs genuinely move, and a forwarder facing a real cost increase in one of the structural categories discussed earlier has a legitimate basis to raise it. What matters is whether the request follows the mechanism agreed during the tender (which category, what evidence, what notice period) rather than arriving as an unexplained new number. A request that doesn't map to any category discussed at tender time, or that arrives without the agreed notice, is worth pushing back on directly.

Deciding between re-tendering and renegotiating depends on what's actually driving the concern. If the issue is a forwarder underperforming against the specific commitments it made, or a rate that's drifted meaningfully out of line with what the current volume and scope would attract in the market, a fresh tender is often the more thorough route, since it re-tests the whole scope rather than just the point in dispute. If the relationship is otherwise working and the issue is narrower — one lane, one surcharge category, one service gap — renegotiating that specific point usually costs less time and preserves a working relationship that a full re-tender would otherwise put at risk for no real gain.

How tenders go wrong, stated plainly

Most of the ways a freight tender fails trace back to points already covered in this article, but it's worth naming them together as a final check before running one.

Sending an RFQ when the actual need was an RFP is the most common starting error — asking for a price against a scope that was never actually defined, then treating the incomparable numbers that come back as a genuine comparison. Guessing at volume rather than pulling it from real shipment history produces rates built on numbers nobody can hold to once actual volume looks different. Leaving "all-in" undefined lets every forwarder answer a slightly different question while appearing to answer the same one. Awarding purely on the lowest headline number, without weighting coverage, customs capability, systems, financial stability, and responsiveness, routinely produces a forwarder that's lowest-priced on paper and most expensive once real shipments start moving through gaps in its actual capability. Treating surcharges as something to ignore during the tender and deal with later, rather than agreeing the categories and mechanism upfront, all but ensures a dispute the first time a surcharge actually appears on an invoice.

Beyond the tender itself, the two failures that cost the most happen after the award: skipping implementation — no SOP, no named contacts, no go-live date — so a carefully negotiated agreement never actually gets followed in practice; and skipping the review cadence, so a forwarder's actual performance against what was promised is never checked until something has already gone wrong.

None of this makes a tender complicated in principle. It makes it a discipline: define what's actually being bought before asking anyone to price it, keep the comparison genuinely apples-to-apples, evaluate on more than the number, and follow through once the award letter is signed. A tender that does those four things consistently outperforms one that does none of them, regardless of which forwarder ends up winning.