A freight RFP is useful only when shipment history, future demand, operating rules, price structure, and award logic describe the same transportation problem.

A freight RFP can be competitive, analytically sophisticated, and still fail on the first week of routing.

The shipper publishes twelve months of lane volume but omits the weeks when it actually moved. The rate sheet asks for dry-van linehaul while the freight regularly requires drop trailers, driver assist, team service, or constrained appointments. Carriers price the visible linehaul and qualify the hidden operation in free-text notes. Procurement compares the numbers as if the bids describe the same service.

They do not.

The test for the event is straightforward:

A freight RFP must let a provider price the future operating obligation and let the shipper test a realistic award—not merely compare cells in a bid sheet.

That requires a clean historical baseline, an explicit forecast, service rules, commercial definitions, carrier assumptions, and award logic that can survive implementation.

Decide what the bid is meant to change

Start with the business problem. “Market test” is too vague to design the event.

The RFP may be intended to:

  • rebid an incumbent network;
  • secure committed capacity on volatile or underserved lanes;
  • add regional or specialized carriers;
  • consolidate providers;
  • build backup coverage;
  • support a facility opening, acquisition, or customer launch;
  • change mode or service design;
  • introduce drop trailers, pooling, consolidation, or other operating changes;
  • improve service, data, safety, environmental, or commercial performance;
  • reset the fuel or accessorial structure.

Name the primary objective and the non-negotiable constraints. A cost-focused event with no realistic volume commitment is different from a capacity event for a launch date. A provider-diversification project should not be optimized into another single-carrier award.

Define scope by legal entity, geography, mode, business unit, shipment direction, service, and planned start date. If brokerage, dedicated carriage, managed transportation, warehousing, or final mile are all present, separate the roles and commercial schedules. The distinctions in UniteCargo's broker, 3PL, and forwarder guide remain relevant during procurement; a brand name does not tell the bidder which regulated or contractual role it will perform.

Build a historical baseline that can be priced

Raw transportation history is not automatically bid data. Normalize it into a shipment-level file and publish a data dictionary.

At a minimum, include fields relevant to the service:

  • unique anonymized shipment identifier;
  • origin and destination postal code, city, state, facility code, and time zone;
  • pickup and delivery date or agreed transit requirement;
  • order date, tender lead time, and material appointment restrictions;
  • mode, service level, equipment, trailer length, and temperature or special handling;
  • number of stops and stop sequence;
  • weight, handling units, dimensions or cube where relevant;
  • commodity description and applicable freight classification;
  • hazardous-material, food, pharmaceutical, high-value, or other controlled-service indicator;
  • actual linehaul, fuel, and accessorial categories where disclosure is appropriate;
  • dwell, cancellations, layovers, redelivery, rejected tenders, and recovery usage;
  • awarded provider type and whether the move was routing-guide, spot, or emergency freight.

Remove direct customer or supplier identity when confidentiality requires it, but do not anonymize away the operating constraint. “Customer A” can still be described as appointment-only, driver-unload, no weekend receipt, or urban limited-access delivery.

Treat blanks as information. A missing dimension is not zero. An absent accessorial field may mean no charge, an unstructured invoice, or a system that never captured the event. Publish known data-quality limitations so carriers can state how they priced the uncertainty.

The Federal Highway Administration's Freight Analysis Framework can support broader network and commodity context, but public flow data cannot replace the shipper's own shipment-level operating history. A national corridor forecast does not reveal the Tuesday appointment pattern at one receiver.

Separate history from the forecast

Carriers are pricing future work. Show what is expected to change.

For every lane or service cohort, provide:

  • historical shipment and volume count;
  • weekly or monthly distribution, not only an annual total;
  • peak week and credible peak-day behavior;
  • forecast base, high, and low range;
  • expected seasonality;
  • known facility, customer, product, sourcing, or service changes;
  • new lanes with no history;
  • discontinued volume that should not be priced;
  • confidence level and forecast owner.

Volume ranges are more honest than false precision. A carrier can plan around “three to five loads per week, with eight during a documented six-week peak” better than an annual forecast of 214 loads with no cadence.

Distinguish addressable volume from awardable volume. Addressable volume appears in the file. Awardable volume remains after incumbency commitments, customer routing, private fleet, mode shifts, minimum-provider rules, and other constraints. If a lane may be split, state the minimum practical share.

Do not promise a volume commitment unless the organization can control the tenders. If the bid is non-binding, say so. Then recognize that providers will price the risk of an award that may not become freight.

Publish the service specification

The lane file says where freight moved. The service specification says what the provider must do.

Define:

  • tender method, required fields, and response clock;
  • minimum lead time and treatment of late tenders;
  • pickup and delivery appointment ownership;
  • operating hours, closures, and holiday calendars;
  • driver work, loading, unloading, pallet exchange, lumper, and securement responsibility;
  • live, drop, preload, or trailer-pool expectations;
  • tracking milestones, transmission method, latency, and exception escalation;
  • proof-of-delivery content and submission timing;
  • cargo, packaging, seal, temperature, hazardous-material, and security controls;
  • claims notice and evidence process;
  • invoice format, support, dispute, approval, and payment workflow;
  • surge, outage, weather, and recovery expectations.

If the service requires a trailer pool, include turns, dwell, balance rules, yard visibility, inspection, damage, and peak inventory. The decision framework for live load versus drop trailer can help define that schedule.

Do not tell carriers to assume “industry standard.” The disputed item is often the one each party believed was standard.

Use eligibility gates before commercial ranking

A low bid does not cure an ineligible provider.

Resolve the bidding legal entity and required role. Define minimum authority, insurance, financial responsibility, operating capability, geographic coverage, equipment, security, claims, technology, and special-service requirements. State which requirements are hard gates and which are evaluated attributes.

FMCSA explains that federal insurance filing requirements vary by entity, authority, cargo, and vehicle type. Those filings are not a substitute for reviewing the insurance actually required by the contract and shipment. Apply the evidence sequence in How to Vet a Motor Carrier before award and again before activation.

Do not create an unexplained safety-score cutoff from public data. FMCSA has documented limitations and appropriate uses of Safety Measurement System data. Define a review process that considers identity, authority, out-of-service status, inspection and crash context, safety rating where present, operating history, corrective action, and shipment fit.

For environmental procurement, use a defined method. EPA SmartWay provides shipper tools, carrier performance information, and guidance; it also publishes sample RFP language. Decide whether participation, reporting, or performance is a gate, a weighted attribute, or a development objective.

Make the price sheet describe the commercial model

Do not ask for one “all-in” number unless the same operating facts truly govern every shipment.

Depending on mode and service, collect:

  • linehaul unit and basis;
  • minimum charge;
  • distance source and mileage method;
  • fuel index, baseline, increment, effective lag, and rounding;
  • freight class, density, dimensional, pallet, or weight basis;
  • stop, detention, layover, truck ordered not used, redelivery, storage, limited-access, driver-assist, lumper, toll, and other relevant accessorials;
  • equipment or team-service premiums;
  • seasonal, surge, and capacity-event rules;
  • rate geography and arbitrary rules;
  • currency, tax, and cross-border assumptions where applicable;
  • validity dates and reopening triggers.

The freight-pricing architecture explains why linehaul, fuel, class, density, minimums, and accessorials must be normalized before comparison. If fuel uses the U.S. Energy Information Administration's weekly retail on-highway diesel series, specify the geography, publication week, baseline, increment, and effective date. “DOE fuel” is not a complete formula.

For detention and related charges, state the clock, free time, notice, evidence, exclusions, and rate. The operational and legal distinctions in UniteCargo's detention, demurrage, per diem, and storage guide should be resolved in the commercial schedule rather than left for invoice disputes.

Require bidders to return exceptions in structured fields:

  • bid as requested;
  • bid with stated exception;
  • no bid;
  • alternative service proposal.

A rate qualified in a buried note is not directly comparable to an unqualified rate.

Let carriers see the operating distribution

Averages conceal the part providers must resource.

For each material lane or cohort, show distributions for:

  • day of week;
  • tender lead time;
  • pickup time;
  • number of stops;
  • weight and cube;
  • dwell;
  • appointment window;
  • cancellations;
  • volume by week;
  • spot or recovery frequency.

If 80 percent of loads tender with two days' lead time and the remaining 20 percent arrive four hours before pickup, the average lead time may look reasonable while the operation remains expensive. If annual volume is concentrated in one retail reset or harvest period, the average week is not the capacity problem.

Allow a structured question period. Publish answers to all bidders when the answer affects scope or pricing. Issue a versioned amendment rather than changing a shared spreadsheet silently.

Normalize bids with executable scenarios

First normalize the rates. Then simulate the operation.

Create representative scenarios:

  • clean standard load;
  • peak-week load;
  • short-lead tender;
  • multi-stop shipment;
  • live load with defined dwell;
  • drop-trailer turn;
  • shipment requiring accessorial support;
  • service failure and recovery;
  • lane with partial award.

Apply each bid's linehaul, fuel, minimums, accessorials, and exceptions to the same scenario facts. Compare total expected cost, not just base rate. Keep current or benchmark cost available as a reference, but do not force an award simply because a bid beats an imperfect incumbent baseline.

Then test an actual award. A solver can minimize modeled cost while assigning every difficult lane to carriers that did not offer enough volume, geographic balance, equipment, or surge capacity. Apply constraints for:

  • maximum and minimum carrier share;
  • carrier-stated lane capacity;
  • facility and regional coverage;
  • incumbent transition limits;
  • service and equipment eligibility;
  • provider-diversity requirements;
  • backup coverage;
  • implementation capacity.

Run failure scenarios. Remove the largest carrier. Reduce the top carrier's accepted volume. Increase peak demand. Apply a fuel move. Model accessorial incidence above the historical average. An award that works only under the base forecast is not a robust operating plan.

Evaluate more than price—but define the evidence

A weighted matrix is useful when every criterion has an observable test.

Dimension Evidence to request
Capacity fit Lane-level weekly commitment, equipment, terminals, drivers, partner-carrier model, surge process
Service Comparable lane history, operating process, references, corrective-action examples
Commercial Normalized scenario cost, exceptions, index logic, validity, payment terms
Implementation Named owner, integration plan, test schedule, training, activation gates
Visibility and data Supported events, method, latency, uptime history, correction and outage process
Risk and compliance Entity, authority, insurance, security, claims, controlled-service qualifications
Network value Facility balance, backhaul fit, geographic density, alternative modes or capacity
Relationship governance Escalation, review cadence, data-dispute process, continuous improvement

Avoid points for polished narratives with no evidence. Ask finalists to work through an actual lane, exception, invoice, and service failure. The answer should reveal owners and decisions, not only technology features.

Use the definitions from the carrier-performance scorecard as the future operating baseline. A bidder should know how acceptance, pickup, delivery, tracking, dwell, claims, and invoice performance will be calculated before accepting the award.

Build the award into the routing guide

An award file is not an operating launch.

For each awarded lane, publish:

  • primary, secondary, and recovery position;
  • awarded share or capacity;
  • effective and expiration dates;
  • accepted rate and version;
  • tender lead time and response clock;
  • service and equipment scope;
  • known exclusions;
  • escalation contacts;
  • activation status.

Load the structure into the routing guide, then test tenders, acceptances, status, proof of delivery, invoice, and exception paths before production freight. X12 maintains distinct transaction sets for rate requests and proposals, load tenders, tender responses, shipment status, and invoices. Whether the parties use EDI, API, portal, or email, those are still different business events and should be tested separately.

Do not activate every new provider on the same day. Use controlled lane pilots, confirm identity at tender and pickup, monitor early loads manually, and close data or payment defects before scaling.

Common freight RFP failure modes

Reject or repair the event when:

  • annual lane volume has no weekly distribution;
  • historical and forecast periods are mixed;
  • service requirements live only in bidder questions;
  • facility codes cannot be reconciled with the rate sheet;
  • accessorial history is absent but the bid assumes it is zero;
  • bidders cannot express assumptions and exceptions structurally;
  • award optimization ignores carrier capacity or shipper control of volume;
  • provider roles and legal entities are ambiguous;
  • a lower linehaul rate is compared with a more inclusive incumbent rate;
  • data, implementation, or invoice testing begins after go-live;
  • the routing guide does not reflect the award;
  • the shipper has no process for honoring, monitoring, or revising awarded volume.

The answer is not a longer questionnaire. It is a tighter connection between operating facts, commercial rules, provider evidence, and the award.

What a usable RFP leaves behind

A good freight RFP makes uncertainty visible.

Publish clean history and known defects. Separate the baseline from the forecast. Show demand distributions and operational constraints. Define the service and price architecture. Gate eligibility before ranking bids. Collect assumptions in a comparable form. Simulate total cost and a feasible award. Stress the award, then install it into a tested routing guide with measurable performance.

That process may produce fewer superficially cheap bids. It is far more likely to produce rates a provider can honor and capacity the operation can actually use.

When the scope and role are clear, use the UniteCargo logistics directory to identify additional carriers, brokers, 3PLs, warehouses, forwarders, and specialized providers for the qualified bidder pool.

Sources and further reading

Best Practices