TMS RFP Mistakes That Lead to the Wrong Shortlist, and What to Ask Vendors Instead
Most TMS shortlists are decided by feature grids and polished demos that never touch your freight. Here are nine RFP mistakes, plus the questions and tests that show which vendors can run your network.
Why feature grids produce a shortlist of lookalikes
Many shippers who regret their transportation management system chose it from three finalists that scored within a few points of each other. That's rarely bad luck. It usually comes from a TMS RFP that measured the wrong things. When every finalist answers "yes" to 400 requirements, the decision falls to price and demo charisma. Neither tells you how the system will handle your lanes, carriers and exceptions.
The first two mistakes happen before a single vendor gets involved.
Mistake 1: Sending a generic feature grid
A downloaded TMS requirements template with rows like "supports multi-modal planning" or "offers carrier rating" just invites yes/no answers. Established vendors, including Oracle Transportation Management, SAP TM, Blue Yonder, Manhattan Active TM, e2open and Descartes, will tick almost every box. The grid can't separate native capability from configuration, partner add-ons or roadmap items.
What to do instead: Make every high-weight requirement scenario-based and ask for one of four answers: standard and configurable by your team, configured by the vendor during implementation, needs custom development, or not available. Then ask the follow-up that exposes the gap: "Which of these answers depend on a third-party product or a release that isn't generally available today?"
Mistake 2: Writing requirements before profiling your freight
Requirements written from a wish list overweight features you'll rarely use. A distributor moving most of its volume as LTL and parcel doesn't need the same planning engine as a manufacturer running multi-stop full truckload and ocean imports.
What to do instead: Before you write a requirement, pull 12 months of shipment data and summarize it by mode, region, lane count, carrier count, shipments per day and accessorial frequency. Weight your TMS evaluation criteria to match. If ocean is 5% of spend, it shouldn't carry 20% of the score.
Demos that run on vendor data prove almost nothing
A vendor's demo environment is tuned to look good. Lanes are clean, rates load perfectly, and every carrier answers a tender within seconds. Your network won't behave that way, so the demo shouldn't either.
Mistake 3: Letting vendors demo on their own data
If finalists show their standard demo, you're comparing three sales engineers, not three systems.
What to do instead: Two to three weeks before the demos, send every finalist the same anonymized data pack: 2,000 to 5,000 representative shipments, your real rate structures (fuel tables, accessorials and minimums included), 20 to 40 locations with dock hours, and your carrier list. Require them to load it and demo on it. A vendor that can't load a realistic rate file in two weeks is telling you something about implementation.
Mistake 4: Scripting only the happy path
Nearly every TMS can plan a load, tender it and print a document. The differences show up when something goes wrong.
What to do instead: Build a TMS demo script where at least half the time goes to exceptions, and run the same script, in the same order, with every vendor. Good scenarios include:
- A primary carrier rejects a tender 90 minutes before pickup. Show the automatic waterfall to backup carriers and the cost impact.
- An order is split after planning. Show re-planning without manual load rebuilding.
- A carrier invoice arrives 6% over the rated amount with an undocumented detention charge. Show the audit, dispute and approval workflow.
- A rate amendment takes effect mid-month. Show who can load it and how fast.
Have your own planners and freight-audit staff drive part of each demo. If the vendor resists handing over the controls, write that down.
Carrier connectivity and compliance in the regions you actually ship
This is where global shippers get burned most often. A vendor's integration library can look huge on paper and still be thin in the countries that matter to you.
Mistake 5: Accepting a carrier connectivity count at face value
"Connected to 50,000 carriers" might mean deep API integration with large North American and European carriers, and email-only connections with the regional carriers in Southeast Asia, Latin America or Africa that move your freight. Tender, status and invoice connectivity are separate things, and many integrations cover only one.
What to do instead: Send your top 30 carriers by spend in each region and have vendors fill in a matrix: for each carrier, is it live today, and for which message types (tender, tender response, status, proof of delivery, invoice), by API, EDI or portal? Then ask: "For carriers not live today, who builds the connection, how long does it take, and who pays?" Also ask whether real-time visibility is native or comes through a partner network such as project44 or FourKites, and whether the partner is billed separately.
Mistake 6: Treating regional compliance as an implementation detail
Freight documentation and tax rules differ sharply by country. India uses e-way bills for many goods movements under GST. Mexico requires the Carta Porte complement for goods in transit. Brazil relies on electronic transport documents such as CT-e and MDF-e. The EU is moving toward electronic freight transport information under its eFTI framework, alongside wider use of e-CMR. Data-protection regimes such as the GDPR in Europe and India's DPDP Act can also affect where shipment and driver data may be hosted.
What to do instead: List the documents and filings each country requires, and ask vendors to show, not just describe, how they generate them. Ask whether compliance comes from native functionality, a localization partner or your own development, and who keeps it current when rules change. Your tax and legal advisors in each country should confirm the requirements list. Don't let the vendor define it.
The optimization benchmark and the true cost comparison
Optimization is usually the biggest source of savings in a TMS business case, and the claim buyers verify least.
Mistake 7: Skipping the optimization benchmark
Vendors will quote typical freight savings. Those numbers come from other networks, other baselines and other levels of planning maturity, so they say little about yours.
What to do instead: Give each finalist the same historical data, say one representative month, along with your real constraints: equipment types, dock windows, driver-hours rules for the relevant countries, and customer delivery windows. Ask them to plan it and return load counts, total distance, cost and run time. Compare the results with what you actually shipped. A hypothetical: say your actual month cost $1.2 million across 1,400 loads. If one vendor's plan comes in at 1,310 loads and $1.12 million while honoring every constraint, you have a defensible savings estimate. If another produces a cheaper plan by breaking dock hours, it fails. Insist on the solved plan file, not just a slide with totals.
Mistake 8: Comparing subscription prices instead of total cost
Subscription fees are often well under half of five-year cost. The table shows the main drivers. All figures are indicative, typical 2026 market ranges in USD that vary widely by contract, region and volume, so confirm them with vendors.
| Cost driver | Typical 2026 range (indicative) | What to ask |
|---|---|---|
| Annual SaaS subscription, mid-market | $40,000 to $250,000 | Is pricing by shipment, freight spend, user or module? What happens when volume grows 30%? |
| Annual subscription, large enterprise | $250,000 to $1,000,000+ | Which modules, regions and environments are included? |
| Implementation services | Roughly 0.5x to 2x first-year subscription | Fixed fee or time and materials? Which partner, and in which time zones? |
| Carrier onboarding | Hundreds to a few thousand dollars per connection | Which carriers are included, and who maintains the connections? |
| Visibility and tracking | Cents to a few dollars per tracked shipment | Native or partner? Billed by the TMS vendor or separately? |
| Localization and e-documents | Varies widely by country | Included, add-on, or partner-delivered? |
Ask every finalist for a five-year cost on the same volume assumptions and compare that number, not year-one pricing.
References, scoring and the decision rule that holds up
Mistake 9: Letting vendors choose the references and scoring breadth
References a vendor hands you are its happiest customers. And a scoring model that adds up hundreds of equally weighted lines favors whichever vendor has the longest feature list, not the best fit.
What to do instead: Ask for references that match your profile: similar modes, similar regions, live for 12 to 36 months, so they're past the honeymoon but still remember go-live. Ask each one three things: "How far was your actual go-live from the original date?", "Which carriers in your region still aren't connected?" and "What did year-two cost compared with the contract?" Industry peer groups and your implementation partner's other clients are good sources of unprompted references.
Then restructure the scoring. For most shippers, this weighting holds up better than a flat grid:
- Scripted demo on your data, including exceptions: 30%
- Optimization benchmark results: 20%
- Carrier connectivity and compliance in your regions: 20%
- Five-year total cost: 15%
- Implementation approach and references: 15%
Score each area independently before anyone discusses pricing, so a low quote doesn't color how evaluators see capability.
The decision rule: if a vendor fails the exception script or breaks constraints in the benchmark, drop it, whatever it scored on the grid. Your next step is concrete. Pull 12 months of shipment data this week and build the freight profile. Every later part of your TMS vendor selection, including requirements, demo pack and benchmark, depends on it.
Common questions
How long should a TMS RFP process take?
For a mid-market shipper, eight to fourteen weeks from requirements to contract is a realistic range: a few weeks to profile freight and write requirements, about three for responses, and four to six for scripted demos, the benchmark and references. Enterprise programs covering several regions often take longer. Cutting the demo and benchmark phase to save time is usually a false economy.
How many vendors should be on a TMS shortlist?
Send the RFP to six to eight vendors, then narrow to three for scripted demos and the optimization benchmark. More than three finalists stretches your team too thin to prepare data packs and run consistent demos. Fewer than three leaves you little negotiating leverage.
Should we use a free TMS requirements template?
A template helps you catch categories you might miss, such as freight audit, dock scheduling or returns. Don't send it as-is, though. Rewrite your high-weight requirements as scenarios drawn from your own freight profile, and cut lines that don't match how you ship.
What data should we give vendors for the demo and benchmark?
Provide an anonymized sample of 2,000 to 5,000 shipments, real rate structures including fuel and accessorials, location and dock-hour data, your carrier list, and your planning constraints. Remove customer names and any pricing you consider sensitive, and put a mutual NDA in place first. Check with your data-protection advisor before sharing personal data, such as driver details, across borders.
Is a TMS from our ERP vendor a safer choice?
Not automatically. TMS modules from ERP vendors can simplify integration and support. But carrier connectivity, optimization depth and regional compliance vary as much there as with specialist vendors, so run them through the same demo script and benchmark instead of exempting them.
Do we need a consultant to run TMS vendor selection?
It isn't required, but an independent advisor can speed up requirements work and bring pricing benchmarks from other deals. Check whether the advisor takes referral fees or has implementation partnerships with any vendor on your list, and keep that firm out of scoring if it does. Contract and data-protection terms should still go through your own legal counsel.
Written by the Logix Gridnode editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.