From Lane KPIs to Quarterly Reviews: A 14-Step Carrier Scorecard Playbook
Most carrier scorecards measure the wrong things, or measure the right things and then get ignored at award time. Here are the formulas, data feeds, weights and review cadence that turn scores into rate decisions.
Steps 1–3: Decide What the Scorecard Must Change
A carrier scorecard is only worth building if it changes who gets your freight and at what price. Before you pick a KPI, write down the decision the score will feed. For most shippers that means three things: lane allocation in the next contract award, which carriers get a quarterly business review, and which go on a corrective action plan.
- Name the decision. "Scores below 70 for two straight quarters lose primary status on the lane" is a usable rule. "We'll monitor performance" isn't.
- Segment before you score. Score by mode (full truckload, LTL, parcel, ocean, air) and by lane group, not one blended number per carrier. A carrier can be excellent on dedicated regional runs and weak on long-haul spot coverage. Blend them and you hide both facts.
- Set the period and the minimum sample. Monthly scores rolled up quarterly work for most truckload networks. Require a minimum volume before a score counts; roughly 20 to 30 loads per quarter per lane group is a sensible floor. Below that, one bad week decides the grade.
Keep the KPI list to six or seven. Past ten, carriers stop knowing what to fix and your own team stops reading the report.
Steps 4–6: Lock the KPI Formulas Before Anyone Sees a Score
Most scorecard disputes aren't about performance. They're about definitions. Publish the formulas to carriers before the first period starts, and don't change them mid-quarter.
| KPI | Formula | Illustrative target range |
|---|---|---|
| Tender acceptance rate | Tenders accepted ÷ tenders offered at contract rate (first offer only) | 85–95% on contracted lanes |
| On-time pickup | Pickups within appointment window ÷ total pickups | 92–97% |
| On-time delivery | Deliveries within appointment window ÷ total deliveries | 93–98% |
| On-time in-full (OTIF) | Loads delivered on time with no shortage, overage or damage ÷ total loads | 90–96% |
| Tracking compliance | Loads with all required milestones received within the agreed latency ÷ total loads | 90–98% |
| Invoice accuracy | Invoices passing audit with no adjustment ÷ total invoices | 95–99% |
| Claims ratio | Claim value ÷ freight spend, or claims per 1,000 shipments | Well under 1% of spend |
Targets vary widely by mode, region and commodity. Treat these as starting points and calibrate against your own history.
- Define the clock. Is "on time" measured against the original appointment or a rescheduled one? Use the original unless the shipper or consignee asked for the change. Otherwise carriers can reschedule their way to a near-perfect score.
- Count tender acceptance on first offer. If a carrier rejects and then accepts a re-tender at a spot premium, that's still a rejection at contract rate.
- Set a floor and a target for each KPI. You'll need both to normalize scores in step 10.
Steps 7–9: Wire the Data Sources and Assign Fault
Your scorecard is only as good as its timestamps. You'll usually draw on four sources:
- TMS records: tenders, appointments, accept or reject responses and load attributes.
- Status messages: in North America, EDI 990 (tender response) and 214 (shipment status). In Europe, EDIFACT IFTSTA or status data from freight platforms. Increasingly, both arrive as API feeds from visibility providers such as project44, FourKites, Shippeo or Transporeon.
- Freight invoices and audit results: EDI 210 or invoice data from your freight audit and payment provider. These feed invoice accuracy and accessorial trends.
- Warehouse receipts and claims logs: for in-full and damage data.
- Close the coverage gaps. In India, Southeast Asia, Latin America and much of Africa, many regional carriers don't run EDI. Use driver apps, telematics or GPS integrations, with manual gate-in and gate-out scans as the fallback. If you collect driver location data, check local data-protection rules first. The EU's GDPR, India's DPDP Act and Brazil's LGPD all touch consent and retention in different ways.
- Add reason codes. Every late or failed load gets a code: carrier-caused, shipper-caused (load not ready, dock congestion), consignee-caused, or force majeure. Only carrier-caused failures count against the score. Skip this and your first QBR turns into an argument about your own docks.
- Give carriers a dispute window. Five to ten business days after the period closes is typical. After that, the data is final.
On tooling cost: scorecard modules come bundled with many mid-market and enterprise TMS platforms. Standalone visibility subscriptions can run from the low tens of thousands of dollars a year to well into six figures, depending on load volume and modes. Those are indicative 2026 ranges that vary by contract and region, so confirm with vendors. With fewer than eight carriers, a well-governed spreadsheet fed by TMS exports is a reasonable start.
Steps 10–11: Weight and Normalize So Scores Mean Something
- Normalize each KPI to 0–100. Use (actual − floor) ÷ (target − floor) × 100, capped at 0 and 100. If the on-time delivery floor is 85% and the target is 97%, a carrier at 94% scores 75. Raw percentages compress everyone into the 90s, so you can't tell carriers apart.
- Weight by what failure costs you on that lane type.
| KPI | Contract FTL | LTL | Ocean / international |
|---|---|---|---|
| Tender acceptance | 30% | 5% | 15% (booking acceptance) |
| On-time delivery / OTIF | 30% | 35% | 30% (schedule reliability) |
| On-time pickup | 10% | 15% | 10% |
| Tracking compliance | 10% | 10% | 15% |
| Invoice accuracy | 10% | 20% | 20% |
| Claims | 10% | 15% | 10% |
These weights are starting points, not a standard. A shipper selling to large retailers that fine suppliers for missed delivery windows should push OTIF higher. LTL invoice accuracy carries a heavy weight because reweighs, reclassifications and accessorials are where LTL costs drift. Review the weights once a year, never mid-period.
Steps 12–14: Run the Quarterly Review and Price the Score Into the Award
- Tier the cadence. Run weekly operational dashboards for planners and send monthly scorecards to every carrier. Hold a quarterly business review only with the carriers that carry most of your spend, usually the top five to ten. A useful QBR covers the score trend, the three worst lanes, root causes the carrier owns, your own shipper-caused failures, a capacity outlook and two or three agreed actions with owners and dates.
- Convert service failures into cost. This is where most programs stop short. Estimate what each failure costs you. A rejected tender costs the spot premium on the re-tender; a late delivery costs chargebacks, expediting or overtime.
- Award on score-adjusted cost, not rate.
Hypothetical worked example: a distributor runs 100 loads a quarter on a lane. Carrier A bids $1,850 per load, rejects 6% of tenders and delivers late on 4% of loads. Carrier B bids $1,780 but rejects 18% and runs late on 11%. Assume a $300 spot premium per rejection and $400 in fines and expediting per late load. Averaged per load, Carrier A costs $1,850 + $18 + $16 = $1,884. Carrier B costs $1,780 + $54 + $44 = $1,878. B's $70 rate advantage shrinks to $6. At that margin, A should get primary status, since it's the more dependable carrier, and B should take backup volume.
Watch for these red flags in your own program:
- Every carrier scores above 90. Your floors are too lax.
- Carriers dispute most of the data every period. Your reason codes or timestamps aren't trustworthy.
- Scores never show up in award decisions. Carriers will notice and stop putting effort into QBRs.
Use this rule for your next bid: if two carriers' score-adjusted costs fall within 2% of each other, award the lane to the higher scorer. Put that rule in your RFP so every bidder knows service is priced in.
Quick answers
How many carriers should get a full quarterly business review?
Usually your top five to ten carriers by spend, or the ones covering your most critical lanes. Everyone else gets a monthly scorecard and an escalation call only if they fall below threshold. A QBR takes real prep on both sides, so save it for relationships where the outcome moves meaningful volume.
Should we share the carrier scorecard with carriers or keep it internal?
Share it, along with the formulas and weights. Carriers can't fix what they can't see, and publishing formulas up front cuts disputes later. You can keep your score-adjusted cost assumptions internal, but tell bidders that service scores factor into awards.
What if our smaller carriers can't send EDI or API status updates?
Score them on what you can measure reliably: tender acceptance from your TMS, delivery confirmation from warehouse receipts, and invoice accuracy from audit. You can also require a driver app or GPS link as a contract condition, phased in over a quarter or two. Check local data-protection rules before you collect driver location data.
Written by the Freight Base System editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.