How TruScore works

The 0 to 100 score is retained so carriers can be compared even within the same letter.

TruScore is built entirely from public records, refreshed on a rotating basis so a grade reflects the latest data we hold:

• FMCSA operating authority and registration records (authority status, age, lapses, revocations, reported fleet size).
• FMCSA insurance filings (coverage on file, filing history, cancellations).
• FMCSA roadside inspection and crash records (out of service rates, crash and fatal crash counts, inspection volume).
• NHTSA vehicle records (fleet make-up and equipment profile).

Four factors build the score. Two are fraud checks that can cap it. Every factor drills down to the exact records behind it.

• Authority Track Record (30%). Length and cleanliness of operating authority. Longer continuous authority raises it; lapses (normalized by tenure) and out of service orders lower it; revoked authority is a hard cap.
• Insurance Stability (25%). We confirm current coverage only from a live snapshot; a filing history alone shows the carrier held coverage, not that it holds coverage today. A long lapse free history raises it; confirmed loss of coverage and churn lower it.
• Safety Posture (25%). Vehicle and driver out of service rates versus national averages, plus crash and fatal crash counts as rates per fleet unit, so a large carrier is not penalized for size. A substantial clean inspection record counts as positive evidence.
• Fleet Profile (20%). Unique vehicles observed at roadside, a newer average model year, and how well the observed fleet matches the reported fleet size.
• Chameleon Risk (fraud check). Reincarnation risk from a shared address, phone, or officer across carriers, plus links to previously revoked or defunct authority. We count related companies, not matching fields: one company reached by both a shared address and a shared phone is a single relationship, and is scored once. Where we cannot tell whether two identifiers point at one company or two, we assume one. One or two related carriers is an affiliation and does not flag on its own; it escalates to a public High flag only with corroboration from a defunct or revoked link. A contact shared with many carriers is treated as a filing service, not a fraud ring.
• Paper Carrier Check (fraud check). Looks for a carrier that reports trucks but has no verifiable operating footprint. On its own that is not treated as fraud: a one or two truck operator can genuinely go years without being stopped for a roadside inspection, so silence is not evidence. It is flagged only with corroboration from a second, independent signal, such as a past authority lapse, an out of service order, contact details shared with other carriers, links to defunct carriers, or a reported fleet far too large to have gone unseen. Without that second signal the carrier is noted for review rather than flagged. A brand new carrier not yet inspected is labeled New.

• The four performance factors are combined on their weights.
• A factor with no data behind it contributes a neutral value rather than a penalty. We do not punish a carrier for data we do not have.
• Either fraud check caps the grade low regardless of the other factors, but only once corroborated. Neither flags on a single signal.
• A revoked carrier, or one currently out of service, is capped near the bottom.
• The top grade (A) is awarded only with full data coverage and a genuinely strong record, so an A is always earned on real evidence.

Confidence reflects how much real data backs a grade, measured by how many of the three enrichment factors (Insurance, Safety, Fleet) we hold for the carrier.

When we hold no enrichment data on a carrier, we return NR (not rated) rather than stamp a letter on a carrier we cannot see. This mirrors how a rating agency declines to rate an entity it lacks data on. It is a deliberately honest posture: we grade every carrier we have evidence on, and we say plainly when we do not. Revoked and out of service carriers are still surfaced even on thin data, because that status is itself a clear signal.

Separately, a private fleet (PF) is not graded at all. A private carrier hauls its own freight, so it files no for hire authority and leaves very little federal record behind. Scoring it against for hire carriers would mostly measure how rarely it appears in public data, not how well it operates, so we decline to grade it and say so. This is different from NR: NR means we looked and found too little, PF means this is not a carrier type we grade. Fraud, revocation and out of service findings are still shown for private fleets.

TruScore is not static. Public records are refreshed on a rotating basis, and when a refresh brings in new insurance, safety, or fleet data, a carrier's TruScore recomputes and a carrier that was NR moves into a rated grade the moment enough data exists. Carriers for whom the public record still holds no operational data remain NR until data appears.

Every factor in every grade drills down to the specific public records it was computed from, so a carrier or broker can see exactly why a grade is what it is. Because TruScore is built on public records, the underlying data is independently verifiable at the source. A carrier that believes a record is wrong can raise it with us.