How the grade works

🛡️ Safety Score (50%)

From FMCSA public records (the SMS bulk files and the QCMobile API):

• Out-of-service rates— vehicle and driver OOS rates, shrunk toward the national averages (20.7% / 5.5%) until a carrier has enough inspections, so 0% on 2 inspections isn't read as a perfect record. We also show how the rate compares to similar-size carriers.
• Crash rate — severity-weighted crashes (fatal and injury weighted heavier than tow-away) per power unit, over a rolling 24-month window. A carrier with an inspection record and no crashes on file counts that as a real (mildly positive) signal rather than a blank — scaled by fleet size and capped, since an absent crash record can also be a data gap: 0 crashes on 2 trucks is weak evidence, and even a large fleet with no crash record never earns a perfect crash score it didn't demonstrably prove.
• Operating authority.A for-hire carrier with no active FMCSA operating authority can't legally take a load — we cap its grade at Cand show a warning. Private and intrastate carriers, which don't need that authority, are not flagged.
• Safety rating — hard cap. An Unsatisfactory rating or an active out-of-service order caps the grade at F (the carrier is unfit to operate). A Conditional rating caps it at D. The strictest cap wins.
• What an F means. F is reserved for carriers that are unfit or dangerous — an Unsatisfactory rating, an out-of-service order, or a real crash/fatality pattern. An active, authorized, crash-free carrier with poor inspection numbers tops out at D("verify carefully"), not F — high out-of-service rates are a maintenance problem, not proof of danger.

We never punish missing data — weights renormalize over whatever signals exist. Thin inspection samples are held near the national average (0% out-of-service on 2 inspections isn't a clean record), and every grade carries a graded confidence label — low (<10 inspections), medium (10–29), or high (30+) — with the observed numbers always shown alongside their sample size.

📈 Stability Score (30%)

How likely is this carrier to still exist — and still pay you — in six months? Built from the carrier's own FMCSA Licensing & Insurancerecord, not from market indices: every carrier in a soft market isn't equally likely to fail, so we score the carrier, not the weather. Four signals:

• Insurance discipline (40%) — cancellations and coverage gaps in the last 36 months. Carriers drop insurance right before they shut down; a 30+ day gap in required liability coverage is one of the strongest distress signals in trucking. A policy switched with no gap is treated as routine, not distress. A carrier with no required insurance on file right now scores 0 and gets a red warning banner. Carriers whose authority has since lapsed are still scored on their filing history — an insurance collapse is usually why a carrier died, so it stays visible.
• Authority age (30%) — years since first operating authority. First-year carriers fail at very high rates; a carrier that has operated 20+ years has survived multiple freight recessions. Carriers without for-hire authority use their FMCSA census first-seen date instead.
• Fleet trajectory (20%)— is the fleet growing or shrinking between census snapshots? Carriers usually downsize before they fail. This signal reads neutral (flagged "limited") until enough snapshot history accumulates.
• Authority history (10%) — involuntary revocations. One revocation reinstated years ago is survivable history; a revocation in the last 3 years, or several ever, is a live red flag.

Reincarnated-carrier check:a carrier sharing an exact street address and phone number with a carrier whose authority was revoked in the last 5 years — and first registered around or after that revocation — is flagged as a possible "chameleon" and capped at D, with a warning naming the revoked DOT. Matching is deliberately conservative to avoid false positives.

⭐ Driver Reviews (up to 30%)

Verified, anonymous reviews from drivers who worked there (phone-OTP verified; we store a one-way hash, never the number). The weight earns its way up with volume:

• Fewer than 3 verified reviews → reviews don't count at all (no swinging a grade on one voice). The grade is carried by safety and stability — we don't inject a neutral placeholder that would drag every carrier toward the middle.
• At 3 reviews they count 10%, growing linearly to 30% — and the count needed for full weight scales with fleet size (2× the truck count, between 8 and 25). Six verified reviews from an 11-truck fleet is proportionally better coverage than 25 from a 9,000-truck one. Safety and stability share the rest in their 50:30 ratio, and the exact weights used are shown on every grade card.
• One review per driver per carrier; structured questions, not blank rants.

Other states

• Inactive / revoked carriers show a warning to verify before signing.
• Zero trucks on file flags a possible broker or shell company.
• Carriers with too little FMCSA history to grade (new, very small, intrastate, or dormant) get a "not enough record to grade" state — their operating-authority and operation status, not a guessed letter.
• Shown as context, not scored: what recruiters tend to promise for a carrier (from drivers' Promise Checks) and reviews split by company-driver vs owner-operator. These are descriptive and never move the letter.

Disclaimer

Readers should not draw conclusions about a carrier's overall safety condition based solely on this data. Unless a carrier has received an UNSATISFACTORY safety rating under 49 CFR Part 385 or has been ordered to discontinue operations by FMCSA, it is authorized to operate on the nation's roadways.

Grades are CarrierGrade's independent analysis of public data; not an FMCSA safety rating.

CarrierGrade is informational and is our independent analysis of disclosed public data — not legal, safety, or hiring advice, and not an FMCSA safety rating. Money never influences a grade.