90Out

Socket Solutions › 90 Out › Method

Method 90out-sms-1

How 90 Out scores fleets

Every number comes from public FMCSA data and a published method, so two agencies looking at the same fleet see the same numbers.

Safety rankings by category

For each of five categories (unsafe driving, hours of service, driver fitness, drugs and alcohol, vehicle maintenance), FMCSA publishes a measure for every carrier each month: violations found at roadside, weighted by how serious and how recent they are, per inspection. 90 Out ranks that measure against every U.S. carrier with a similar number of inspections, using the groups and minimums in FMCSA's Safety Measurement System methodology (version 3.21). "Worse than 80%" means 80% of the carriers in that group have a better measure.

The Fleet Report Card labels each category by quarter: "Top quarter" when fewer than 25% of similar carriers do better, "Bottom quarter" when 75% or more do, and "Middle half" in between.

Federal law keeps FMCSA from publishing these percentiles for trucking companies, so these are 90 Out's calculation with FMCSA's method, not official FMCSA scores or alerts. A fleet with too few violations to rank shows no bar.

Two known differences from FMCSA's own ranking, because the public file lacks the fields: FMCSA drops carriers whose violations are all older than 12 months, and it groups unsafe driving by violation count and fleet type.

Out-of-service rate

The share of a fleet's roadside inspections in the last 12 months that found at least one out-of-service violation, compared with the same share for every roadside inspection nationwide.

Better or worse than last year

We compare the last 12 months with the 12 before. A change is called only when a standard statistical test says it is too large to be chance at 90% confidence: a two-proportion test for the out-of-service rate, and an exact test for crash counts. Otherwise the fleet shows "No clear change". Under 10 inspections in either year shows "Not enough inspections".

Shop Score (90 Out estimate)

An estimate of how likely a fleet is to change liability insurers at its next renewal, from signals such as its insurer pulling back from trucking, years with the same insurer, past switches and its safety trend. It is a 90 Out model, not a fact about the fleet. Every score is checked against FMCSA filings after the renewal passes, and the switch rate by score range is tracked.

Sources: FMCSA Safety Measurement System methodology v3.21 (June 2026); FMCSA SMS, inspection, crash, census and insurance files on data.transportation.gov.