Carrier vetting 101. Spoiler: It’s not about safety.

Three words that are not synonyms

Almost every conversation I have about this runs aground on the same problem. Safe, compliant, and risky are three different words, yet the industry uses them as if they mean the same thing. They do not.

Compliant means a carrier meets the minimum standard the federal government will accept before it lets that carrier keep its operating authority. Compliance is a floor. It is pass-or-fail, and the bar is set low on purpose because the Federal Motor Carrier Safety Regulations are a national minimum, not a best practice. When a carrier is compliant, what you actually know is that, on the day someone checked, it was not so far out of line that the government would withdraw its authority. That is the whole guarantee.

Safe describes an operational outcome. Did the truck get from point A to point B without hurting anyone or damaging the freight? Safe is a backward-looking track record, mostly a function of things hard to see from the outside.

Risk is the one almost nobody defines, and it is the one that matters. Risk is exposure. It is the amount of financial and human damage a carrier is capable of causing, the likelihood it will cause it, and the question of who pays when it does. Risk is forward-looking, probabilistic, and encompasses an entire category of factors unrelated to safety or compliance.

Here is the part people miss. A carrier can be compliant and unsafe. A carrier can be safe and out of compliance on paper. A carrier can be both safe and compliant and still be a genuinely bad risk. After Montgomery, when that carrier causes a catastrophic crash, safe-and-compliant is not the shield people think it is, because the plaintiff’s lawyer is not going to argue compliance. He is going to argue risk, and that you could have seen it coming.

Risk 101: frequency and severity

If you have ever sat across the table from an underwriter, you already know how they think. Risk has two axes. Frequency and severity. How often something bad happens, and how bad it is when it does. Multiply them and you get expected loss, the actuarial spine of every commercial auto premium ever written.

Now notice what is not in that equation. There is no line item for safety rating. There is no field for compliant, yes or no. Those things might correlate loosely with frequency. They tell you almost nothing about severity, and severity is where companies die.

When an insurer retains me to assess a fleet, I am not grading the fleet’s character. I am estimating frequency, estimating severity, and pricing exposure. Carrier vetting, done correctly, is the exact same lens applied to a hiring decision rather than a policy.

The FMCSA data trap

So let us talk about the data the industry actually leans on. FMCSA assigns exactly three safety ratings, and only after an on-site compliance review. Satisfactory, Conditional, Unsatisfactory. That is the entire scale.

Here is the number that should reframe the conversation. Roughly 92 percent of carriers have no rating at all. They are Unrated. Out of something like 780,000 motor carriers, the overwhelming majority have never had an auditor walk through the door.

So when a broker or a shipper pulls up a carrier and the rating field says Unrated, that does not mean safe. It means nobody from the federal government has looked. The absence of a bad rating gets read as a good rating. It is not one. Unrated is not a clean bill of health. It is an empty file.

Satisfactory has its own problem: time. A satisfactory rating can be six, eight, or ten years old, indicating a company that may no longer be functioning. Conditional, the rating that does carry weight, carries it in the wrong direction. It means FMCSA has already found serious deficiencies. That was the exact fact pattern in Montgomery. Caribe carried a conditional rating. That is not a coincidence in this case. That is the case.

SAFER is a thin, lagging, mostly empty dataset that was never built to measure exposure. It was built to help FMCSA triage which carriers to investigate. The industry borrowed it as a hiring tool because it was free and available. Free and there is not the same as adequate.

What actually goes into a risk profile

If FMCSA data is a thin slice, what is the whole picture? A real 360-degree carrier risk profile goes well past SAFER.

It starts with financial health, because a carrier under financial stress makes predictable choices. It defers maintenance, stretches trade cycles, cancels the telematics, and lets the safety director’s seat sit empty. None of that shows up as a violation until after it shows up as a crash. It includes culture, as OSHA history will tell you, because how a company runs its shop tells you how it runs everything else. It includes corporate structure, the shared officers and addresses and recycled authorities that expose a chameleon carrier, the operation that lets a bad authority die and reincarnates clean under a fresh DOT number. It includes federal litigation history, one of the loudest risk signals there is and one almost nobody reads before tendering a load. It includes catastrophic exposure and the PHMSA incident history, which captures the low-frequency, high-severity tail risk that ends companies.

Almost none of that is compliance data. Most of it is not safety data. It is risk data and public. The reason most carrier vetting does not look like this is that very few people in this industry have actually sat in the loss control chair.

That is why my partner and I built Tea Technologies and the CarrierVerifi platform to consolidate that work: doing it by hand across a dozen separate public systems for every carrier you touch is not realistic, and not realistic is not a defense a jury will accept anymore.

The vetting process in brief

If safety ratings are not the standard, what is the actual process? Think about it in three phases.

Before you tender, run a real pre-tender screen. Confirm active authority through live FMCSA data, not a number somebody emailed you, and match the MC and DOT numbers, legal name, phone and email domain to catch double-brokering and identity theft. Treat authority under 12 to 18 months as a reason to look harder, not a reason to look away. Verify insurance properly, current, in force, limits that actually fit the exposure, and get the certificate from the agent, not the carrier. Read the FMCSA data correctly, weighting a high Unsafe Driving percentile differently than a paperwork category, and weighting a fatal crash differently than a tow-away. Then go past FMCSA into the financial, corporate, litigation and PHMSA picture.

Make the process itself defensible. Have a written carrier vetting SOP, follow it the same way every time, and document that you did. In discovery, the first thing a plaintiff’s attorney requests is your vetting policy and the file on the carrier that crashed. An undocumented process and a nonexistent one look identical to a jury.

Monitor continuously, because a risk profile is a film, not a photograph. The carrier you vetted in January can have lapsed coverage, three new crashes and a downgraded rating by June. Track trajectory, not just absolute scores. A carrier whose violations are climbing poses a different risk than one trending down.

This carries down to the driver level, too, where compliance is again just a floor. A once-a-year MVR pull is compliant, but it means a February DUI stays invisible until the following January. A road test is not federally required, but if you skip it, the question before the jury becomes whether ten minutes of riding down the road to confirm that a driver could operate the truck was too much to ask to keep someone alive. The gap between the minimum and the defensible standard is the entire point.

The insurance blind spot

There is one piece almost nobody puts in a risk profile, and it may be the most expensive omission of all. Insurance itself.

Freight brokers are not federally required to carry liability insurance. What a broker must have is a $75,000 surety bond, the BMC-84, which exists to ensure carriers and shippers get paid, not to cover a wrongful-death verdict. Kavanaugh flagged exactly this gap in his Montgomery concurrence. A large share of the brokerage world is mom-and-pop, single-member LLC operations with no office, no assets and no liability policy. Montgomery made those brokers liable. It did not make them collectible. When the named defendant has nothing, the exposure travels up the chain to whoever does have coverage and assets, which increasingly means the shipper.

Even among companies that do carry insurance, a policy is not the same as protection. Two things go wrong. Limits, because catastrophic verdicts land above policy limits and somebody absorbs the excess. Language, the exclusions, and sublimits are buried in the document. I worked on a cargo loss on a 1.5 million dollar piece of oilfield equipment that came off an open-deck flatbed. A one-million-dollar cargo policy was in place. It looked adequate. It was not. The policy limited open-deck cargo damage to 150,000 dollars, and the shipper absorbed the rest. Nobody read the open-deck language until the claim was filed.

Insurance belongs in a risk profile, but not as a yes-or-no field. The real questions are what the limits are against the realistic severity of a loss, and what the language does when a claim is actually filed.

The right question

So when the industry keeps asking what makes a carrier safe, I would push back. Safe is a backward-looking outcome; compliant is a minimum floor; neither tells you what you need to know before you put a carrier on the road with your name attached to the load.

The question that does the work is this. What is this carrier’s risk profile, and how much exposure does it carry, to the motoring public, to its insurer, and now, after Montgomery, to the broker and shipper who selected it. And the companion question a deposition will eventually ask you directly: how good is your own process for vetting, qualifying and screening carriers, and would it survive an afternoon of cross-examination.

I get to be on television and write columns, and I value that. But the day job underneath it is this. I am an expert witness in commercial motor vehicle litigation and a loss and risk control advisor to insurers, underwriters and the attorneys who try these cases. They taught me a long time ago that risk, not safety and not compliance, is where this whole thing actually lives.

Safe is not always exposure-free. Compliance is not always safe. What “safety” means does not matter. Risk is the only one of the three that decides who pays.

I have published a longer working primer that walks through the full three-phase vetting process, the driver-level program, and the 120-question risk control assessment we use in the field here.