The Intervention Threshold Is 65. You Approved a Carrier at 73.

The broker I'm thinking of — I won't name them, they'll know who they are — spent four years approving carriers with mid-range BASIC scores without a second thought. Nothing dramatic. Not 95th-percentile disasters. Just carriers sitting in the 70s. Their working theory: the carrier had eighteen months of authority, no crashes on file, a Satisfactory rating, and enough tenure that a percentile felt like a rounding error.

Then a driver for MC-1247893, DOT-3567102 crossed the median on I-80 west of Council Bluffs at 3:14 a.m. and killed two people in an oncoming Camry. The carrier's Hours of Service Compliance BASIC had been at 73rd percentile for four months before the load tender. The broker had a SAFER printout in their carrier file, dated the morning of the tender, showing that 73rd percentile in the little bar chart at the bottom.

Six months later, in a deposition I'd rather not have watched, the plaintiff's expert opened FMCSA's Safety Measurement System methodology document to the page listing intervention thresholds. He read four words into the record.

Above the intervention threshold.

He asked the broker if he knew what the intervention threshold was for HOS Compliance. The broker did not.

The Number Nobody Tells You

FMCSA's Safety Measurement System — the thing that produces the BASIC percentiles you see in SAFER — is not just a ranking. It's a triage system. FMCSA uses specific percentile thresholds to decide which carriers get pushed toward compliance reviews, warning letters, and roadside enforcement priority. Above the threshold, you're on the government's radar. Below it, you're not.

For property-carrying carriers, the HOS Compliance BASIC intervention threshold is the 65th percentile. So is Unsafe Driving. So is the Crash Indicator. The other three BASICs — Driver Fitness, Controlled Substances/Alcohol, and Vehicle Maintenance — sit at the 80th percentile. Hazmat carriers get the 80th too. Passenger carriers are different again.

This isn't buried in a PDF nobody reads. It's on FMCSA's public SMS methodology page. Anyone with a browser can find it. And it means something specific: if your carrier is at 66th percentile in HOS Compliance, they are — by FMCSA's own published criteria — worse than 65 out of 100 comparable carriers, and the agency itself considers that above-average risk worth intervening on.

Most brokers I talk to have never looked at those numbers. They see "42nd percentile" and think low, "73rd" and think high, and they don't know where the line is that FMCSA itself uses. Plaintiff's counsel, post-Montgomery, absolutely knows where that line is. Because the Montgomery v. Caribe Transport II opinion opened state-law negligent-selection to freight brokers in every circuit, and a plaintiff's expert can now put a percentile and a threshold on the same slide, hand it to a jury, and ask a question you're going to have trouble answering.

What the HOS BASIC Actually Measures

Before I get into the litigation angle, the mechanics matter. The HOS Compliance BASIC pulls violations from three main regulatory buckets:

49 CFR § 395.3 is the driving-time limits — 11 hours of driving in a 14-hour window, 30-minute break after 8 cumulative hours, the 60/70-hour weekly caps. Every driver moving property is bound by these. Every roadside inspector who runs a Level I or Level II inspection is looking at these.

49 CFR § 395.8 covers the record of duty status. In the ELD era this is the electronic log, plus the supporting documents (fuel receipts, bills of lading, dispatch records) that let an inspector reconcile what the log claims against what actually happened. Falsification cites here. False log lines cite here. "Log not current" cites here.

49 CFR § 395.22 is the ELD mandate itself — the carrier's obligation to install a compliant device on each covered vehicle and to keep it functioning. ELD malfunctions, failure to maintain the eight days of unassigned driving records, and inability to produce ELD data on demand all cite under § 395.22.

The BASIC percentile is not a raw count. FMCSA weights violations by severity, groups them by inspection, normalizes by exposure (mostly by power unit counts), applies a time-weighting curve, then ranks the carrier against its "safety event group" — carriers with a similar number of inspections. So a carrier that's been inspected 22 times gets ranked against other carriers inspected 20-40 times. A 73rd percentile in that group means 27% of comparable carriers had a better violation profile. Not 27% of all carriers. Twenty-seven percent of carriers at their same inspection density.

That's the number the broker in my hypothetical printed and filed away without thinking about it.

What "Above the Threshold" Actually Signals at Load-Tender Time

Here's the load-tender question you need to be able to answer, because you'll eventually get asked it in a deposition:

When your carrier's HOS Compliance BASIC is above 65th percentile at the moment you tender the load, what did you do about it?

Options that hold up:

• You confirmed the carrier had no active out-of-service order, pulled the L&I insurance record, spoke to a dispatcher who could name the driver assigned, and documented all of it.
• You had a written policy that carriers above intervention threshold require additional review, and the file shows the review happened and the reviewer signed off.
• You added a monitoring alert so any new HOS-related enforcement event on that carrier would flag before the next tender.

Options that don't hold up:

• Nothing. You just tendered.
• "It's just a percentile." (The percentile is the number FMCSA uses to run their own enforcement program. Calling it "just a percentile" in a deposition is a bad idea.)
• "The score has been that way for months, so it's stable." (Stable at above-average risk is not stable good.)

The specific FMCSR reg that a plaintiff's counsel will lean on at load-tender time isn't in Part 395. It's in the Montgomery opinion and the state common law duty of care it revives. But the citation stack behind their expert witness will absolutely be § 395.3, § 395.8, § 395.22, and FMCSA's SMS methodology publication. Those four documents, together, let them paint a picture where you, the broker, had free access to a risk signal, filed it in your carrier file with a timestamp, and did nothing.

I don't want to be the broker in that picture. Neither do you.

The 65th-Percentile Carrier Isn't a Coin Flip

There's a mental model I hear from brokers who dismiss BASIC scores: "72nd percentile means they're a little worse than average. Not a red flag." That framing is wrong on two levels.

First, the scale isn't linear. FMCSA's percentile calculation is skewed by the fact that the worst carriers cluster at the top of the distribution. Moving from 50th to 65th percentile is a modest step; moving from 65th to 80th is a much larger practical step in violation density. A carrier at 73rd is not "modestly worse than average." They're materially worse in a way that's compounded by the shape of the distribution.

Second — and this is the part most brokers miss — the BASIC score doesn't capture what's happening in the cab right now. It captures what inspectors found in the last 24 months. Roadside inspection catches a fraction of actual violations. FMCSA's own eROADS and ELD-data studies suggest inspectors catch somewhere between 10% and 25% of ongoing HOS violations depending on the carrier's operating pattern. So a carrier at 73rd percentile has been caught more than the average. What's uncaught is worse.

I don't say this to be a doomer. I say it because when a driver has an at-fault crash after hour twelve of a fourteen-hour window, the ELD download will show it, the plaintiff's expert will show it, and the question in the deposition will be: what did you know about this carrier's HOS pattern before you tendered them a load with a 750-mile drive on a 14-hour clock?

The Dispatcher Call That Should Follow

If I'm approving a carrier whose HOS Compliance BASIC is above 65th percentile, I want to talk to their dispatcher before I tender. Not an email. A phone call. I want to know:

• Who is the assigned driver, and where are they starting from?
• What is the driver's available hours at pickup? (A dispatcher who can't answer this is a red flag on top of a red flag.)
• Does this driver typically run this lane on this schedule? (Novel routing under tight timing is where fatigue crashes cluster.)
• Are they using paper logs, an AOBRD grandfathered device, or a certified ELD? (The regulatory landscape here matters. Grandfathered AOBRDs were phased out December 2019 under § 395.22. Nobody should be running them today. If the dispatcher says "AOBRD," the answer is either they're confused or they're non-compliant. Both are relevant.)

This is what post-Montgomery due diligence looks like when the BASIC score has already given you a signal to look harder. A five-minute phone call that documents the driver identity, the hours available, and the ELD status is not a heavy lift. It is the difference between "the broker did nothing" and "the broker made a documented judgment based on specific verifiable facts."

Where DOTScreener Fits

I'll be direct about this because it's what I built the product to do: DOTScreener pulls the HOS Compliance BASIC and puts it in the Carrier Diligence Summary with the intervention threshold visible right next to the carrier's percentile. If they're at 73rd and the threshold is 65th, the flag is there. If they were at 45th six months ago and are at 71st today, the trend is there.

More importantly — because a single approval event isn't due diligence, it's a starting point — DOTScreener's Continuous Monitoring re-files the Carrier Selection Record every time a material change lands. New CS/A enforcement action, new OOS order, insurance lapse, BASIC score crossing an intervention threshold — the record updates and the timestamp updates. When plaintiff's counsel subpoenas your carrier file, they see a chain of dated evaluations, not one screenshot from twelve months ago that's already stale.

That's the paper trail. That's what "reasonable diligence" looks like when a court asks. And it's most of what stands between "the broker checked the carrier" and "the broker knew and didn't act."

How I Document This

The specific practice, for HOS Compliance BASIC specifically:

• At approval: Screenshot the SAFER snapshot or, better, run the screening through DOTScreener so the percentile is captured with the intervention threshold flagged. If the carrier is above 65th, note it in the carrier file.
• If above threshold and I'm proceeding: Log the additional verification. Dispatcher call notes with date, time, name of person spoken to, driver identity, and hours-available answer. If I asked about ELD status and got a clean answer, that's in the file too.
• Ongoing: Continuous Monitoring alerts on any HOS-related enforcement or new BASIC threshold crossing. When one lands, the file re-files with a new timestamp and a note about what triggered it and whether I chose to keep the carrier active.
• On tender: Fresh screen. Not a screenshot from three weeks ago. The file shows what I knew at the moment I tendered this specific load.

Four steps. Each one is timestamped. Each one is defensible in a room I hope you never have to sit in.

The broker in my opening story didn't do any of this. His file had one printout, dated the morning of the tender, showing a 73rd-percentile HOS Compliance BASIC. That printout, and his answer that he'd never heard of the intervention threshold, is why the case settled instead of going to trial.

The information was there. He just didn't know what he was looking at.

— Mason Lavallet

Founder, DOTScreener.com