Samsara report: top 10% of risk-ranked drivers account for 47% of crashes

Samsara has released a Compounding Risk Report that concentrates crash risk in a small group of drivers. According to FreightWaves, the top 10% of drivers ranked as highest risk by Samsara's Risk Model accounted for 47% of crashes. The Business Wire headline for the release rounds that to roughly 50%.

Arpan Podduturi, Samsara's VP of AI safety, told FreightWaves that 30% of drivers account for more than three-fourths of crashes, so "who you coach, who you zoom in on, really, really matters."

How the model works

The Risk Model evaluates more than 50 factors across driving behavior, exposure, situational context and driver development. These include speeding, harsh braking, distracted driving and mobile phone use, along with conditions such as night driving and weather.

Rather than scoring each safety event on its own, the model looks at how factors occur together. Samsara calls this "compounding risk." Its example is speeding, in rain, in the dark, at the end of a shift.

What the report found

Combinations of behaviors were tied to higher risk than single behaviors. Drivers with mobile phone use alone were 2.7 times more likely than the overall driver population to land in the model's highest risk tier. That rose to 4.5 times when phone use was paired with harsh braking, and 5.4 times when phone use, distraction and harsh braking all appeared.

For individual behaviors, FreightWaves reports these associations with crash likelihood:

• Harsh braking: 1.8 times higher
• Mobile phone use: 1.7 times higher
• Distracted driving: 1.7 times higher
• Speeding events: 1.5 times higher

Samsara said that when it compared drivers who later had a crash with those who did not, the model ranked the crash-involved driver higher about three out of four times. That held for both next-day and seven-day windows.

Limits

Samsara says the model ranks drivers by relative risk and does not predict whether a specific driver will crash on a given day. The findings are statistical associations and do not show that the behaviors caused crashes. The report uses aggregated data from July 1 through Dec. 15, 2025, covering multiple industries and regions.

Samsara also said night driving, freezing temperatures and urban exposure can amplify risk but are not coaching targets on their own.

Coaching Priority

The model powers a new Samsara feature called Coaching Priority. It puts risk factors into one view so managers can see which drivers, behaviors and conditions are driving elevated risk. Podduturi said fleets should coach on behavioral patterns, not isolated events, and decide who needs human-to-human coaching and where coaching can be automated.