Driver-Risk Study Finds 10% of Drivers Linked to 47% of Crashes
A Samsara analysis found that the top 10% of drivers ranked by its risk model accounted for 47% of crashes, sharpening the case for targeted coaching built around recurring behavior patterns.
A new analysis from Samsara found that the top 10% of drivers ranked by its risk model accounted for 47% of crashes. The top 30% accounted for 76%, suggesting that a relatively small group can absorb most of a fleet safety team's coaching time.
The model evaluates roughly 50 factors across driving behavior, exposure, operating context and driver development. Samsara says it ranks relative risk; it is not intended to predict whether a particular driver will crash on a particular day.
Combined behaviors carry a stronger signal
The report found that mobile-phone use alone made a driver 2.7 times more likely than the overall driver population to land in the highest-risk tier. Mobile use paired with harsh braking increased that figure to 4.5 times. Mobile use, distraction and harsh braking together pushed it to 5.4 times.
The analysis also found that the model ranked a driver who later crashed above a driver who did not about three out of four times, across both next-day and seven-day evaluation windows.
Use the score as a queue, not a verdict
The useful fleet-management idea is prioritization. A composite risk ranking can help safety teams decide where to review video, confirm context and schedule coaching first. It should not become an automatic disciplinary score or replace investigation of mileage, route, weather and job exposure.
Fleet platforms including Samsara, Geotab, Motive, Fleetio and Proaction increasingly bring multiple safety and operating signals into one workflow. The quality of the intervention still depends on what managers do next: validate the event, coach a specific behavior and measure whether that pattern changes.
