It’s the time of year when holiday parties are everywhere, and we all have a bit of extra fun at night. So at Zendrive, we wondered: what’s the safest way to get home from the party?

Over the past several months, we’ve conducted a secret study on ridesharing, taxis, and for comparison, the average driver in San Francisco. Using Zendrive’s technology, which uses the sensors on a cell phone to measure a driver’s safety behaviors, like speeding and cell phone use, we took over 1,300 miles in the city, sticking to very limited routes. We sent secret passengers into Uber, Lyft, Sidecar, and taxi rides, and had their phone sensors track the safety of the ride. We modified our usual technology to account for the difference in motion of a passenger’s phone vs a driver’s phone, and let the secret passengers manually (and discreetly) log the driver’s risky and illegal use of their cell phone while driving. The taxi and rideshare companies and drivers had no knowledge of this study. For non-professional drivers, we used the unaltered Zendrive technology for measurement.

Here’s what we found:

Studies have shown that speeding is a factor in 31% of traffic fatalities. While everyone speeds sometimes, we found that taxis were by far the worst of the bunch, speeding about 50% more than rideshares. This was even worse during peak hours, when taxis sped about 2.5 times as much as rideshares did.

You’d think that with all this speeding, taxi drivers might just be better at it than other drivers. However, when we measured unsafe speeding – that is, speeding that included another unsafe behavior, like cell phone use, or hard brakes (which imply a location where speeding was even less appropriate than normal) – taxis were still 25% more likely to engage in it than rideshares.

While the average driver seems safer than taxi drivers in all metrics, the results are mixed when comparing them to rideshare drivers. (Full details below.) The average driver is better than taxis in all of these metrics, but sometimes drives more safely than rideshares, and sometimes not.

What’s encouraging about all of these statistics is that they’re only a snapshot of where we are today. With the availability of driving analytics, such as our Fleet Safety service, things could improve quickly. For the first time, many companies (both rideshare, taxi and delivery fleets) are capable of collecting objective and accurate insights to optimize their road safety.

We believe that five years from now, with the rapid growth in popularity of rideshare services, driver-analytics solutions will make collisions an exceedingly rare event.

See our results below, and if you’re curious, check out our full list of references.


  1. All data on safety violations like speeding for taxi, rideshare, and average driver rides from Zendrive study, 2014.
  2. Speeding fatalities data from the National Highway Traffic Safety Administration
  3. Taxi vs rideshare arrival times from App-Based, On-Demand Ride Services: Comparing Taxi and Ridesourcing Trips and User Characteristics in San Francisco, UC Berkeley, 2014
  4. Taxi vs rideshare insurance coverage information references from LyftUberSidecarSF Municipal Transportation Agency, and the New York Times
  5. Passenger screening information from UberLyftSidecar, and San Francisco Cab Drivers’ Association
  6. DUI tolerance information from UberLyftSidecar, and SF Municipal Transportation Agency
  7. Background check information from Uber, Lyft Customer Support (from representative Austin M.), Sidecar Customer Support (from representative Cassandra), and NBC Bay Area report on taxis