How would you increase the average Uber driver rating in your area from its current 4.5?
- Marty Ross
Clarifying Questions
- When we say Uber Driver rating, is it the rating provided by the riders living in a certain area to the drivers who are picking up and dropping them off? – Yes
- How long do we measure the average driver ratings – daily/weekly/monthly – Do it monthly to get a sense of the overall rating
- Is 4.5 an aberration in terms of the rating, as in where does this locality stand in comparison to other localities in the city – It is somewhere in the middle
- Do we know the reasons why users are giving bad ratings to certain drivers – there is no information
Structure of the response: Increasing the average rating can be done in multiple ways:
- Increase the number of 5-star rides
- Reduce the number of lower-rated rides
- Convert badly rated rides to high rated ones
I will focus on reducing the number of lower-rated rides in the following approach:
- Identify patterns for lower ratings
- Explore further reasons why a certain issue could be happening.
- Discuss a few solutions about how this problem could be solved
- Prioritize the solutions in a certain order
Patterns for Lower Ratings
- Drivers who have consistently received lower ratings from the folks in this area
- If it is not driver-specific, I will further drill down to understand the reasons for lower ratings by various factors
- Ride Length – Short rides or Long Rides
- Ride Type – UberGo, UberX, UberXL etc
- Car Type – Older cars vs Newer cars
- Ride Time – Morning rides vs Evening rides – Yes
- Ride by Day of the Week – Weekday vs Weekend Rides
Problem Evaluation – It has been noticed that morning rides are consistently getting lower ratings as compared to evening rides.
Common reasons why riders provide a poor rating:
- Before Booking the Ride
- No cars available in the area
- Have to try multiple times to book a ride
- Waiting for the Ride
- A mismatch between wait time estimate and actual time
- A mismatch between estimated ride length and actual length
- A high number of multi-contact rides
- Precision pick up
- During the Ride
- Mismatch in predicted route and a new route
- High traffic leading to longer ride time
- Broken or Unsanitary Cars leading to poor ratings – Air conditioning could not be working
- Chatty drivers
- After taking the Ride
- Precision Drops
- Issues with payments
These are the various reasons why this could take time, but I will consider the top reasons why this could be happening:
- Multiple contact rides –
- Training the drivers to better use the app to see
- Termination of drivers if they have many multi-contact rides
- Precision pick-ups – Adding more pick up points so that the drivers and riders can meet at the point easily. Notifying the user about the distance he or she needs to walk before taking the ride

Lyft
Uber