“Matching platform” means the feature of Uber RideShare which match Drivers and Riders when Riders make a booking

UberX is the standard of Uber Rider Share

“Describe and assess” means how to measure the success of the feature

Match drivers and riders based on

• Distance

• Availability

• Ride density of driver

Increase user satisfaction

User = Rider

Before matching

• Riders dind and pick a destination and pick up location on the mobile app

• Riders request a ride

While matching

• System finds a driver to match with the booking

• Riders wait until driver accepts the request

After matching

• Riders receive confirmation from system and drivers

• Driver drives to pick up riders

Before matching

• Ratio of Demand (Riders) / Supply (UberX drivers)

• # of zeros (no available UberX drivers nearby when opening apps)

• Avg time before booking

While matching

• Avg matching time

• # and % of successful booking at the 1st, 2nd, 3rd, …., nth time

• # of drivers’ decline per booking

After matching

• Avg waiting to be picked up

• Avg distance from driver to rider

• Rider satisfaction rate

• Driver satisfaction rate

Criterias to evaluate

• Impact to our goal (Increase Rider satisfaction)

• Complexity to measure

• Risk (is it easy to be skew?)

Before matching

• Ratio of Demand / Supply => Medium / Low / Low

• # of zeros (no drivers nearby) => High / Low / Low

• Avg time before booking => Low / Low / Low

While matching

• Avg matching time => Extremely High / Low / Low

• # and % of successful booking at the 1st, 2nd, 3rd time, …. => High / Low / Low

• # of drivers’ decline per booking => High / Low / Low

After matching

• Avg waiting to be picked up => Medium / Low / Low

• Avg distance from driver to rider => Low / Low / Low

• Rider Satisfaction rate => Medium / Medium / High

• Driver Satisfaction rate => Medium / Medium / High

# of zeros

Avg matching time (North star metrics)

# and % of successful booking at the 1st, 2nd, 3rd time, …

# of drivers’ decline per booking