Explore

Lyft case

Team

Team

Role

Value

Designer

Tech Lead

Product Manager

Luis Rodriguez

There are no rows in this table

⁠

Problem Statement

Currently only 60% of requested trips find a match with a driver on the Airport - Downtown route or vice versa, causing a 23% increase in users failing to find a match.

Need to maximize Lyft's utility in the next 12 months without increasing the prevailing rate.

Who is experiencing this problem?

Users who want to go from Downtown to the Airport or viceversa and are not able to find a driver, hence are getting a “Failed to find driver” experience.

These users will take a flight or will land into this airport. Also, this will consider users who go to the airport to pick up someone who just landed.

Why is it important to our users and our business?

This is important to these riders because it's the way they can get to their destinations on time without having to consider options other than Lyft.

What problems are you not intending to solve?

It is necessary to improve the experience when the user is requesting a trip but cannot find an available driver without affecting the net profit of the company.

What is the problem we’re solving?

Having the ability of scheduling a trip in advance determined by some variables that may change depending on the place and time in which the rider needs their trip so that the match rate increases without affecting the net earnings of the company.

At the same time, we need to find a mechanism that can reduce driver churn but even more user churn, which has been shown to drop 23% more if they don't find a ride. Why now?

“When I go to request a ride from the airport, I always request it at the same time in Uber and Lyft, depending on the application that works best for me in the city, I keep requesting rides for that one.” - Alexis Sanchez

The speed in which our application can find a driver when looking for a ride to or from the airport is a constant concern as it causes a large drop in users who usually end up requesting rides with our direct competitors.

Another constant complaint from the Marketing team is the cost of acquiring a new user and that since they don't get an available driver they end up not returning to the application.

⁠

image.png

⁠

This graph shows the growth in the cost of acquisition per client that has had an increase in the last 4 years, being one of the variables that most generates costs for the company. Consider CAC compared to 33% churn when users can't find a driver answer why now.

Current Experience Screenshots

Currently, users are able to schedule a ride from Lyft’s application, selecting day and time.

Setting the ride

⁠

image.png

⁠

Scheduling the ride

⁠

Hypothesis

Offering a solution that allows users to schedule their trips depending on the departure or landing time of a flight, will allow more previously scheduled trips, which will reduce the rate of users leaving Lyft and increase the match rates.

Experiment Idea (How are we going to do that?)

Overview

Users will have the ability of entering their flight number so that Lyft can know the status in real time and schedule a driver automatically a time before taking off or after landing determined by the user. The user will be able to determine the time after or before they want

More questions:

What information does the user give to Lyft? Lyft should be able to schedule the ride only with the number of the flight having some integrations with realtime flight status platforms.

How much time in advance can the ride be scheduled? Users will be able to schedule a ride 30 days before the original date of flight.

What happens is a flight got delayed? Departure: If the flight is delayed, Lyft will have already fulfilled its mission, which is to take the user to the airport. This can open an opportunity for when flights are cancelled, to send a push notification.

Landing: Lyft should be able to know in real time the status of a flight and scheduling or re-scheduling accordingly.

Can customers use this on all of their devices? Yes, Android, iOS and web platforms will be available.

Will this be available worldwide? Initially, this will be available in Toledo, Ohio, US. Which features will I care about the most? For this specific solution, users will be focused on the “Schedule a ride” feature. This option will be displayed there and user will only have to enter the flight number.

What is the most likely cause of failure if this fails (Risks): Users who landed in the city and for some reason took much longer than the estimated time in immigration, picking up suitcases, having a coffee, etc.

For this, it is necessary to send progressive alerts to the user that allow the driver to reschedule or request waiting. A good opportunity to request tips or charge for waiting by the driver.

Future customer quotes:

“Sometimes my flight is delayed, however I have the peace of mind that Lyft knows it and always when I arrive, regardless of the time, there will be a driver waiting for me.” - Korey Haid “I'm not sure what they're doing but for some reason now whenever I look for a ride to or from the airport it's easier to find a driver” - Amber White “When I go to the airport, I don't have to worry about finding a ride because I already have it scheduled depending on my flight and it always arrives on time.” - Alexis Martinez

Business case:

⁠

Opportunity Size

Absolute Number

Reduce in 13% the drop in users who experienced a “Failed to find driver”.

Match rate increase in 22%

Calculation for above

Total users in the US: 2’805.000 / monthly

Total users in Toledo, Ohio: 342,000

User who dropped (Total * % drop): From 112,860 to 68,400 users

Match rates currently at 60%, jumping to 82% at least.

Estimated Impact (What does success look like?)

This is going to have an impact in the long term when there is a general knowledge of the functionality on the part of the users. As we have more real-time data, this will not only allow rides to be scheduled in a better way depending on the status of a flight, but it will also allow us to have more data to make better decisions regarding the way ride scheduling is handled.

Metrics to watch

Positive Metrics

Feature adoption (Schedule a ride based on four flight)

User acquisition

Match rate

Neutral Metrics

User retention

Negative Metrics

Churn rate

Investment

Front End - (3 developers) High

Back End - (4 developers) High

Product - (1 Product Manager) High

Design - (1 Product designer) Medium

Tech lead (1 Tech lead) High

Privacy - Low

Marketing - High

Compliance - None

HR - None

Security - Low

Fraud - Low

Precedence (What has been done before?)

Uber has a similar feature in some specific cities that can be used for benchmarking

⁠

image.png

⁠

Your Suggestion

We should work on this project immediately

Want to print your doc?
This is not the way.

Try clicking the ⋯ next to your doc name or using a keyboard shortcut (

CtrlP

) instead.