You are a PM at Uber or Ola. You have to upsell Uber Go to Uber Auto customers. How would you go about it? What product changes would you make?

CQ:

1. Define Uber Go - Hatchback based commutation

2. Goal of the exercise - Go is only slightly more expensive however has potential to improve upon the overall experience

3. Geography - India

4. Constraints & Timelines - NA/3 months

Approach:

1. Understand why people use Uber Auto and build recommendations on top of these

   1. Short distance

   2. Cheap

   3. Availability

   4. Used to taking auto from behavioural aspect

2. Recommendations around the same -> I’ll go with collecting reasons why people wish to book Uber Auto over Uber Go and give above reasons for selection and do below as per input selected by the user

   1. Behavioural issue

      1. Auto-upgrade to Uber Go for such customers

      2. Explain benefits of travelling in Uber Go

   2. Availability

      1. Show availability of Uber Go as well on App to influence decisioning

      2. Show overall trip time to influence decisioning

   3. Cheap

      1. Show the fare difference and explain the benefits in terms of safety and overall time saved to non-frequent traveller. Auto-upgrade as per budget

      2. In case the rider is a frequent traveller, talk about no hours saved/month/year. Auto-upgrade for frequent traveller for upto 3 trips/as per budget

   4. Distance

      1. Explain about the overall time saved and safety index of the trip

Approach should be to sell the advantages of Uber Go to users while booking and right till a cutoff time when a trip has been booked.

How to target users:

1. Do an A/B testing and track below metrics for the control group and rest population

   1. No of unique trips made

   2. % trip made using Go vs Auto

   3. Churn rate

   4. % people switching from Auto to Go as per suggestion made

   5. Impact on revenue

2. If the results for the control group is positive then scale the offering in phases basis geography or type of users

I will start with the basic understanding of the product. Uber is a platform whose mission is to make travelling smooth, Convenient and safe just like running water.

Clarification Question:

Why we want to do that?

Let’s say to increase revenue.

When you say upsell Uber Go to Uber Auto so basically we want to onboard uber Auto users to Uber Go platform.

Let’s say yes.

For this we need to understand the motivation behind using Uber Go and Uber Auto.

Persona1: Ravi is a 25 year old work in a company which is at 7 km from his locations. The office timing is from 10:00 AM So Ravi always book an auto from his location at 9:35 AM. It takes 20-25 minutes to reach at the office. One day he is 15 minute late and there is meeting that starts from 10:05 AM. He is the main presenter there. So reaching office on the time is important.

Strategy:

1.    1. New Algo+switch to Uber Auto button: Develop an algorithm which identifies the User behavior and whenever it notices a displacement (e.g. Ravi always take auto near 9:35 AM and suddenly he is late and books auto at 9:50 AM ) in that it should show use Uber Go to reach destination at the right time using Uber Go. There Should be a single tap option which allows you to book cars from Uber Go.

2.     2. Reward+Referral Program: Show incentives to uber auto users as get 30% off on your first 3-4 ride.

I would like to focus on the first one as it really shows that we are solving a critical problem. Second one is just like an incentive once user have taken the three rides there is no guarantee he is going to use Uber Go instead of Uber Auto. They will easily figure out that Uber Auto is cheap and affordable compare to the Uber Go.

Metric:

1.     1.  # of people who have Clicked on the Uber Go button.

2.    2.   # of people who have Clicked on the uber Go button and have completed the ride.

3.    3.   % increase in Uber Go rides.

4.   4.  User ratings + feedbacks.

Objective / Why do we want to upsell?:

• Higher margins in Go
• Auto rides have experienced comparatively higher user complaints, lower/negative ratings than Go
• Reduce rejections in Auto category(supply-demand mismatch)

1. Price difference
2. Supply - Autos are easily available with a lower wait time than Go cabs
3. Quicker than Cabs

Parameters for experiment:

1. Price difference (base charge, surcharge, rate/min, rate/km)
   1. Consider trips where the price difference between Auto and Go is less than 'x%' on the base of Auto fare.
   2. Experimentation - To begin with consider x to be 20. Based on user response
2. Trip distance - Compare the trip distance v/s trip volume between Auto and Go for a defined time period. If the preference of auto reduces as distance increases, then we can determine a threshold distance for trips which would be eligible for the upsell.
3. Supply - At the time and location of trip request, compare the ratio of trip to Autos with trips to Go. Upsell whenever the ratio is better for Go.
4. Previous affinity for Go - Based on the previous usage of Go cabs, we can extend marginal discounts to users (<5%). The value of discount given would be on the basis of the category in which a user falls:
   1. Never used Go
   2. Used Go at an avg of <0.5 times a day in the past month [ Don't compare volume. Normalise the ride quantity metric]
   3. Used Go at an avg of 0.5-1 times a day in the past month
   4. Used Go at an avg of >1 times a day in the past month

1. Budget constraints affecting the volume of discounts that can be extended
2. Resource constraints to build and scale the experiment system to operate at a ride level
3. Complexity of using the multivariate experiment system to optimise results

• Phase 1
  • Consider only 'trip distance' parameter. Define threshold on the basis of historical data analysis and iterate on it for a short period to get a high success rate. This would validate the user base which responds to reduced price variation and which doesn't. The users who respond positively need not be extended a discount while the ones who don't, can be evaluated on a different/additional set of parameters.
  • Frontend changes - None. Promote Go by showing it above Auto even if price is higher. To be more aggressive, hide Auto from top options list.
  • Pros - Minimal development effort, no marketing spend/increased CAC
  • Cons - Constant recalibration would require frequent manual effort
• Phase 2
  • Consider 'price difference' and 'previous affinity for Go' parameters. Price difference can be determined in real time and the user base for affinity can be determined based on analysis for the past month hence, a static list. Add these parameters for users who did not respond positively to previous phase.
  • Frontend changes - None. Promote Go by showing it above Auto even if price is higher. To be more aggressive, hide Auto from top options list.
  • Pros - Builds on additional knowledge, no marketing spend/increased CAC
  • Cons - Dev effort needed to do run-time checks
• Phase 3 - Evaluate the performance and do more analysis before implementing the supply parameter. That may require a lot of effort to build.

1. Sanity metrics - The metrics which should not be negatively impacted while experimenting.
   1. Number of successful trips
   2. Reject to request ratio for Go
2. Success metrics - Success rate of upsell request getting accepted, started, completed
3. Inform business leads managing Go and Auto categories about the experiment so that they are aware about the experiment.
4. Inform teams tracking reject to request ratio for Auto and Go categories about the experiment.
5. Pass an identifier with the trip requests which were part of the experiment.
6. Maintain a control user base whose ride requests are eligible for the experiment so that we can compare the sanity metrics such as conversion etc.

Requirements for a solution to a problem gather from 4 sources:

EMUC: Employees, Metrics, Users & Customers (users who generate revenue and are decision makers)

First, Ill try to know what sources the problem has emanated from:

1. Employees - Is it because of the feature problems from the development team to pursuade an UberAuto sale? - No

2. Metrics - Is is because of high customer churn , low revenue and low profits from UberAuto feature?- Yes

3. Users -  Users find it costly to travel UberGo? - Maybe Yes

4. Customers - Is B2B revenue not much profitable to corporate partnerships? - No

After filtering, it turns out that Metrics and User dissatisfaction are the factors responsible for the problem in this case study whivh needs an effective Value Capture Pivot (The Lean Startup).

Metrics - For customer churn, i would try retention programs like:

1. Incentivize the travellers with coupon codes for movies, Baskin Robins, etc. 

2. Charge them same for UberGo for first 5 rides so that they become accustomed to UberGo comfort through classic Loss Aversion (UX behavioural psychology) way.

Users- 

1. Create user-personas after segmentation of UberAuto users to ascertain what reason could entice their personality to come onboard with UberGo. Example - A student who doesn't earn should be able to book an UberGo at their University gates,libraries, museums, amusement parks, cinema theatres.-  Personalised Advertisment and Increase avalaibility

2. API integration with cinema theatres to know end timings of movies so that the availability can be made just at the ending of movies in a timely manner

3. API integration with Airlines and Ralilway authorities to know the arrival timimgs to make seamless availability of UberGo cabs.

4. Make the ride interactive and enjoyble by integrating Wifi, Netflix, JioSaavn, etc during the ride and inform the users about the feature to leverage against UberAuto.

5. Seasonal Ad-Hoc rides - If its too much cold, raining or maybe the loo season, make UberGo create a bandwagon effect (UX psychology) that most of our customers choose UberGo with these benefits during the extreme weather seasons.Also, tell them about the ambient conditions inside ther UberGo.

Before everything... Learn.
1. What are the benefits of UberGo Vs. other transportation solutions or Uber X
2. What Are the needs of UberGoers: From social needs to physical needs.
3. What are the restrictions (From GDPR to know about the other person to max people inside an Uber)

With Data, test mockups.
1. Would you go to UberGo if there was a gamification (Guess the drivers name, for example)
2. Would you care to earn points everytime you order UberGo?
3. If you go UberGo.. You can pick the song on the car!

After you'll find what users want.. Just do it!
1. Soft launch with exciting UberGoers
2. Official launch to any UberGoers
3. Launch this to non-Uber-go to see if you can "convert" them :-)

 

What is the motivation for Uber to move customers from one product offering to another? 

possible reasons I can think is 

1. increase user experience & safety, which can lead to customer retention and stickiness.

2. increase in gross revenues per ride.

let us understand possible reasons for customers choosing Uber Auto

1. Lower cost of transport

2. Shorter wait time (since autos are available in every nook and corner)

3. Emergency

4.one or two passengers/travelers

5. faster ride (since this is a smaller vehicle which can accommodate small lanes/ densely populated areas)

6. Local drivers

With all the above factors, it's important to get the current 'Uber Auto' ride data - and measure factors like 

1. Ride distances

2. Ride frequency per user per day/ month

3. Customers requesting for UberAuto ride, after one or few failures attempts to acquire UberGo or other services

With all the above data, we need to understand what is the percentage(monthly/ weekly) of a customer using Auto ride Vs UberGo and others?

Proposed plan

1. Promote Saftey - add some kind of safety logo for all UberGo and above product offering (Since cars are all safety certified and equipped with Function Safety features)

2. Offer a discount promotion to a targeted customer (who has more than a certain percentage of UberAuto use in their over Uber rides). 

3. NO cancellation policy - Ensure Drivers cannot cancel the UberGo requests for short-distance rides

4. Reward Program for frequent users of UberGo, irrespective of ride distances 

5. Urgency flag while looking for a ride - especially if customers are booking a ride for hospitals, airports, bus stands, railway stations, etc.