Problem Statement

Background

This task is all about improving an existing product, Meals on Bikes, a food delivery service where people can order from a choice of restaurants online and have the food delivered to their home.

Comments from our feedback form

• “I’ve tried Meals on Bikes three times now and every time my food has been late! I’m totally fed up. You suck.”
• “Why is it that every time I order from your site I get stuck with extra fees I didn’t know about?!”
• “Your service is pretty good but it always takes longer than the ‘expected delivery time’”
• “There are 2 pizza places in my neighborhood that have Meals on Bikes signs in the window but they never appear in the list when I go to order. What the?!”
• “That guy, Barry, who delivered my last order was a total jerk. I’m never ordering from you again.”

From our analytics (per week)

• 75k home page views
• 30k order page views
• 1100 orders
• 95 refunds

What the Support reps say

• We’re getting a lot of complaints about late deliveries
• Customers are confused about whether their order has been dispatched or not
• It can take up to 7 hours for us to get back to customers

What we’d like you to do

Come prepared to spend 15-20 minutes taking us through:

1. What you make of the feedback
2. How you’d identify and decide what problems to solve first
3. How you would onboard the team that’s going to work on it
4. How you decide the features you would include, how would you approach release planning and why
5. Describe any check-in points or other notes

Final Presentation

https://www.canva.com/design/DAGY13Ce2Ls/LrXznmRqStIWQ5UrsOffbg/edit

Detailed Solution

Approach

Output for slide 1:

• Base slide with problem statement, assumptions taken details of me, the solver.
• Nomenclature or slide format (Colors, section, etc.)

Title:

Step 1: Identify customer problems

Thoughts on this:

• Correlate customer feedback, user journey funnel data of clicks and visits and support reps’ feedback to identify an overall list of problems
• Create a model customer journey and map these problems to various areas of customer journey
• Identify metrics to check to validate the problems

Output for slide:

• One simple strip with customer journey, with a layer of problems mapped on it and a sticky note of metrics to check for each problem’s validation

User journey:

Customer discovers app → Customer downloads app → Customer completes onboarding → Customer searches for the relevant food items and adds them to cart → Customer goes to order placement flow, selects coupons, payment options → Customer makes transaction → Restaurant accepts the order and starts preparing it → Delivery partner is allotted and reaches the restaurant → Delivery partner picks the food order → Delivery partner delivers the food order → User receives the food order → User eats the food → User provides feedback → Churn / Repeat

User journey funnel and drops:

Add to cart rate (Homepage to order page CTR) → 30k/75k → 40%

Transaction rate (Order page to orders) → 1.1k/30k → 3.6%

Refund rate (Orders to refunds) → 90/1100 → 8.1%

Problems identified and sources

Customer journey	Problems	Location of symptoms	Additional checks
Search for relevant food items	Lack of restaurants	Customer feedback: – “There are 2 pizza places in my neighborhood that have Meals on Bikes signs in the window but they never appear in the list when I go to order. What the?!” Add to cart rate → 40%	Metrics: # clicks to ATC, scroll depth to ATC, top failure keywords
Customer goes through order placement flow	Confusion about final prices	Customer feedback: – “Why is it that every time I order from your site I get stuck with extra fees I didn’t know about?!” Transaction rate → 3.6%	Metrics: # customer issues related to transaction fees, difference in amount of dishes vs total payout
Delivery partner allotment → pick up → Delivery	Late delivery	Customer feedback: – “I’ve tried Meals on Bikes three times now and every time my food has been late! I’m totally fed up. You suck.” – “Your service is pretty good but it always takes longer than the ‘expected delivery time’” Support rep feedback: – We’re getting a lot of complaints about late deliveries Refund rate → 8.1%	Metrics: Delivery time, delivery distance, promised vs achieved time, # of customer issues correlating highest with deliveries
Delivery partner allotment → pick up → Delivery	Poor delivery service	Customer feedback: – “That guy, Barry, who delivered my last order was a total jerk. I’m never ordering from you again.” Support rep feedback: – Customers are confused about whether their order has been dispatched or not Refund rate → 8.1%	Metrics: Delivery avg rating, Delivery distance vs order repeat rate, delivery guy vs delivery rating
User eats the food → Provides feedback	Poor customer service	Support rep feedback: – It can take up to 7 hours for us to get back to customers Refund rate → 8.1%	FRT, TAT, Customer happiness score post ticket resolution

Assumptions:
– Only does food delivery, not courier, etc.

Step 2: Map the problems to their sources and verify them

• Build hypotheses on potential causes of each problem and map them to the user journey
• Define metrics needed to capture and identify these problem sources
• Identify the relevant stakeholders and speak with them to validate the cause of problems

Customer journey	Problems	Potential sources	Metrics to check
Search for relevant food items	Lack of restaurants	– Higher commission rates – Slow / Stuck onboarding – Poor recommendation engine – Poor search capability – multi keyword, multi lingual, etc.	–
Customer goes through order placement flow	Confusion about final prices	– Poor UX design – Special cases of cutlery being included – Compulsory tip being added	–
Delivery partner allotment → pick up → Delivery	Late delivery	– Poor allotment logic – Very high order clubbing – Poor traffic conditions / seasonality	–
Delivery partner allotment → pick up → Delivery	Poor delivery service	–	–
User eats the food → Provides feedback	Poor customer service	– Bad process – Low service providers	–

Step 3: Prioritize the problems based on impact

• Layer the initial funnel with additional metrics to find where the largest drop is in terms of final revenue

Output:

• Assumed values of metrics
• Final goal metrics – Customer satisfaction, Avg revenue
• Prioritization of problems based on impact on goal metrics

Step 4: Onboard teams to sensitize, solution and execute

• Identify stakeholders to onboard based on the user journey
• Onboarding goals – Sensitize about the problems, brainstorm ideas on solutions and discuss potential solutions and their feasibility
• Onboarding process
  • Invite for a group dinner and place orders and see issues being faced
  • Identify metrics that are owned by each team and then send a pre-read on these problems and invite to solve them

Output:

• Business stakeholders (Delivery, Food supply), Tech stakeholders (DS, EMs, Devs), Support and ops stakeholders, Product team – Make design of all people this represents
• Add a meme of 12 great men or some other movie reference but with a TV with a PPT in the screen
• Pre read document with:
  • What metrics
  • What agenda
  • Who is needed
  • What’s expected homework

Step 5: Identify features to build

• Identify the features to build:
  • Get a list of features from the previous onboarding exercise
  • Find common solutions built in the industry
  • Do top down thinking and find solutions
  • Ask customers for solutions

Output:

• Some meme to show that a single problem is being pulled at by a lot of people
• Table of problems with:
  • Potential solutions
  • Design of solutions or inspiration of it from other companies, etc.
  • Feasibility concerns
  • Adoption concerns
  • Potential impact scope
  • Expected support / dependency
  • Source of solution
  • Final prioritization of features to experiment

Step 6: Experimentation, validation and detailing of features

• Build feature mock-ups, designs or wireframes
• Speak with customers to validate feature ideas
• Run experiments to validate POC:
  • Hypothesis
  • Cohort design based on hypothesis, user groups
  • Success metrics and criteria
  • Experimentation hygiene – Minimum size, timeline
• Finalize top features and create detailed PRDs. Components to include:
  • To be planned

Output:

• Sample designs / wireframes of one or two features
• Customer questions list – Take pointers from that book of how to ask questions

Step 7: Launch

• Create a rudimentary adoption plan
• Get effort estimate on the features:
  • Bandwidth estimate from development teams and discussion on feasibility
  • Adoption effort estimate rom development team and adoption effort estimate from other teams based on a rudimentary adoption plan
• Create development roadmap and include features in the roadmap
  • Group features into existing themes based on either user journey or development pods
  • Prioritize features based on RICE matrix.
  • Check bandwidth available to the relevant developers and allot accordingly
• Development cycle process:
  • Regular scrums
  • Resolve doubts, reprioritize and reduce scope in case of development delays
  • QA
  • UAT
  • Partial release

Step 8: Adoption

• Develop a rudimentary adoption plan right from the time of ideation
• Categorize in three parts based on timeline
  • Pre-development
  • Pre-launch
  • Post-launch
• Categorize based on type of action:
  • Internal team training
  • External customer education
  • Monitoring and feedback loop
• Identify adoption metrics to track and define co-owners for it