Why Uber’s CPO delivers food on weekends | Sachin Kansal
Sachin Consul, CPO at Uber, discusses his extreme dogfooding practice, driving and delivering hundreds of trips to inform product decisions. He emphasizes a "ship, ship, ship" mentality, PM career advice, Uber's AV strategy, and tips for riders/eaters.
Deep Dive Analysis
16 Topic Outline
Introduction to Sachin Kansal and His Dogfooding Approach
Sachin's Personal Dogfooding Routine and Scale
Building Empathy and Understanding Drivers Through Experience
Balancing Quantitative Metrics with Qualitative User Experience
Operationalizing Dogfooding and Fostering a Culture of Empathy
The 'Ship, Ship, Ship' Mentality: Importance of Rapid Delivery
Why Live Demos are Crucial for Product Announcements
Career Advice for Early-Career Product Managers
Evolution of Product Management with AI: Constants and Changes
Collaboration Between Engineers and Product Managers
Uber's Vision for Autonomous Vehicles and a Hybrid Network
Uber's Path to Profitability and Sustaining Innovation
Balancing Data-Driven Decisions with Gut Instincts
Leveraging AI Tools in Product Management
Lessons Learned from Palm's Decline and Career Failures
Final Thoughts and Lightning Round
5 Key Concepts
Dogfooding
Dogfooding involves personally using and experiencing your own products as an end-user, often from various user roles like a driver or delivery person. This practice helps product teams uncover inefficiencies, build deep empathy, and improve product quality by experiencing visceral reactions in real-world conditions.
Ship, Ship, Ship Mentality
This product development philosophy emphasizes the rapid delivery of functional code into the product to quickly impact end-users. It focuses on minimizing the cycle time from identifying a problem to deploying its solution, rather than mindlessly shipping, ensuring solutions reach users as soon as possible.
Product Sense / Judgment
This refers to an innate ability to make correct product decisions, developed through extensive experience from shipping many products and making countless micro-decisions. It's a refined 'gut' feeling that distinguishes good product managers from great ones, becoming increasingly vital in an AI-driven environment.
Hybrid Network (Uber AVs)
Uber's strategy for autonomous vehicles (AVs) integrates AVs from partner companies into its existing human driver network. This hybrid approach aims to optimize vehicle utilization and efficiently fulfill demand by leveraging human drivers during peak times or in areas where AVs are not yet fully scaled.
Concentric Circles Portfolio Strategy
This framework for managing a product portfolio places the core product (e.g., Uber rides/delivery) at the center, requiring constant attention for flawlessness and efficiency. Successfully managing this core grants the 'license' to expand into outer circles, representing new growth bets and innovations.
11 Questions Answered
Sachin regularly takes 5-10 Uber rides and places 3 Uber Eats orders weekly. Additionally, once or twice a month, he dedicates half a day to driving and delivering as an Uber driver/courier, completing 10-12 trips each time.
It provides a visceral understanding of the user experience, revealing how features designed in an office feel in a real-world, moving environment, and builds deep empathy for drivers and couriers by experiencing their daily challenges and emotions.
Uber leaders model the behavior, centrally organizes sign-up processes, runs quarterly driving/delivering competitions for employees, and sets six-monthly OKRs for teams to fix hundreds of dogfooding-identified issues.
Always ask for permission before taking a phone call in the car, and when exiting, close the car door softly rather than slamming it.
Uber is building a 'hybrid network' that integrates autonomous vehicles (AVs) from partner companies with its existing human driver network, aiming to optimize utilization and fulfill demand across various cities and conditions.
The company embraced a 'grind' to identify and remove inefficiencies, particularly by optimizing core marketplace logic like batching delivery trips, which allowed savings to be reinvested and contributed to sustained profitability.
While obsessing over metrics and understanding friction points is crucial, numbers alone cannot capture user emotions. A strong product culture allows for the coexistence of both quantitative analysis and qualitative, emotional insights to drive better product decisions.
Seek roles where you can ship multiple products quickly to gain experience and develop product judgment, as greatness comes from thousands of micro-decisions rather than a few macro strategic ideas.
AI is rapidly changing how product managers conduct market research, design mocks, and tell product stories, making these tasks easier. However, the fundamental need to deeply understand what end-users want and to develop strong product judgment (gut feel) remains constant and becomes even more critical.
In competitive consumer markets, it's difficult for third or fourth players to keep up with top competitors unless they have a unique angle. This teaches the importance of speed, hustle, resilience, and never taking the status quo for granted.
Turn on the porch light at night to help couriers easily find your house and the correct door for food placement.
23 Actionable Insights
1. Practice Extreme Dogfooding
Set aside half a day or a day once or twice a month to personally use your product as an end-user, performing 10-12 sessions. This helps you experience the product’s real-world impact and user emotions, which is crucial for product development.
2. Document Dogfooding Learnings
After dogfooding, take screenshots and write detailed reports (e.g., 40 pages) of issues and suggested fixes, tagging relevant team members. This ensures learnings are not forgotten and provides concrete evidence for product improvements.
3. Follow Through on Dogfooding Feedback
Develop impatience for fixing identified product issues and actively follow up with teams to ensure they are prioritized and resolved quickly. This transforms dogfooding from a fun activity into an impactful process for product improvement.
4. Operationalize Dogfooding Culture
Leaders should model dogfooding behavior, organize company-wide driving/delivering weeks with competitions, and set OKRs (e.g., fix 300 issues per half) to ensure feedback is acted upon and rewarded. This embeds a culture of user empathy and continuous product improvement.
5. Embrace a “Ship, Ship, Ship” Mentality
Focus on shipping actual code in your product as quickly as possible, rather than just documents or designs. This minimizes the cycle time between identifying a problem and delivering a solution to end-users.
6. Minimize Decision-Making Cycle Time
Actively cut down the time spent between development activities (e.g., requirements, design, coding) by making quick decisions, especially for “two-way door” choices. This speeds up the shipping process and reduces unnecessary delays.
7. Prioritize Shipping Experience Early
For early career product people, seek jobs where you can ship multiple products (3-5 cycles) as fast as possible over 2-3 years. This helps develop innate judgment and “product sense” through numerous micro-decisions.
8. Focus on Micro-Decisions for Judgment
Understand that great product management comes from thousands of micro-decisions (e.g., button placement, copy) rather than a few macro-strategic ideas. Rapidly making these decisions builds a strong “gut” or product sense.
9. Deeply Understand End-User Needs
The constant in product management is the need to deeply understand what end-users want, which is harder and requires more effort than typically assumed. This involves going beyond data to grasp their emotions and context.
10. Acknowledge Non-Reciprocal User Relationship
Understand that users’ lives are complex and not centered around your product; they only think about it for brief moments. Your goal is to dazzle them in those brief moments, not to take over their lives.
11. Unblock Teams Proactively
Product leaders should be hands-on when needed, stepping in to write PRDs or run daily stand-ups to remove ambiguity and accelerate decision-making for stalled teams. This helps maintain momentum and ensures critical projects move forward.
12. Insist on Live Product Demos
For product announcements, insist on live demos to tell a compelling story about user value and create rigor among the team before launch. This also provides a moment of pride for the team seeing their work in action.
13. Foster Co-Creation with Engineers
Encourage engineers to co-create products from the start, participating in whiteboarding and document writing, rather than waiting for a fully specified PRD. This leverages their expertise and avoids blocking progress.
14. Balance Core & Growth Bets
Maintain a “concentric circles” framework for your product portfolio, obsessing over the flawless and efficient operation of your core product. This provides the license and resources to expand into new growth bets.
15. Challenge Data with Gut Feeling
Sometimes, pursue big bets or features even if initial data doesn’t fully support them, especially if you intrinsically understand a deep user problem. This requires trusting your developed product sense and judgment.
16. Cultivate Speed, Hustle, and Resilience
Always be paranoid about competitors and never take the status quo for granted, fostering a sense of urgency and hustle. Develop resilience to power through difficult times and continue building for end-users.
17. Use AI for Document Summarization
Leverage AI tools like ChatGPT or Gemini to summarize long user research documents (e.g., 50-100 pages). This helps product managers quickly get acquainted with diverse user sentiments and regional insights.
18. Utilize AI as a Research Assistant
Use AI’s deep research features as a thought partner to evaluate potential features, identify gaps in thinking, and generate new ideas for brainstorming sessions. This enhances productivity and strategic insight.
19. Dogfood Non-Consumer Products
For non-consumer products, get as close to the customer as possible by sitting next to them to observe product use or setting up test accounts to simulate their experience. This helps identify friction points even in complex B2B flows.
20. Be Respectful to Service Providers
When interacting with service providers (e.g., Uber drivers), ask permission before being on the phone and close car doors softly. This shows respect for their personal space and property, potentially leading to better ratings.
21. Aid Delivery Drivers at Night
If receiving a food delivery at night, turn on your porch light to help the driver easily find your house and doorstep. This simple act improves their experience and safety.
22. Study Iconic Tech Leaders’ Context
Read biographies of iconic tech leaders (e.g., Steve Jobs, Elon Musk) to understand the context that shaped their product decisions and world-changing innovations. Apply these principles to your own context, rather than just imitating their behavior.
23. Learn Resilience from Leadership Books
Read books like “The Hard Thing About Hard Things” to understand the challenges of leadership and how to persist through difficult times. This helps develop the resilience needed to keep building for end-users.
6 Key Quotes
You don't ship documents. You don't ship brainstorming meetings. What you ship is code in your product.
Sachin Kansal
What makes a great product manager is not five amazing strategic ideas. It's the thousand micro decisions that you made.
Sachin Kansal
Until I get behind the wheel or I'm a passenger or I'm using Uber Eats, what I miss is the visceral reaction that you get when something happens.
Sachin Kansal
I'm a big believer in then bringing that emotion to my work.
Sachin Kansal
No one actually really cares about the feature that you want to talk about. People are way too busy. You're obsessed about your products, your products are your life, but for your end users, your products are maybe 10 minutes of their life once a week.
Sachin Kansal
Focus on the inputs. You have input metrics in life and in work and you have output metrics. You can't really control the output metrics. All you can control is inputs.
Sachin Kansal
6 Protocols
Sachin Kansal's Personal Dogfooding Process
Sachin Kansal- Take 5-10 Uber rides and place 3 Uber Eats orders weekly.
- Once or twice a month, set aside half a day or a full day to drive and deliver as an Uber driver/courier (aiming for 10-12 trips).
- During or immediately after, take many screenshots and document everything learned that is not ideal, adding details and screenshots to a dedicated document (e.g., driver document, rider app document, Uber Eats document).
- Tag relevant team members who work on specific features in the document.
- Share the document with the team for review and prioritization.
- Follow through on reported issues to ensure they are fixed as quickly as possible.
Tips for Riders to Avoid Annoying an Uber Driver
Sachin Kansal- If you need to be on the phone, ask the driver for permission first.
- When exiting the car, close the door softly; do not slam it.
Developing Product Sense / Judgment
Sachin Kansal- Seek jobs where you can ship multiple products as fast as possible, aiming for 3-5 product cycles in 2-3 years.
- Focus on making thousands of 'micro decisions' (e.g., button placement, copy, launch markets) rather than just a few macro strategic ideas.
- Continuously increase your cycle time to develop an innate sense of judgment and 'gut' feel for correct product decisions.
Uber's Portfolio Strategy for Growth and Efficiency
Sachin Kansal- Obsess over making the core product (e.g., 33 million daily trips) as flawless and efficient as possible, continuously seeking to remove even small inefficiencies.
- Once the core is solid and efficient, use the resources and trust gained to expand into new growth bets (e.g., Uber Eats, Uber Groceries, Uber Reserve, Uber for Teens, Taxis).
Using AI as a Product Manager
Sachin Kansal- Use AI tools (e.g., ChatGPT, Gemini) for summarization of long documents, such as 50-100 page user research reports, to quickly grasp key insights.
- Leverage 'deep research' features in AI tools by feeding them hard prompts about potential features or strategies to get initial ideas and identify gaps in thinking, treating the AI as an excellent research assistant and a starting point for brainstorming.
Tip for Uber Eats Customers to Avoid Annoying a Courier
Sachin Kansal- If receiving a delivery at night, turn on your porch light.