You are the PM at DoorDash, and you are about to release a new feature. A/B testing of that feature shows that the number of orders from users goes up, but the number of restaurants that were ordered from goes down (say from 100 to 90). Will you launch this feature or not?

Introduction

The trade-off we're facing at DoorDash involves a new feature that increases the number of orders from users but decreases the number of restaurants ordered from. This scenario presents a complex decision-making challenge that requires careful analysis of both short-term gains and long-term ecosystem health. I'll approach this by examining the context, metrics, and potential impacts before recommending a course of action.

Step 1

Clarifying Questions (3 minutes)

• What specific user behaviors or interface changes are driving the increase in orders?

• Why it matters: Understanding the mechanism behind the increase helps assess sustainability and potential for optimization.
• Hypothetical answer: The feature recommends personalized meal bundles, increasing order size and frequency.
• Impact: If true, we might focus on refining bundle recommendations to include more diverse restaurants.

• Are there particular types of restaurants experiencing a more significant drop in orders?

• Why it matters: Identifies potential negative impacts on specific segments of our restaurant partners.
• Hypothetical answer: Smaller, independent restaurants are seeing a larger decrease in orders.
• Impact: We may need to develop targeted solutions to support these vulnerable partners.

• How does this change affect our key revenue metrics, such as take rate and average order value?

• Why it matters: Ensures we're considering the financial implications for DoorDash.
• Hypothetical answer: Average order value has increased by 15%, offsetting the decrease in restaurant diversity.
• Impact: We'll need to balance short-term revenue gains against long-term platform health.

• What's the user feedback on this new feature?

• Why it matters: Helps gauge user satisfaction and potential long-term adoption.
• Hypothetical answer: Users report higher satisfaction due to convenience, but some miss discovering new restaurants.
• Impact: We might need to incorporate discovery features to maintain restaurant diversity.

• How urgent is this decision, and what are the implications of delaying the launch?

• Why it matters: Determines the timeline for decision-making and potential for further testing.
• Hypothetical answer: We have a 2-month window before a major marketing push, but competitors are working on similar features.
• Impact: We have some time for refinement but need to move decisively to maintain our competitive edge.