Lecture Notes: Building and Scaling Data Teams with Jessica Lax

Introduction

• Presentation by Lenny with guest Jessica Lax
• Jessica Lax: VP of Analytics and Data Science at DoorDash
  • 10+ years at DoorDash
  • Previously founded GS Simple, social gifting startup
  • Career began in investment banking at Lehman Brothers
• Discussion Topics:
  • Building and scaling a data organization
  • Structuring data teams
  • Importance of metrics
  • Examples of data-driven decisions at DoorDash
  • Early days of DoorDash

Key Insights on Building Data Teams

Philosophy of Analytics

• Analytics should drive business impact, not just serve as a support function
  • Not just answering 'why' but 'what do we do now?'
• Ownership and involvement in various tasks to understand data (e.g., calling customers)

Structuring Data Teams

• Centralized Org Model: Preferred over embedding analytics in business units
  • Central team divided into pods aligned with product, engineering, marketing, etc.
  • Shared goals with partner teams to ensure alignment of incentives
• Dual Goals: Retain high talent bar and ensure growth opportunities
• Benefits of Centralization:
  • Consistent and high talent bar
  • Better growth opportunities for employees
  • Consistency in methodologies and metrics
  • Identification of shared issues across different teams
  • Strong team culture and brand

Practical Examples and Stories

Data Team’s Impact

• Example of discovering fraudulent behavior in referral programs through deep dives during hackathons
  • Importance of diving deep into data and experimenting to find true insights

Advice on Balancing Tasks

• Proactive insights vs. answering immediate questions
  • Prioritization and clear communication with business partners make this balance possible
  • Making trade-offs explicit can help align priorities

Hiring Best Practices

• Technical Skills: A non-negotiable baseline for hiring
• Curiosity: Self-motivated to dive into unexpected issues beyond given tasks
• Examples of assessing curiosity:
  • Case studies with intentional anomalies
  • Asking candidates about past experiences

Metrics and Incentives

Importance of Right Metrics

• Short-term metrics that drive long-term impact
• Simple metrics over complex composites
• Example: Merchant Health Score simplified into actionable sub-metrics

Cross-Company Metric Comparisons

• Use of common currency to compare diverse areas of the business
• Translate improvements in specific areas into core business metrics (e.g., GOV, volume)

Learning from Fail States

• Set goals around extreme cases to prevent them (e.g., “never delivered” orders)

Overcoming Background Limitations

• Jessica’s non-traditional shift to data science from a business background
  • Embracing opportunities and tackling problems head-on
  • Importance of diverse backgrounds within teams

Global Data Team Management

• More similarities than differences across locations
• Complexity due to different currencies, regulations, and languages

AI in Data Work

• Example: “Ask Data AI” - AI tool for autonomous data queries within DoorDash

Final Thoughts

• Non-traditional backgrounds can thrive in data science
• Importance of a strong, diverse, and motivated team culture
• Encourage truth-seeking in data and decisions

Lightning Round (Fun and Personal Insights)

• Favorite books, movies, TV shows, and life motto
• Influence from professional mentors and parents
• Key moment realizing DoorDash's eventual success: gaining #1 market share, public recognition