Loading...
July 14, 2024 • 1hr 19min
Building a world-class data org | Jessica Lachs (VP of Analytics and Data Science at DoorDash)
Lenny's Podcast: Product | Growth | Career
Jessica Lachs is the VP of Analytics and Data Science at DoorDash, where she has built one of the largest and most impactful data teams in tech over the past 10 years. In this episode, she shares insights on structuring and scaling high-impact analytics organizations, defining effective metrics, fostering a culture of extreme ownership, leveraging AI, and building diverse data teams.
Jessica comes from a non-traditional background, having started her career in investment banking and founding a startup before joining DoorDash as its first GM. She built the analytics function out of necessity, teaching herself SQL and Python to help set goals and measure performance across markets in DoorDash's early days.
Building a world-class data org | Jessica Lachs (VP of Analytics and Data Science at DoorDash)
0:00
-0:00
Key Takeaways
- Centralized analytics model is superior to embedded teams - provides consistency in talent, methodologies, metrics, and culture while still allowing analysts to be embedded with partner teams
- Data teams should drive business impact, not just answer questions - focus on finding opportunities and having a point of view on decisions
- Hire for curiosity and problem-solving skills over just technical abilities - look for people who will dig deeper unprompted
- Keep metrics simple and actionable - avoid complex composite metrics in favor of clear, understandable metrics teams can move
- Translate all metrics to a common "currency" like order volume to enable trade-off decisions across teams
- Set goals around edge cases and fail states, not just averages - e.g. "never delivered" orders
- Encourage extreme ownership culture where analysts go beyond their typical role to solve problems
- AI tools like chatbots can empower non-technical users to do more data analysis themselves
- Build diverse teams with complementary skills and backgrounds - mix of startup and big company experience, different functional expertise
Introduction
Jessica Lachs is the VP of Analytics and Data Science at DoorDash, where she has built one of the largest and most impactful data teams in tech over the past 10 years. In this episode, she shares insights on structuring and scaling high-impact analytics organizations, defining effective metrics, fostering a culture of extreme ownership, leveraging AI, and building diverse data teams.
Jessica comes from a non-traditional background, having started her career in investment banking and founding a startup before joining DoorDash as its first GM. She built the analytics function out of necessity, teaching herself SQL and Python to help set goals and measure performance across markets in DoorDash's early days.
Topics Discussed
Centralized vs. Embedded Analytics Teams (4:59)
Jessica strongly advocates for a centralized analytics model over embedding analysts in different business units:
- Centralized model benefits:
- Consistent high talent bar across the org
- More growth opportunities for analysts
- Consistency in methodologies and metrics
- Ability to see patterns across teams
- Stronger team culture and brand
- How it works at DoorDash: Analysts are divided into pods that map to product/engineering/ops structure, but report centrally
- Key point: "We are very much to the earlier point, we have a seat at the table. We are business partners, we are thought partners with our product counterparts, with our engineering counterparts, with our ops counterparts."
Balancing Proactive and Reactive Work (15:10)
Jessica emphasizes the importance of carving out time for exploratory work and deep dives, not just answering inbound requests:
- Set goals for the team around finding insights through self-directed work
- Hold hackathons to dedicate time to exploring interesting problems
- Example impact: A hackathon deep dive into referrals uncovered fraud issues and led to major improvements in the program
Advice on Pushing Back Effectively (20:45)
For data teams struggling to balance reactive requests with proactive work:
- Establish a culture where leadership sets expectations on prioritization
- Align goals with business partners to easily say no to low-priority requests
- Make trade-offs explicit: "Is this data pull that you want me to do more important than these other three things that I was going to be working on? Yes or no?"
- Regularly re-evaluate priorities with business partners
Hiring for Curiosity and Problem Solving (24:20)
Jessica looks beyond just technical skills when hiring:
- Key trait: Curiosity - people who will dig deeper unprompted
- Interview approach: Present cases with intentional issues to see if candidates notice and how they react
- Look for: How candidates handle ambiguity, being wrong, and making decisions with imperfect information
Coming from a Non-Traditional Background (28:57)
Jessica shares her experience building the analytics function without formal training:
- Learned SQL and Python out of necessity to help set goals and measure performance
- Focused on solving immediate problems rather than building a grand vision
- Key insight: "If you think about things from first principles about what you need right now in front of you to unblock yourself or solve a problem, and you just focus on that instead of thinking about, like, you know, a global that you're trying to build. And I think that that helps."
The Early Days and Culture at DoorDash (34:40)
Jessica shares stories that highlight DoorDash's early culture:
- Extreme ownership: Sales team helping with consumer acquisition despite not being their direct goal
- Customer-first mentality: Entire company jumping on customer support during outages
- WeDash program: All employees, including executives, regularly do food deliveries or customer support
Encouraging Cross-Functional Roles (40:39)
Jessica discusses how DoorDash encourages analysts to go beyond their typical role:
- Example: Data scientists making customer calls to understand why a feature didn't work as expected
- Encourages team members to do product management or engineering work when needed
- Results in analysts moving to other functions like product or operations
Defining Effective Metrics (44:39)
Jessica shares insights on choosing the right metrics:
- Find short-term metrics that drive long-term outcomes - e.g. inputs that drive retention rather than retention itself
- Keep metrics simple and understandable - avoid complex composite scores
- Quantify all levers in common terms (e.g. order volume) to enable trade-off decisions
- Focus on edge cases and fail states, not just averages - e.g. "never delivered" orders
Simplifying Metrics for Better Outcomes (46:30)
Jessica provides an example of simplifying metrics:
- Problem: Complex "merchant health" composite score that was hard to understand and act on
- Solution: Broke it down into simpler, actionable metrics like "% of new merchants getting an order in 7 days" and specific input metrics
- Key point: "Maybe we missed number four, five, and six on the list of things, but you got one through three, and that's 95% of it anyway."
Focusing on Edge Cases and Fail States (55:28)
Jessica emphasizes the importance of looking beyond averages:
- Example: "Never delivered" orders - rare but terrible experiences that drive churn and are costly
- Set concrete goals around eliminating these edge cases
- Key insight: "Just because something doesn't happen frequently doesn't mean that it's, that it's not important."
Managing a Global Data Organization (1:00:12)
Jessica shares her experience managing data teams across different countries:
- More similarities than differences in data scientists and consumer behavior across countries
- Added complexity from different currencies, languages, and regulations
- Benefit of having seen similar problems before in other markets
Leveraging AI for Productivity (1:02:31)
Jessica discusses how DoorDash is using AI to empower non-technical users:
- Ask Data AI: Chatbot that helps users write and edit SQL queries
- Goal is to reduce bandwidth needs for the analytics team
Building Diverse and Skilled Data Teams (1:05:25)
Jessica shares her approach to building diverse teams:
- Hire people from different functional backgrounds (e.g. operations, marketing, finance)
- Mix of startup and big company experience
- Encourage teaching and learning across team members with different expertise
- Key point: "We just have a group of people with different backgrounds who can all teach each other how to be better. And we're not all carbon copies, you know, of each other."
Conclusion
Jessica Lachs provides valuable insights into building and scaling high-impact analytics organizations. Her approach emphasizes centralized teams, a focus on business impact, hiring for curiosity, simplifying metrics, and building diverse teams. Key themes include fostering a culture of extreme ownership, balancing proactive and reactive work, and leveraging AI to empower non-technical users. Jessica's non-traditional background and problem-solving approach offer inspiration for aspiring data leaders without formal training.
