Lenny's PodcastJessica Lachs: Why centralized data beats embedded analysts
Through pods aligned to product and ops, DoorDash centralizes data; analysts share goals with their teams, and case interviews hire for curiosity over skill.
At a glance
WHAT IT’S REALLY ABOUT
How DoorDash Built a Centralized, Impact-Driven World-Class Data Organization
- Jessica Lachs, VP of Analytics and Data Science at DoorDash, explains how she built one of tech’s most respected data orgs by positioning analytics as a strategic, impact-driving partner rather than a service function.
- She strongly favors a centralized “center of excellence” data model with embedded pods and shared goals, enabling higher talent bars, consistent metrics, cross-functional mobility, and a strong team culture.
- Lachs dives into how to hire great data people, design effective and simple metrics that truly drive behavior, and create space for proactive, exploratory work that uncovers high-leverage opportunities.
- She also shares stories from DoorDash’s early scrappy days, the company’s culture of extreme ownership and customer obsession, and how her own non-traditional background shaped her approach to building a global analytics organization.
IDEAS WORTH REMEMBERING
5 ideasCentralize the data org, but embed pods and share goals with the business.
DoorDash runs a centralized analytics org with pods aligned to product, ops, and marketing; analysts report centrally to maintain standards and community, but share the same goals and roadmaps as their partner teams to ensure tight alignment and impact.
Treat analytics as a seat-at-the-table partner, not a service bureau.
Lachs rejects a Jira-ticket, dashboard-only mentality; her team is expected to proactively find opportunities, bring opinions, and answer “so what/what now?”—earning their seat by driving material business outcomes.
Hire for deep curiosity and structured problem-solving, not just raw technical skills.
Technical competence is table stakes; the differentiator is people who pull on odd threads unprompted, handle ambiguity, accept being wrong, and can still make a call without perfect information—tested via real-world case interviews with intentional imperfections.
Use simple, understandable metrics and short-term proxies for long-term outcomes.
Instead of goaling teams on slow-moving metrics like retention or opaque composite scores, DoorDash identifies measurable input metrics that predict long-term value and keeps them simple enough that everyone can reason about and act on them.
Explicitly quantify trade-offs via a common business currency.
By expressing different levers (price cuts, faster delivery, new merchants, better login flows) in a shared unit like orders or GOV, DoorDash can compare investments across functions and choose the highest-ROI actions more quickly and objectively.
WORDS WORTH SAVING
5 quotesFor me, analytics is a business impact driving function, and not purely a service function.
— Jessica Lachs
I strongly disagree [with embedding]; I believe a central model, a center of excellence, is superior.
— Jessica Lachs
Ultimately, you want to find a short-term metric you can measure that drives a long-term output.
— Jessica Lachs
Retention is a terrible thing to goal on. It's almost impossible to drive in a meaningful way in the short term.
— Jessica Lachs
Yes, you are a data scientist, but your goal is to figure out what's happening, and if that means that you're gonna pick up the phone and call customers, then that is what you're going to do.
— Jessica Lachs
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome