Building a world-class data org | Jessica Lachs (VP of Analytics and Data Science at DoorDash)

Lachs describes pods aligned with partner teams while reporting through a central analytics organization. Shared goals help analysts work as partners rather than a detached request queue. The arrangement is her account of DoorDash's structure, not proof that it fits every organization.

She expects analysts to bring opportunities, trade-offs and a deeper understanding of business problems to the table. A seat in the conversation must be earned through that contribution. Completing requested data pulls alone is not the whole role she describes.

A central organization can support a consistent hiring bar and shared metrics. Lachs points out that separate definitions of the same business measure make collaboration harder. Reusing methods also gives several teams a way to improve one approach instead of rebuilding it separately.

When a new request arrives, Lachs recommends comparing it with the work already planned. Ask which item should move, rather than quietly trying to do everything. Leadership should define an operating model so junior people do not carry every difficult prioritization conversation alone.

Her referral example describes an aggregate that hid groups with very different behavior. A deeper analysis changed the understanding of the channel. The lesson is to inspect distributions and cohorts before deciding an average explains the opportunity.

Lachs wants a measurable input that can be changed quickly and tested against a longer-term outcome. Choosing a plausible input is only the start: experimentation should check whether it actually moves the intended result. She also favors measures people can understand over opaque composite scores.

She values technical skills alongside different functional backgrounds and company stages. People who have worked at startups and people who have seen scale can teach one another. This is a team-design choice, not an argument that technical competence can be skipped.