In a candid retrospective, Eugene Yan describes how his first 100 days as a data science lead felt nothing like his previous work as an individual contributor (IC). The core change, he says, is that your own technical output no longer defines your impact. Instead, your value comes from making the team more effective—through code reviews, design guidance, and removing obstacles that slow others down.
A second major shift is communication. As a lead, you spend far more time translating between business stakeholders and the team, clarifying vague problems, and negotiating priorities. This means less time in notebooks and more time in meetings, but those meetings become the lever for the team's direction.
Finally, the post stresses that measurement changes. Personal metrics like lines of code or model accuracy give way to team delivery and business adoption. Yan argues that letting go of IC-era habits—like hoarding hard problems—is one of the hardest adjustments, but also the most necessary for the role to work.