This document discusses key metrics and analysis that are important for data scientists. It emphasizes that daily active users, retention rates, time spent, and revenue are classic good metrics to analyze. Good analysis should describe user needs and inform decisions. Complexity should be avoided as an anti-goal. Measuring impact and having observable consequences are important. Assumptions, small data, talking to users, and making predictions can help overcome objections around data and analysis. Confidence intervals, testing intuitions with data, and the value of information are also discussed.