Product Analytics

Rodrigues works through the analytical toolkit for consumer products, from defining metrics to building models that explain user behaviour. The book starts with behavioural theory and the argument that a metric needs a causal story behind it, then moves into descriptive analysis, cohort and retention modelling, and the statistics that experimentation depends on. Later sections cover causal inference outside the experiment, including matching methods and instrumental variables, along with simulation and forecasting. Code examples are given in both R and Python, with the reasoning explained alongside. Suited to an analyst or a technically inclined product manager who wants to answer questions an A/B test cannot.