You can't really invest in a data team too early, as it's something that grows with the company, but as soon as possible is ideal for early-stage startups to establish product usage trends, create business insights, and identify areas to improve their product. Hire a generalist first who can do a little bit of everything, rather than an expert from a large brand name, to create data insights, develop business hypotheses, and build a scalable data architecture from scratch. A small data team can answer big questions by thinking in terms of scalability and efficiency, and should focus on preventing issues in the future. The tools used include a centralized data warehouse, ELT tools like Segment and Fivetran, dbt for cleaning raw data, Census to send data back to other platforms, Metabase for self-serve data dashboards, Propel for in-app analytics, and Courier for routing notifications. The data team is evolving again with the goal of enabling data scientists to build subject matter expertise while maintaining centralized data standards.