Traditional data teams are often criticized for being mere "curators" who pull out metrics without using their full potential. Former CTO Erik Bernhardsson of Better and Spotify's music recommendations feature thinks that data teams should be used to investigate and explore which metrics would make the biggest impact on a company's goals, rather than just following instructions from executives. To build a truly data-driven company, leaders need to identify quality metrics that can measurably impact their business, and give data analysts more freedom to model questions before trying to answer them. The right reporting structure is key, with some companies adopting an embedded model where data teams report to a centralized head of data but perform day-to-day tasks in assigned business units. Successful data teams are built on creative people who care about the business outcome, rather than just tool-oriented engineers. By loosening the reins and giving data analysts more autonomy, businesses can tap into their full potential and drive real business success.