Companies in the software-intensive systems industry collect vast amounts of data from their products on a continuous basis. While many companies have more data than what they will ever use, the volumes keep growing and the frequency at which data is collected is only increasing. Still, there is little guidance for how to effectively manage and make use of this data and companies struggle with how to turn raw data into customer and business value. In this paper, we study data practices in software-intensive systems companies and we identify beneficial patterns that lead to successful use of data. Also, we identify detrimental anti-patterns that companies should avoid when working with data.