Despite heavy investments to acquire talented data scientists and take advantage of the analytics boom, many companies have been disappointed in the results. The problem is that those scientists are trained to ask smart questions, wrangle the relevant data, and uncover insights--but not to communicate what those insights mean for the business. To be successful, the author writes, a data science team needs six talents: project management, data wrangling, data analysis, subject expertise, design, and storytelling. He outlines four steps for achieving that success: (1) Define talents, not team members. (2) Hire to create a portfolio of necessary talents. (3) Expose team members to talents they don't have. (4)  Structure projects around talents.