How to Talk to Data Scientists

Rather than teach a diluted version of data science, this book describes how to start projects and how to detect and fix problems - the moments when leadership is critical. Insights are provided through real world examples, including a ‘Playbook’ featuring common projects.

---

Every major company has or will soon have a data science program. Most fail, expensively, imperiling their executive sponsors. Unfortunately, executives have been misled to focus on the latest buzzwords. Although buzzwords change—big data, data science, machine learning, deep learning, and artificial intelligence –the distraction from fundamentals manifests as a predictable trajectory from exuberant program launch, to stagnation, to awkward decommissioning.

After architecting data science programs at over a dozen companies, across sectors and scales, Dr. Elser has formulated a reliable framework for successful data science programs. Surprisingly, software and algorithms are secondary. Rather, the key is understanding how the available data aligns to the problem to be solved. The business executive understands the problem sufficiently to enforce this alignment, while data scientists act on it. But executives tend to underestimate their role and thereby fail to construct the necessary connective tissue with their data scientists.

This book provides business executives with a concrete exercise, populating a "Master Table," accessible to nontechnical managers and data scientists, which serves as the connective tissue between them. Rather than teach a diluted version of data science, this book describes how to start projects and how to detect and fix problems—the moments when leadership is critical. Insights are provided through real world examples, including a Playbook featuring common projects. The intended audience is executives (C-suite through VP). However, ambitious mid-level managers and data scientists will also benefit.