How to Lead in Data Science

Jike Chong, Yue Cathy Chang

• Publisher: Manning
• Year: 2021
• ISBN:
  9781638356806
• Topic: Computer Science
• Subtopic: Data Processing
• Index: Computer Science

A field guide for the unique challenges of data science leadership, filled with transformative insights, personal experiences, and industry examples. In How To Lead in Data Science you will learn: Best practices for leading projects while balancing complex trade-offs
Specifying, prioritizing, and planning projects from vague requirements
Navigating structural challenges in your organization
Working through project failures with positivity and tenacity
Growing your team with coaching, mentoring, and advising
Crafting technology roadmaps and championing successful projects
Driving diversity, inclusion, and belonging within teams
Architecting a long-term business strategy and data roadmap as an executive
Delivering a data-driven culture and structuring productive data science organizations How to Lead in Data Science is full of techniques for leading data science at every seniority level—from heading up a single project to overseeing a whole company's data strategy. Authors Jike Chong and Yue Cathy Chang share hard-won advice that they've developed building data teams for LinkedIn, Acorns, Yiren Digital, large asset-management firms, Fortune 50 companies, and more. You'll find advice on plotting your long-term career advancement, as well as quick wins you can put into practice right away. Carefully crafted assessments and interview scenarios encourage introspection, reveal personal blind spots, and highlight development areas. About the technology
Lead your data science teams and projects to success! To make a consistent, meaningful impact as a data science leader, you must articulate technology roadmaps, plan effective project strategies, support diversity, and create a positive environment for professional growth. This book delivers the wisdom and practical skills you need to thrive as a data science leader at all levels, from team member to the C-suite. About the book
How to Lead in Data Science shares unique leadership techniques from high-performance data teams. It's filled with best practices for balancing project trade-offs and producing exceptional results, even when beginning with vague requirements or unclear expectations. You'll find a clearly presented modern leadership framework based on current case studies, with insights reaching all the way to Aristotle and Confucius. As you read, you'll build practical skills to grow and improve your team, your company's data culture, and yourself. What's inside How to coach and mentor team members
Navigate an organization's structural challenges
Secure commitments from other teams and partners
Stay current with the technology landscape
Advance your career About the reader
For data science practitioners at all levels. About the author
Dr. Jike Chong and Yue Cathy Chang build, lead, and grow high-performing data teams across industries in public and private companies, such as Acorns, LinkedIn, large asset-management firms, and Fortune 50 companies. Table of Contents
1 What makes a successful data scientist?
PART 1 THE TECH LEAD: CULTIVATING LEADERSHIP
2 Capabilities for leading projects
3 Virtues for leading projects
PART 2 THE MANAGER: NURTURING A TEAM
4 Capabilities for leading people
5 Virtues for leading people
PART 3 THE DIRECTOR: GOVERNING A FUNCTION
6 Capabilities for leading a function
7 Virtues for leading a function
PART 4 THE EXECUTIVE: INSPIRING AN INDUSTRY
8 Capabilities for leading a company
9 Virtues for leading a company
PART 5 THE LOOP AND THE FUTURE
10 Landscape, organization, opportunity, and practice
11 Leading in data science and a future outlook

1 What makes a successful data scientist?

This chapter covers

• Learning what is expected of data scientists
• Examining the challenges of a data scientist’s career progression

Data science (DS) is driving a quantitative understanding of the world around us. When the technologies to aggregate large quantities of data are paired with inexpensive computing resources, data scientists can discover patterns through analysis and modeling at scales that were not possible just decades earlier. This quantitative understanding of the world through data is being used to predict the future, drive consumer behavior, and make critical business decisions. The scientific process used to improve our understanding of the world allows us to craft solutions based on testable and repeatable results.

Leadership is the ability to amplify your capabilities by influencing, nurturing, directing, and inspiring people around you to produce more significant impact than what can be achieved as an individual. There are opportunities to lead as a technical individual contributor and as a people manager.

001	Leadership is the ability to amplify your capabilities by influencing, nurturing, directing, and inspiring people around you to produce more significant impact than an individual can achieve. There are opportunities to lead as a technical individual contributor and as a people manager.

Building a DS function in a company to produce industry-leading, data-driven innovation is currently within reach for many nimble organizations. However, 95% of the companies with DS teams have teams of fewer than 10 members [1], [2]. Leadership talent who can lead projects, nurture teams, direct functions, and inspire industries are scarce and in high demand. This book lays out many paths for every data scientist to navigate for the next stages of their career. It also shares the expectations of the roles in great DS teams and organizations.

This chapter introduces the historical and current expectations for data scientists, discusses the hard capabilities and soft psychosocial virtues crucial for data scientists, and shares interview and promotion challenges in case studies. It aims to help you contextualize real opportunities and challenges in the workplace. Let’s begin!

1.1 Data scientist expectations

In 2010, Drew Conway introduced the well-known data science Venn diagram [3] (figure 1.1), which clarified three pillars of skills required for success in the nascent field of DS: math and statistics knowledge, hacking skills, and substantive expertise. The Venn diagram pushed the DS field forward by crystallizing a unique set of skills in an uncommon group of talent who can unleash extraordinary opportunities for nations, businesses, and organizations.

Figure 1.1 Data science Venn diagram from 2010 by Drew Conway

Dr. Conway later founded multiple technology companies, including Datakind, Sum, and Alluvium. Countless blogs and books have since referenced the Venn diagram he introduced. By 2021, over 200,000 DS practitioners worldwide earned the title of data scientist. How has the field evolved?

1.1.1 The Venn diagram a decade later

While many of the 2010 original terms and ideas are still valid, there have been updates, debates, and even battles on the topic of the DS Venn diagram; a simple image search of these words would yield scores of variations. The role of a data scientist has significantly expanded since its inception. In 2021, the math and statistical knowledge pillar has broadened to a more general technology capability. The technology capability includes tools and frameworks for you to lead projects more effectively. They are used to frame the problem, understand data characteristics, innovate in feature engineering, drive clarity in modeling strategies, and set expectations for success.

The hacking skills pillar has extended to execution capabilities and now includes the practices for you to specify projects from vague requirements and prioritize and plan projects while balancing difficult trade-offs, such as speed versus quality, safety versus accountability, and documentation versus progress.

Substantive expertise has expanded to include having expert knowledge to clarify project alignment to the organizational vision and mission, account for data source nuances, and navigate structural challenges in the organization to launch projects successfully. While these are the pillars that make a successful data scientist, we find that it is difficult, if not impossible, to locate individuals who are strong in all three dimensions.

For example, a data scientist entering the field of DS with an academic background often has strong capabilities only in the technology dimension. A data scientist with years of experience in the industry can usually pick up execution best practices on the job, including the ability to deploy scalable and maintainable DS solutions. A seasoned DS practitioner with a long tenure in a domain with substantive expert knowledge is rare to find and could be highly valuable to the right employer.

Are these three capabilities, technology, execution, and expert knowledge, sufficient for succeeding in the field of DS today? Let’s find out!

1.1.2 What is missing?

As with any practitioners in the field, we had our share of blind spots in building teams. While we diligently assessed candidate capabilities in technology, execution, and expert knowledge, our hiring mishaps showed up when candidates were vetoed in final-round executive interviews or, worse, were hired and then had to be managed out of the team. Many of these failures were summarized as “not a cultural fit.” But what does that mean?

What is the culture for the DS field that we’re looking to “fit”? How is that distinct from an organization’s culture or an industry’s culture? To analyze these failures, this book expands the interviews, reviews, and promotion criteria of a data scientist to consider not just the capabilities but also the virtues of a data scientist in pursuing a DS career.

According to the Greek philosopher Aristotle, virtues come from years of practicing being good to benefit oneself as well as society. They are the individual’s habitual actions etched into one’s character.

Virtues in DS are nurtured. We highlight three dimensions of virtues to nurture into hab...