What to Expect When You're Expecting: Data Science Edition
Today, data science is among the most desired capabilities for hiring firms and job candidates alike. Firms yearn to leap ahead of the competition by enlisting teams of scientists to improve their utilization of data and predict business outcomes. Scientists seek to fill these positions to sharpen their mathematics, programming, and research toolkits. This scenario looks perfect on paper, however in practice, with data science being such a new initiative for many companies, hiring firms and applying scientists often have very different expectations of what it means to be a data scientist.
As a brief illustration, let’s consider the duties of an average data scientist working at a midsize company. In future articles, we’ll explore what a data scientist is, does, and does not do in greater detail, but for now, let’s start with a high-level overview of what a data scientist should be working on and the type of output organizations can expect.
This table can be thought of as the happy medium for each party (firm and scientist). If a scientist is doing some mixture of these practices on a day-to-day basis, they should feel effective and challenged in their career. If a firm has a need for some mixture of these practices on an ongoing basis, they should see tangible outcomes and feel validated in their decision to hire full-time scientific talent.
To illustrate further, let’s examine some typical data science workflows in the diagram below.
As you can imagine, not every firm will require all four data science practices in their day-to-day operations. In fact, some may not require any data science practices in their day-to-day at all. Companies with large legacy data systems may need simpler or more advanced ways to efficiently and optimally utilize or analyze their data, while others may need organized, usable databases but have no use for scientific experimentation and modeling.
These complex challenges are fueling the misalignment in job expectations between firms and candidates, and the questions raised here are only a handful of the considerations that must be addressed in determining how and whether to bring data scientists and data science capabilities into an organization. This series of articles seeks to peel back the curtain of the sometimes murky process of onboarding scientific talent. Our purpose is not to dissuade readers from hiring data scientists nor to belittle firms for past hiring foibles, but to highlight and help resolve some common challenges that employers and data scientists struggle to address. We’ll explore a number of scenarios playing out in the current business environment and examine the unique profiles of different data science-related roles (e.g. data scientist, data engineer, statistician, and other related positions) to better equip scientists and firms to evaluate their needs and expectations and build more symbiotic relationships in the future. By the end of this series, we hope to have given readers a clearer understanding of how to answer critical questions such as:
When would a company benefit from hiring a full-fledged data scientist as opposed to a database analyst, statistician, etc?
What are the differences between these roles?
When should a firm hire a full-time resource, and when should they opt for a consultant or contractor?
Look out for our next post where we’ll breakdown the ecosystem of data science roles and their related counterparts. In the coming weeks, we'll also hear firsthand from Valkyrie scientists about their experiences working in a variety of different types of companies!