The Kinds of Data Scientist

Yael Garten

In 2012, HBR dubbed data scientist “the sexiest job of the 21st century”. It is also, arguably, the vaguest. To hire the right people for the right roles, it’s important to distinguish between different types of data scientist. There are plenty of different distinctions that one can draw, of course, and any attempt to group data scientists into different buckets is by necessity an oversimplification. Nonetheless, I find it helpful to distinguish between the deliverables they create. One type of data scientist creates output for humans to consume, in the form of product and strategy recommendations. They are decision scientists. The other creates output for machines to consume like models, training data, and algorithms. They are modeling scientists.

1. Data science for humans: the consumers of the output are decision makers like executives, product managers, designers, or clinicians. They want to draw conclusions from data in order to make decisions such as which content to license, which sales lead to follow, which medicine is less likely to cause an allergic reaction, which webpage design will lead to more engagement or more purchases, which marketing email will yield higher revenue, or which specific part of a product user experience is suboptimal and needs attention. These data scientists design, define, and implement metrics, run and interpret experiments, create dashboards, draw causal inferences, and generate recommendations from modeling and measurement.
2. Data science for machines: here the consumers of the output are computers which consume data in the form of training data, models, and algorithms. Examples of the work products of these data scientists are: recommendation systems which recommend what shirt a customer might like or what medicine a physician should consider prescribing based on a designed optimization function, such as optimizing for customer clicks or for minimizing readmission rates to the hospital. Depending on the engineering background of these data scientists, these work products are either deployed directly to the production system, or if they are prototypes they are handed off to software engineers to help implement, optimize and scale them.

The elusive full stack data scientists do exist, though they are hard to find. In most organizations, it makes sense for data scientists to specialize into one type or another. But data scientist are curious creatures who thrive from being able to creatively dabble; there are benefits to giving them flexibility to work on projects that touch both “types” – both for them and for the organization.   (The sidebar offers more detail on how the two types of data scientists differ not only in their skills and the work they do, but in whom they partner with and their measures of success.)

To read full article