Data Science Overtaking The Data Scientist

The data scientist is dead. Long live data science!

Well, not dead, but certainly dying. Up until late 2012, the Google search popularity for "data scientist" tracked that for "data science" but thereafter has sagged.

This trend is even confirmed, though to a lesser degree, in job postings:

Why is this? I can think of three possible reasons:

1. Purple Unicorns

The term "data scientist" can mean different things to different people. In one extreme, it can mean expert statistician, machine learning engineer, Big Data Engineer, expert at data software (R, SAS, Jupyter, Excel, SQL, Linux command line), and domain expert all in one. Many are shy of even appearing to claim all of that, and so opt for less lofty titles like "Data Analytics". Still, Data Science itself must be accomplished, and so it does, regardless of the individuals' titles.

2. Data Science is a Team Sport

As is often said at the Data Science Association, data science is a team sport. Whereas in 2012 when a company wanted to kick off a data science initative and they thought, "oh, let me advertise for a data scientist," now companies know they need a team that includes Big Data/database engineers and data curators.

The concept of a data science team has been gaining popularity since 2013.

3. The Long-Sought Dream of Self-Service Data Science

"Self-service data science" has been touted by various software vendors since at least 2013, perhaps in response to industry frustration in finding purple unicorns. Today, Google reports over 16,000 hits for "self-service data science". So it seems there are some who wish to practice data science without data scientists at all. Is it realistic? Some parts of data science can be automated, but not all.