What’s Happening With Data Science Hiring?
We have all seen the immediate impacts of COVID. What has been the impact on data science hiring?
Vin Vashishta | Originally Published: May 12th, 2020
The last two months have changed the hiring landscape. How has it effected demand for Data Scientists? There have been several attempts to analyze the job market using surveys and looking at open jobs. None of those tell the story because it is hard to get any sort of granularity or comparison to the overall market. Most data points represent lagging indicators.
Recruiters’ behavior is a strong indicator of demand. They get an early look at new jobs. Internal recruiters know what skills are needed within their teams. They also know what areas are going to be cut back or frozen.
Everyone tells Google the truth, including recruiters. Their searches are easy to track because they all use common search strings. Boolean search is a staple of the industry so Google Trends can give us insights into their behavior.
I have evaluated a few areas of the data science field as well as roles within technology but outside of the field. I used generic searches for resumes as a bit of a baseline. Job searches follow that baseline trend of drop and recovery. The extent of that recovery differs. What follows is a picture of how demand has changed over the last two months.
At the beginning of March, the market started to shift across the board. By that I mean fall off a cliff. The slowest weeks in the hiring world are Thanksgiving and Christmas. In the next images you see those as the sharp V-shaped spikes (the smaller one around the end of June is due to the fiscal year end). The middle of March to the middle of April saw numbers that were close to those weeks with regards to search traffic.
For that month, searching saw the impacts of a shift out of the office and to working from home. It was just a bit better than when everyone is on vacation. That’s expected.
What’s interesting is searches for data scientists in the financial space (red) increased significantly. It peaked in the beginning of March. Although activity has dropped from the peak, it is still above where it started this year.
The medical field (green) is a strong market for data scientists. Search activity is consistently at the top of the range. Over the last six weeks, that search volume has fallen off by 20%.
Retail and manufacturing each have less than half the search volume as finance does. While the medical and financial markets have continued to decline, retail and manufacturing have rebounded to near their average search volume. That’s the upward curve you see at the far right in overlapping blue and yellow.
There is a recovery in demand from the lows, but search data suggests that each industry is recovering at a different pace. Some have declined more than others. Some have rebounded better than others.
Data scientists with python and spark (red) are still in the highest demand with machine learning/data scientist and python (overlapping yellow and blue) close behind. Searches for both dropped but are approaching their average volume again.
Looking at deep learning or NLP data scientist searches (overlapping green and purple), those have not recovered as well. Those fell further and have stayed under their average volume by about 10%.
It looks like companies are still looking for the most common data scientists. Hiring for more specific skills and deep learning have likely been frozen or reduced. It’s also possible that companies are shifting open positions from more senior roles to mid-level roles to save money without losing positions altogether.
Comparing data science hiring to other technical roles, recruiters are still searching harder for data scientists. Compared to product managers, data analysts, business analysts, software developers, and web developers (not pictured), data scientists are still searched for more. Search volumes for data scientists with deep learning experience, the least searched data scientist type, are the same as business analysts.
Keep in mind there are a lot more data analysts than data scientists so the similarity in volume is odd. There are twice as many open data analyst positions than data scientist positions on Indeed. Activity around data science hiring is still stronger than other technical roles.
Demand is recovering. It looks a lot like the first week of the year when the business world is starting to come back from vacations and ramping back up on hiring. Obviously in this case it wasn’t a vacation. We’re not in a January economy. The upswing is slower and may be plateauing to a new range below previous demand.
I mentioned the small spike at the end of the fiscal year. It is about 6 weeks until we hit the start of that again this year. It will be a deeper spike as companies trim their budgets. Hiring freezes will turn to layoffs.
Data science is a fortunate field. Some roles in technology like systems administrators and C# developers are not rebounding as well if you believe the search data. Roles outside of technology are recovering inconsistently. States are recovering inconsistently.
I’ll leave off with a comparison between searches for unemployment benefits versus searches for LinkedIn and Indeed. Unemployment searches surged but have recently fallen behind the two large job boards. You could interpret that as optimism. People are looking for work instead of being discouraged from participating.
However, in states like Nevada, where I live, and Hawaii where I grew up, the search for unemployment is still higher than the job boards. This is going to be a lopsided recovery.
Even in tech, it is time to look at transitioning into safer roles. There are many positions under threat. Those outside of stable fields will need help moving into new roles. The jobs they have lost may never come back.