Editor’s note: Michael Semaca is operations associate at recruiting firm Burtch Works in Evanston, Ill.
In recent years, the distinction between the career paths within predictive analytics and data science has lessened, but despite this convergence, predictive analytics professionals (PAPs) and data scientists have some key differences that warrant direct comparison.
PAPs possess well-refined quantitative skills that can be used to describe or derive insights from large data sets. At Burtch Works we classify data scientists as a subset of PAPs; in addition to the skills that predictive analytics professionals possess, they also have the computer science skills necessary to work with unstructured or streaming data, including formats like audio or video files.
Burtch Works recently released its annual comprehensive salary study focusing on these two increasingly popular areas of the analytics field. While Burtch Works has been publishing these reports for six years, the 2019 edition is the first to combine the two professions into a single report, with the intent of comparing the two groups.
The study focused on compensation and various demographic information, including education, residency status, industry, gender and experience levels. The report also included general trends for the overall data science and analytics hiring market.
2019 quantitative hiring market trends
The most noticeable market trend was that professionals in these fields have more options than ever as demand for their talents has grown. An increasing number of companies and industries employ PAPs and data scientists, which has made these fields more technically diverse.
Another area where there are more options than ever are the education options for those looking to enter these fields. New education programs, such as coding bootcamps and online programs, have allowed professionals to transition their skills to data science and analytics roles from related fields. Although the majority of data scientists and analytics professionals have a graduate degree, the study found that an increasing number of professionals with eight or less years’ experience had a bachelor’s degree.
As more professionals are supplementing their skill sets through online programs, more and more companies are relying on technical screens testing to better evaluate talent coming from a variety of educational backgrounds. These technical screens allow companies to evaluate a candidate’s technique and tool fluency, as well as problem-solving capabilities.
2019 salaries of predictive analytics professionals and data scientists
Compensation data was conducted with a sample size of 1,840 PAPs and 421 data scientists, with each group being categorized into six job levels indicating their level of work experience and/or management responsibility. Changes in salaries were compared using the median of the years 2019 versus 2018.
These job levels were defined the same way for both data scientists and PAPs, and were divided between individual contributors and managers. Individual contributors at level one typically have zero to three years of experience, those at level two have four to eight years of experience and those at level three are subject matter experts with nine or more years of experience.
For managers, job level is typically informed by team size and management responsibility. Those at level one typically have one to three years reports and are more tactical or leading small groups; level two typically have four to nine reports and are responsible for executing strategy with moderately–sized teams; and those at level three, the executive level, have more than 10 reports and are responsible for determining strategy for large teams.
For predictive analytics professionals who were individual contributors, median base salaries increased at all job levels, with the largest increase being at level one. The story was a little different for PAPs that were managers. At levels one and two, salaries remained steady, while level three managers had an increase of 4%. Across all levels of all PAPs, the mean and 75th percentile salaries increased.
The median salaries for individual contributor PAPs in the sample are as follows: $80,000 for level one, $97,000 for level two and $132,000 for level three. Level one PAPs saw an increase of 4%, while levels two and three each saw an increase of 2%.
For PAP managers, the study found that level one had a median salary of $130,000, which was unchanged from 2018. Level two also showed no change, and a median salary of $180,000. Level three managers saw an increase of 4% in their compensation, which rose to $248,500.
The story was similar for data scientists, who also saw slight increases or no change to their salaries when compared to 2018.
Individual data scientists at level one had a median salary of $95,000, which showed no change from last year. At level two, the median salary was $130,000, an increase of 1% from 2018. And for level three, the median was $167,000, also a 1% increase from 2018.
Compensation at levels one and two increased slightly for data scientist managers, who had median salaries of $146,000, an increase of 1%, and $190,000, an increase of 3%, respectively. Level three managers saw no change in their salaries, with a median salary of $250,000.
With this data in mind, Burtch Works then compared the salaries of professionals in predictive analytics versus those in data science roles. At all job levels, data scientists earned more than their PAP counterparts. The largest difference for individual contributors was 34% at level two and the smallest at level one, where there was a 19% difference in compensation.
Data scientist managers still earned more than PAP managers, but the difference between the two was less pronounced than for individual contributors. The largest difference here was between managers at level one, where data scientists earned 12% more than PAPs at this level; at level three, this shrank to a difference of 1%.
Education profiles
Burtch Works also used the data collected from the sample to examine and compare demographic trends for both PAPs and data scientists, including their education profile.
The study found that 86% of PAPs sampled had an advanced degree, with 71% having a master’s degree and another 15% holding a Ph.D. Data scientists were even more likely to have an advanced degree, with 94% having one; 47% held a master’s degree and another 47% held a Ph.D.
As one would expect, PAPs with an advanced degree tend to earn more than those with a bachelor’s degree. PAPs with a Ph.D. out-earned holders of a master’s degree at all levels.
Data scientists who hold a Ph.D. earn more than those with a master’s degree at every level except level two and three managers. In some managerial positions at higher levels, leadership and management skills often matter more to employers than education.
Both PAPs and data scientists were most likely to have studied mathematics and statistics based on their highest degree. However, PAPs were significantly more likely than data scientists to have a degree in economics or business, while data scientists were more likely to have a computer science or engineering degree. Among others, business degrees include master’s in business analytics degrees, which have become increasingly popular in recent years.
Median compensation for PAPs that were individual contributors for those with a master’s degree are as follows: $80,000 at level one, $95,000 at level two and $130,000 at level three. Holders of a Ph.D. earned $10,000 to $20,000 more than their counterparts with a master’s degree depending on the level of experience.
Data scientists with a master’s degree had a median salary of $90,000 at level one, $125,000 at level two and $160,000 at level three. Their counterparts with a Ph.D. can earned $11,000 more at level one, $12,000 more at level two and $20,000 more at level three.
Gender
Gender distribution between PAPs and data scientists was also studied. For PAPs, the sample was 74% men and 26% women; for data scientists, it was 83% male. Compared to last year, the sample showed an increase of 2 percentage points for women in predictive analytics and an increase of 5 percentage points of women data scientists.
The largest percentage of women for both data scientists and PAPs was observed among level one individual contributors, indicating that more women are entering the field. This may contribute to a more even gender ratio as these early career workers become more experienced.
Women who were PAPs earned slightly less than men at most levels for both individual contributors and managers except level one managers, where women’s salaries were 102% of men’s salaries. This pay gap was relatively small, however; the largest gender salary difference in the sample was among level three individual contributors, where women earned 94% of what men did.
For more demographic trends as well as general market trends for data science and analytics, the full report can be downloaded for free on Burtch Works’ website (registration required).