The results of our 2018 SAS, R, or Python survey have been released! Click here to see our latest analysis.

As many of you probably know, over the past few years we’ve been gauging statistical tool preferences among data scientists and predictive analytics professionals by sending out a flash survey to our network: the first two years weighing SAS vs. R, and then adding Python to the mix last year as their libraries expanded.

Now, as one might imagine, the discussion is always rather spirited in nature – we’ve had over 1,000 responses every year – and reading the comments has become one of our favorite parts of doing the survey, so feel free to chime in!

To keep the comparison simple, we only asked one question: Which do you prefer to use – SAS, R, or Python?



Over the past four years we’ve seen preference for open source tools steadily climbing, with 66% of respondents choosing R or Python this year. Python climbed from 20% in 2016 to 26% this year.


Each year we also match responses to demographic information, to show how these preferences break down by factors like region, industry, education, years of experience, and more.


Similar to last year, the largest proportion of Python supporters are on the West Coast and in the Northeast, however, the Mountain region is close behind the Northeast, and all regions saw at least some increase in Python preference. R preference is highest in the Midwest and SAS preference is highest in the Southeast.

Open source preference remains high in Tech/Telecom and preference for SAS continues to be higher in more regulated industries like Financial Services and Pharmaceuticals.


Professionals with a Ph.D. are the most likely to prefer open source tools, likely due to the prevalence of R and Python usage in research and academic programs, and the foundation of experience it establishes as they move into business.


Preference of open source tools is by far the highest amongst professionals with 5 or less years’ experience. Even as the specific proportions have changed, this trend has remained fairly constant over the years that we’ve done the survey.


Here you can see the comparison in tool preference between traditional predictive analytics professionals working with structured data versus data scientists working with unstructured or streaming data. There is a clear preference for Python among data scientists, whereas predictive analytics professionals’ preferences were more mixed.

This distinction is interesting when compared to the tool preferences of professionals in the Tech industry vs. the Marketing/Advertising Services and Financial Services industries. As we’ve found in our salary studies for Data Science and Predictive Analytics, 41% of data scientists work in Technology, whereas in predictive analytics 52% of professionals work in Marketing/Advertising Services or Financial Services – only 12% work in Technology.

Support for Python grew slightly among predictive analytics professionals, from 16% in 2016 to 20% this year, however among data scientists the increase was much higher – from 53% in 2016 to 69% in 2017.


A Few Lively Comments

We’ve always encouraged respondents to share their opinions so that we can post some of them here, and yet again they came through! Here are just some that we collected:


R – I am a reformed SAS user.

My vote is for SAS. It’s a great end to end tool hands down!

I’m going to have to go with Python, though I’m a recent migrant from R

R is still going strong – SAS is losing steam and Python is just around the corner

I actually use c++, so I guess none??

SAS, but I’m over 40. Those young kids and their cool sounding software…

First choice is Python, second R and thank god we fired SAS.  😉


I’m a SASsy guy!

Python by a billion miles


If you’d like to see the full results from last year, check out our 2016 SAS, R, or Python blog post, or click here to see more flash surveys, including hiring trends and what sorts of perks analytics professionals receive at work.


Interested in our salary research on data scientists and predictive analytics professionals? Download our studies using the button below.

Click to download our free salary reports

38 Responses to “2017 SAS, R, or Python Flash Survey Results”

  1. PaulB

    I’d be curious to know whether you can be any more specific about the types of data scientists in this survey. For example, Gartner segments data scientists into corporate data science, line-of-businesss data science, and “adjacent, maverick” data science. The latter category seems to favor R/Python while the first two probably favor data science platforms using technologies like SPSS and SAS. So it may be that the closer you get to “corporate data science teams” (and I’m guessing the first two categories are a relatively small subset of data scientists overall), then the more likely it is that data scientists are using things like SAS. Any insight on whether this explains the seeming discrepancy between what the research firms say about the platforms and what the various surveys like yours say about the popularity of R among data scientists overall?

    • Linda Burtch

      Thanks for your question, Paul. Data Science is a title that is defined in a seemingly endless number of ways. At Burtch Works we segment people as either Predictive Analytics people(PAPs) or Data Scientists. PAPs work with structured data, and Data Scientists work with unstructured, streaming, messy data. The trend that is emerging in the corporate world now is a blending of the skills – many traditional PAPs are picking up experience with unstructured massive data, and I see many very enthusiastically embracing new tools and technology. It’s a great time to be in analytics! -Linda

  2. Larry Boyer

    Perhaps the most telling chart is the prefered software by age. Because of a learning curve, software use is sticky. SAS should be concerned about the trend and business should be paying attention too so they find the right skills and technology to use for their employees. It’s not your Daddy’s analytics anymore.

  3. Beau Martin

    All of these are overused. While as a company we’ve written plenty of custom programs, why would anyone ever want to write a program to analyze anything unless they absolutely have to? In most cases this becomes a matter of reinventing the wheel, missing the refinements incorporated into existing tools, by people that are likely more experienced and expert at solving a problem than you. While not all existing programs are worthwhile many are. Your first search should be for an existing tool, not a programming language, library, or function.

  4. Francisco Azeredo

    Nice presentation, Linda. I always enjoy voting and seeing the results afterwords. This year, however, I voted for Julia!

  5. pharmaxyz

    My guess is that SAS is losing its market share is due to the cost. Most startups or consulting companies just can’t invest half million dollar a year for a robust SAS server or SAS tools. SAS is a very nice end to end tool, it’s also a very easy-to-use tool for people who are not trained in programming, it’s just very pricey and could be difficult to implement.

    • Phil Rack

      And that’s what WPS was designed to address — cost. Basically, WPS is a SAS Language compiler with a dozen or more data access engines, Procedures and support for the macro language. See for information on what is supported in WPS. Btw, it is used extensively in finance, healthcare and retail. Single workstation licenses start at $1229 and servers start at less than $7K for a small server.

  6. Christopher Shannon

    Prefer R for data exploration and Python for productionalizing.

  7. ML Cottingham

    Perhaps it is because I live embedded in the engineering community, but the choice of SAS vs R or Python seems like it would bias the results in favor of R & Python since most of us cannot justify the cost of SAS since it is not what we do all day, every day.

    • PaulB

      Yes, it depends how you define the question. A survey of free versus non-free software (as is the case for all Python/R vs SAS surveys) will show Python/R dominating, particularly amongst Data Scientists. if the question is market share for analytics software based on revenue (such as the studies done by IDC and other big market research firms) you find SAS growing at 5-8% annually with SAS market share for Advanced and Predictive Analytics software dominating everyone else at over 30% (the other competitors are IBM, Mathworks, etc.) 5-8% is doing pretty well these days in the software world, particularly when you consider that SAS will probably book nearly $1 Billion in revenue this year just in the advanced analytics segment (they also have revenues in the “Business Intelligence” and “Data Integration” segments, as well as other more marketing-oriented segments.)

      • Jon

        The 5-8% SAS growth is happening in a period of exceptional growth for analytics. For now this growth is offsetting their haemorrhaging of share of users to Python and R. It’s a cash cow for now but it is going the way of COBOL.

  8. John G

    To be completely fair (scientific?), I think we should only ask the question of analysts who have pretty good knowledge of all three languages … if there are any such people. Then they could tell us how they stack up against each other doing the same kind of work. I’m a SAS guy and I can remember having a conversation with a younger co-worker who was telling me how great R was compared to SAS. I said “ok, here are three or four things that I regularly do in SAS, show me how to do them in R”. I never got a response.

  9. Jack D

    You can’t judge a language’s “power” by its number of users. That’s a biased metric. Free software will definitely have more users. IS FREE SOFTWARE GOOD? In heavily regulated industries, like banking/insurance, models and software have to be validated. Free codes will have a tough time passing these model risk regulations.

    What if the “free” code has an error and caused the models to spit out bad valuations that made the traders lose billions? Who will you blame?

    I can cite more examples. That’s why SAS is a mainstay in financial services.
    To be qualified to work in model validation for , say a bank, you need a lot of modeling experience, education (MS/PhD) and/or industry certifications. So, the average age of these professionals is higher. Newbies simply aren’t qualified yet to do these tasks.

    As model validation gets more established, free software simply will not cut it.



  1.  R leads, Python gains in 2017 Burtch Works Survey | A bunch of data
  2.  R leads, Python gains in 2017 Burtch Works Survey – Mubashir Qasim
  3.  R leads, Python gains in 2017 Burtch Works Survey – Cloud Data Architect
  4.  R leads, Python gains in 2017 Burtch Works Survey – Iot Portal
  5.  SAS vs. R 2015: The Deeper Dive - Burtch Works
  6.  The Deeper Dive: SAS vs. R - Burtch Works
  7.  R leads, Python gains in 2017 Burtch Works Survey – KAYENRE Technology
  8.  Modernize Your Analytics Tools for Faster Results—at Less Cost | Trifacta
  9.  Python Overtaking R? | NEURALSCULPT.COM
  10.  4 Major Trends Influencing the 2017 Predictive Analytics Hiring Market - Burtch Works
  11.  2018 Predictions for the Analytics & Data Science Hiring Market - Burtch Works
  12.  Is Python Edging Out R in the Data Science Space?
  13.  Reflections on the Job Market – Stella Min
  14.  Python or R ? that is the question – Siddartha Kandikonda’s blog
  15.  2018 Data Science & Analytics Hiring Survey Results - Burtch Works
  16.  Data Scientists: How to Succeed in America’s Top Job - Burtch Works
  17.  6 Tips for Hiring Data Scientists (Besides High Salaries) - Burtch Works
  18.  2018 SAS, R, or Python Survey Results: Which do Data Scientists & Analytics Pros Prefer? - Burtch Works
  19.  Etapas delduelo para científicos de datos y de TI lograndoque el código libre funcione en las corporaciones paranoicas
  20.  Etapas del duelo para científicos de datos y de TI – Business Empresarial
  21.  What’s the Best Programming Language for Data Science?
  22.  R Programming Language's Reported Decline: Premature?
  23.  R Risks Python Swallowing It Whole: TIOBE

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.