Friday, April 4, 2014

The Great Debate: SAS vs. R

I’ve been recruiting analytics talent for over 30 years, and now over the past few years I have watched open source R seemingly catapult to popularity alongside the proprietary standby SAS. Despite hearing more about R from clients and candidates than ever before, determining whether R was actually more popular than SAS proved difficult. A quick Google search for “R vs. SAS” returns more than a few pages dedicated to each side, as well as several heated LinkedIn discussions relating to the topic, with no definitive answers.


For my latest “flash survey” I wanted to quantify the preferences of the Burtch Works network, and asked one simple question: Which do you prefer to use, R or SAS?


With even more participants responding that couldn’t seem to pick just one or picked neither, this tells me that such a seemingly simple question has a more complex answer. Here are just a few of the entertaining responses we received:

  • “I am a purist, so SAS.”
  • “R - unless you have a ‘both rock’ category – it’s a close one.”
  • “Never learned how to use R. Too damn old.”
  • “R. But isn’t the debate more between R and Python?”
  • “SAS as long as I’m not paying for it.”
  • “SAS. What’s ‘R’? (Joking…)”

Curious as to how these results may vary by factors like industry or years of experience? I know I am! I can’t wait to dig into the data, and in the next few weeks will be posting a full write-up on the blog with our findings. Thanks to all who participated and stay tuned!

Edit: To see our deeper dive into the survey results, including how preferences vary by years of experience, education, region, and industry, please visit The Deep Dive: SAS vs. R

To be the among the first to see all our latest job postings, blog posts and news - including our upcoming Data Scientist Salary Study - be sure to follow Burtch Works on LinkedIn.

7 comments:

Bertin said...

This is a great post. The debate SAS vs. R is definitely going grow because multiple reasons including cost reduction.

http://www.nonodata.com/2014/02/24/case-nosql-database-windows/

Stewart Pratt said...

I'd love to see a break down by years of experience.

Luigi Roggia said...

I think R has much greater opportunities to get further developed in the next few years.
R is opened to big data, parallel computing, cloud solutions and much more.
SAS, like Oracle, Sap and AS400, belongs to the past decades and does not respond anymore to the real needings of the market.

Dean said...

Well, if predictive analytics is the holy grail of business intelligence (as most believe it is), then why should it not be available to everyone? If you had a modern UI for R that anyone without a PhD in stats could use -or- trying to put SAS on everyone's desk, which would you choose? SAS was modern once too, just a very, very, long time ago.

kwanyuetho said...

I use R because it is free. I do not use SAS not because I am against it but I have no access to it.

Ralph Winters said...

I think that SAS has acknowledged the lack of access, and has now enabled free access to academics.

http://www.sas.com/en_us/industry/higher-education/on-demand-for-academics.html

Sean Dayson said...

There is a modern UI for R that anyone can sue: RStudio. It takes no more effort to learn RStudio than Excel, it is incredibly well designed, and could be on the desktop of every analyst today if they wanted it.