SAS is #1…In Plans to Discontinue Use

I’ve been tracking The Popularity of Data Analysis Software for many years now, and a clear trend is the decline of the market share of the bigger analytics firms, notably SAS and SPSS. Many people have interpreted my comments as implying the decline in the revenue of those companies. But the fields involved in analytics (statistics, data mining, analytics, data science, etc.) have been exploding in popularity, so having a smaller slice of a much bigger pie still leaves billions in revenue for the big players.

Each year, the Gartner Group, “the world’s leading information technology research and advisory company”, collects data in a survey of the customers of 42 business intelligence firms. They recently released the data on the customers’ plans to discontinue use of their current software in one to three years. The results are shown in the figure below. Over 16% of the SAS Institute customers surveyed reported considering discontinuing their use of the software, the highest of any of the vendors shown. It will be interesting to see if this will actually lead to an eventual decline in revenue. Although I have helped quite a few organizations migrate from SAS to R, I would be surprised to see SAS Institute’s revenue decline. They offer excellent software and service which I still use, though not anywhere near as much as R.

The full Gartner report is available here.

SAS Attrition Plot


About Bob Muenchen

I help researchers analyze their data, teach workshops on data analysis using R, and write books about research computing.
This entry was posted in Analytics, R, SAS, SPSS, Statistics, Uncategorized. Bookmark the permalink.

15 Responses to SAS is #1…In Plans to Discontinue Use

  1. Bob,
    While I agree that many organizations currently using SAS are considering alternatives, it’s important to note that this Gartner report focuses on BI, where SAS has never been particularly strong. Out of 1,500 sites surveyed by Gartner, only 30 say they use SAS BI.

    Also note that SAS and others are losing out to newer products like Tableau and Qlik, which are very easy to use. In other words, this survey is bad news for SAS, but not good news for R.



    • Bob Muenchen says:

      Hi Thomas,

      Thanks for that interesting information. Given how much emphasis SAS Institute has given BI on its web site and at conferences over the years, I’m surprised that they haven’t made more headway there.

      Easy-to-use software is more widely used than programming languages and, since the release of newer software such as Tableau, Spotfire, KNIME, and RapidMiner, it’s likely to become even more so in the future. Luckily for R fans, virtually all of them can call R.


  2. R Lover says:

    While R is very popular, there are some areas where it’s lacking behind python. Example – deep learning. There is just one library darch, but apart from that There are not many good ones.

    Another area R particularly lacks is natural language processing. Libraries like NLTK beats R very easily, when it comes to NLP.

    R has to pick up on these areas very fast, if it’s to stay competitive…..

    • Bob Muenchen says:

      Hi R Lover,

      Thanks for the comments. Competition is a wonderful thing!


    • rob says:

      eh.. disagree with the implications.

      Of course R will be lacking in these two and many other areas in which it was not designed/created for (or at least, not with these things in mind).

      R will be the leader in statistics for the foreseeable future; it will make up ground in other applications where people are coming from an r background and want to use r in different areas.. but may take a long time; python will make up ground in statistics, but it would take a long time to catch up also.

      • Robert Young says:

        Well. The key to that observation is the word “statistics”. More and more, BI, and quant generally, is being consumed by the “Big Data” meme. In that venue, R doesn’t do as well. First, because it’s memory resident, and fixes to support RBAR style calculation aren’t native (likely, can’t be, but that’s a guess). And second, because Big Data doesn’t do much with statistics, if one exclude descriptive statistics from the definition.

        R is best suited to sample based inferential problems, whether one adopts frequentist or Bayesian approach. No, I haven’t categorized the 6,000+ packages on CRAN. Yet. If you have every attribute on every customer, for example, there’s not much need for inference. There are other things one can do with such data, but they tend into the area of OR.

      • jaehyeonkim says:

        I’m not sure if memory limitation is still be a big issue for R. These days it doesn’t seem to be expensive to run R on cloud. Also it is reported that even big companies like Yahoo or MS find it more effective to do much of the analysis in a dedicated single server rather than multiple servers. Of course raw data should be processed appropriately for it beforehand but this seems to be what SAS or R suits.

        As far as I’ve informed, inference can be extended to prediction. And, for prediction, I don’t think anyone has all data.

        It is hardly possible to imagine a ‘one-size-fit-all’ type of tool and analysis workflow should be ‘pipelined’ in an effective way. In my opinion, R can play quite a useful role in a workflow. And the fact that R can be called or embeded in other languages or BI tools could be a definite advantage to SAS.

      • Bob Muenchen says:

        Hi Jaehyeonkim,

        Good point. Large memory cloud machines are cheap for short-term use and memory is pretty cheap too.


      • jaehyeonkim says:

        this seems to be what SAS or R suits. –> the doesn’t seem to what SAS or R aims for.

    • Of course, if you really know what you are doing you know that Deep Learning can be done with most neural network packages, of which there are many in R.

      • R Lover says:

        That’s a defeatist mindset. If we use that logic, we can as well argue that we can go to C, instead of R. It’s also a matter of conveniecne.. Not just what can be put together..

      • No, doing Deep Learning in C would be very difficult because you would have to build the foundation algorithm from scratch. Any well-informed user can perform Deep Learning using a number of R packages, without a lot of coding.

        Of course, if you’re not well-informed…

  3. SAS are moving their BI applications to a new set of tools. The evaluated tools of this research , probably, SAS BI Dashboard and SAS Web Report Studio are being replaced by a new generation of an integrated analytics platform, ready to Big Data capabilities, such as Hadoop, In-Memory and In-Database processing. I´ve being worked for many year with differente BI applications and believe that SAS is now achieving the ideal point between BI capabilities and statitistics integrated in single and robust platform. SAS spent time and money to get a better interface and develop a interactive and easy-to-use tool,. See more and say me what you think about:

    • Bob Muenchen says:

      Hi Washington,

      Thanks for the info on their new tools. I wish them luck with them!


    • Robert Young says:

      I looked at the link, but didn’t find any mention of in-database processing. I gather that means something along the lines of PL/R (postgres) and Oracle and SAP functionality to run R (or other stat engine) routines from the SQL statements.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s