How SAS Has Remained Relevant Over a 44-Year Timespan

Oliver Schabenberger of SASsoftware | eSPEAKS with Chris Preimesberger

eWEEK Editor Chris Preimesberger talks with EVP, COO & CTO Oliver Schabenberger about the new partnership between SAS and Microsoft.

SAS, which stands for Statistical Analysis System, was an early intelligence software pioneer in the 1970s and has stayed in the forefront ever since, despite intense competition from numerous other companies. eWEEK spends some time with the CEO-to-be of the Cary, N.C.-based software giant.

Successful organizations must adapt to changing customer needs. A 44-year-old company could get stale and stuck in its ways, but not SAS software. It has remained a pioneer in its industry and is constantly reimagining better ways to provide data analytics and surrounding services for its customers.

One of the ways SAS has better provided for customers is by partnering with Microsoft Azure Cloud. In this eSPEAKS video, we cover:

  • The expansion of the partnership between SAS software and Microsoft
  • Why Microsoft was chosen
  • Similarities between the two companies
  • How SAS has remained a successful company over the years

To provide an insider's perspective, eWEEK spoke with: Oliver Schabenberger, EVP, COO and CTO, SAS software. Host: Chris Preimesberger, Editor, eWEEK.


With the recent announcements of a strategic partnership with Microsoft for the Azure cloud and the anticipated release of the cloud-native rewrite of SAS Viya, one of the oldest independent software providers has reached a key milestone in its generational transformation.

SAS (Statistical Analysis System), started up by Jim Goodnight in 1976, is one of the few survivors of the first generation of independent software firms; Oracle, co-founded by Larry Ellison in 1977, also is in that camp. A common thread for both firms has been the continued strong leadership of the founders who to this day remain very much engaged. And if you add in a later comer--Informatica--you find yet another company that has had to undergo a similar transformation exercise.

For years, SAS was synonymous with data mining, a discipline that has since morphed into data science. The differences between the two are subjective, but the transition echoes the expansion of the data estate and the Cambrian explosion of open-source modeling frameworks, programming languages, relatively inexpensive cloud compute, and endless on-premises and cloud-based storage that have made this a far more multidimensional discipline.

No longer are quants, data miners or now, data scientists, defined by their knowledge of the SAS language, because there are many paths to generating those analytics models.

New-Gen Development Approach 

To say that SAS faces generational change is an understatement and hardly a new development. For the data and analytics sector at large, generation change has been a fact of life since the early days of “big data” more than a decade ago. That’s where the spotlight on innovation in data management and analytics started coming from innovators such as Google, Facebook, LinkedIn, Twitter and a few others, who had to invent their own technologies to deal with analytic challenges blowing the scale off anything available commercially.

Likewise, SAS’ transformational journey among classic enterprise software companies is hardly unique. For SAS, what’s notable are the milestones that are occurring this week. It is finally embracing cloud-native architecture in its flagship software, and it has entered into a strategic partnership with Microsoft that could see SAS’ analytic services eventually become ubiquitous if your organization uses Azure services, ranging from analytics and machine learning to Microsoft 365. That won’t happen tomorrow, but if SAS is successful, not only in transforming its software but also the way it conducts business, it could open itself up to a much wider audience. Or, as EVP Oliver Schabenberger puts it: “Analytics hidden in plain sight.”

It takes a while to turn around a ship. The shift of SAS Viya to a cloud-native, microservices and container-based architecture is the culmination of a process that started roughly five years ago. At the time, there was the realization that SAS software needed more than a new face. It had to meet data scientists where they lived, and it needed to broaden the reach of SAS analytics beyond the elite pools of SAS practitioners. Traditionally, SAS skills were so specialized that it practically defined its own job title.

Exciting the New Generation of Data Science Grads

And so SAS bit the bullet. It started accommodating open source--if you can’t beat ‘em, join ‘em. That exposes open-source algorithms and modeling frameworks to the SAS base, but what about exciting the new generation of data science grads who are enamored with Python and Anaconda? Hold that thought.

The most visible sign of change is the forthcoming Viya 4 release that will become generally available in Q4. SAS had a classic enterprise software portfolio of separate tools that were added over the years, many with point-to-point interfaces or connections. Viya was a chance to not just design a single engine but also reimagine and refactor the portfolio. 

A good example is Visual Investigator, which mashes up functions from a bunch of tools in the classic portfolio including Visual StatisticsData Prep and Enterprise Miner. When originally conceived, the primitives of Viya were built as microservices, but because the customer base was not ready for cloud (and standards like Kubernetes had yet to happen), it was released as traditional, monolithic software. You could run it on-premises or in the cloud, but it was packaged and released using the typical waterfall cadence of legacy software.

What’s being formally announced this week is Viya reimagined for containers. As for the managed cloud service, that will come out later this year starting on Azure, it will be with ongoing upgrades with the cadence expected in the cloud. As a customer, you will be able to choose, within a range, the release intervals you want based on your change management schedules. Because it’s containerized, the analytics will be decoupled from the underlying operating environment.

SAS Will Need to Reshape Internally

Of course, as anyone on the business side will tell you, technology is typically the easy part of the battle. Products become online services. Development morphs to continuous integration/continuous delivery modes. The hard part for SAS will be reshaping itself internally to deal with a customer base that should theoretically grow more diverse and expand through channels. 

It means continuously engaging with customers through digital and analog means rather than relying on annual contract talks and support calls. It means not only enabling use of open source but also becoming part of the community and contributing to it. It also means taking more risks with the ability to fail fast, learn and move on. Some call this digital transformation, but this is simply a means to an end. We’ll stick with generational change because that also means thinking differently.

SAS’ story is not unique, and neither is the prescription: the need for SAS to deploy its own products. As it delivers services along CI/CD principles, that’s also the way the rest of the business has to think––not just the dev team. For SAS, the story of generational transformation is one that won’t start or end with any single announcement.

Tony Baer of PUND-IT contributed to this article, which originally ran in eWEEK earlier this year.