Corporate Profile: SAS – Where it’s been and where it’s headed
By Kathy Lange
AS is the world’s largest privately held software company, with 70,000 customer sites in 138 countries, including 93 of the top 100 companies on the 2014 Fortune Global 500 list. SAS’ strengths in advanced analytics – statistics, data mining, machine learning, econometrics, forecasting and operations research – along with data management capabilities lie at the heart of its success since it was founded in 1976.
SAS software grew out of a North Carolina State University project to analyze crop yield data for the U.S. Department of Agriculture. It didn’t take long for industry to also see the value of SAS software for data analysis. Growth in the SAS product portfolio has paralleled industry increases in computing power. Data transformations and numerical algorithms are a foundation of product research and development at SAS.
The SAS product portfolio evolved significantly over the years with the addition of business-focused solutions that serve a variety of industry needs while building upon SAS’ highly regarded analytics software. As SAS extended its reach into industry-specific solutions, it also pushed into new geographical markets. The result: With annual sales of $3 billion, SAS is now the preeminent supplier of advanced analytics to businesses, governments and other organizations.
An aerial view of SAS’ picturesque world headquarters campus in Cary, N.C.
The uses of SAS software are innumerable, but here are a few examples:
- A retailer runs markdown optimizations in minutes rather than days, increasing profit and revenue.
- A financial institution uses data to spot anomalies so quickly that the firm can stop fraudulent credit card purchases before the transaction is completed.
- A pharmacy management company lowers costs and keeps patients out of the hospital by identifying those who need assistance managing their chronic conditions and prescription medications.
- An insurance company uses predictive analytics to determine the best new lines of business to enter.
“We have a very close working relationship with our customers,’’ says Radhika Kulkarni, SAS vice president of Advanced Analytics Research and Development, and a longtime advocate of operations research and advanced analytics. “We work on some of the most challenging problems customers need help solving.”
Reputation Built with People
SAS has had a remarkable uninterrupted run on Fortune’s list of 100 Best Companies to Work For in the U.S., where it is currently ranked second . This springs from SAS CEO Jim Goodnight’s philosophy: “Treat employees like they make a difference and they will.”
SAS attracts and retains creative employees because doing so directly contributes to consistent revenue growth and profitability. As CEO, Goodnight is fond of saying the company’s most important “assets” drive out the gate every evening, and he wants to see them drive back in the gate the next morning.
“In our industry, rapid innovation and extreme customer care are essential,” Goodnight says. “The best way to make that happen is by supporting people. We have spent decades perfecting a culture and work environment that encourages creativity by addressing the day-to-day stresses and concerns that employees inevitably bring to work.”
The approach extends to its global offices. Last year, SAS ranked No. 2 on the elite Top 25 World’s Best Multinational Workplaces  list from Great Place to Work® .
SAS Advanced Analytics and Operations Research
With a culture that supports creativity and innovation, SAS helped shape the role advanced analytics and operations research plays in organizations. SAS continues to support the demands of researchers and practitioners by adding new advanced analytics functionality every year, supported by a highly trained staff of research statistician developers and operations research specialists with advanced degrees.
SAS has significantly increased R&D investment in its entire advanced analytics portfolio but especially in operations research, and more specifically, optimization, discrete event simulation and network analysis. SAS’ new optimization modeling language and a full suite of optimization solvers have been developed over the last decade, along with a new discrete event simulation application. The family of optimization solvers play an important supporting role in other analytical tools from SAS, but they also enable optimization-based business solutions like SAS Marketing Optimization and SAS Revenue Management and SAS Pricing Optimization, which extend the benefit of operations research to a broader class of business users and organizations.
The Advanced Analytics R&D organization also houses the Advanced Analytics and Optimization Services group that supports customers with modeling and algorithmic expertise. One ongoing project of this group was joint work with a large consumer packaged goods company, where SAS collaborated on the application of statistical and optimization techniques for product portfolio analysis. This project was a finalist for the 2014 Wagner Prize for Excellence in Operations Research Practice and involved developing an innovative algorithmic and computational solution.
Another example is a discrete event simulation customer project from Duke University Hospital. The organization’s neonatal intensive care unit used SAS Simulation Studio to study the effect of staffing levels and other policies on health outcomes and operational efficiency.
“When we can engage with customers, it is very rewarding for us as researchers and developers, since we can see tangible, real-world results from our R&D efforts,” says Manoj Chari, SAS senior director, Advanced Analytics R&D. He has overall responsibility for SAS R&D Operations Research and related areas.
SAS R&D endeavors to share insights with the broader research and practitioner community by active involvement in professional communities, such as INFORMS and the American Statistical Association.
SAS OR R&D strives for excellence with the goal of continuing SAS’ leadership in the advanced analytics market. In the era of big data, many of the challenges that customers face are the sheer volume, scale and complexity of data, as well as new classes of business questions.
A significant amount of current SAS R&D efforts involve developing software infrastructure, as well as techniques to address these new data paradigms and computing architectures. SAS has significantly invested in developing integration with Hadoop-based data ecosystems. SAS has extended its analytics product portfolio, enabling key functions to run in high-performance, massively parallel distributed computed environments, including batch analytics computations, as well as interactive and visual analysis of any size data.
Support for the Next Generation
With its roots in academia, SAS has a strong commitment to supporting the development of future operations research practitioners, statisticians and data scientists. The company’s SAS Analytics U program provides free or low-cost licenses for the advanced analytics tools most used by academics, with the option to install the solutions onsite or access them on demand. Support is available for undergraduates, graduate students, academic research and institutional research.
SAS offers scholarships to conferences, opportunities to present at regional SAS user group conferences, SAS certification and a program known as the Data Mining Shootout – a competition to solve real-world data mining problems. Winners receive cash prizes and a trip to the annual SAS Analytics Conference, as well as a case study competition co-sponsored with the Analytics Section of INFORMS. The case study competition winners attend the INFORMS Conference on Business Analytics and Operations Research.
Classes and students using SAS routinely win praise for solving real-world problems:
- Oklahoma State University students modeled ways to market and time flu shot campaigns to reduce the viral illness’ impact on the community.
- Lehigh University students helped define a market opportunity in using liquefied natural gas and fuel cells to replace diesel fuel for selected railroad and mining industries.
- Franklin College students modeled various crimes as a function of time, weather, demographic information and other phenomena in multiple cities, and to recommend plans of action to prevent and deter these crimes based on future prediction of those variables.
SAS’ commitment to education also includes financial and practical support of many of the proliferating master of analytics programs, including the flagship N.C. State University Institute for Advanced Analytics. This program offers a master’s degree in analytics with the goal of producing individuals who have mastered complex methods and tools for large-scale data modeling, and who have a passion for solving challenging problems through teamwork.
A New INFORMS Fellow
SAS has been a longtime supporter of the INFORMS community, and as one recent example SAS R&D VP Kulkarni was inducted as an INFORMS Fellow in recognition of her contributions to practice and service to the profession. Kulkarni was hired 31 years ago as one of the first SAS/OR developers, but over the years she has taken increasingly significant leadership roles in the SAS R&D organization. Kulkarni was primarily responsible for laying down the vision of the role that operations research could play in the larger SAS Advanced Analytics portfolio and building an R&D team that could implement the vision. She has also been very active in supporting the INFORMS community through her own services and through her encouragement of her team’s professional development and organizational service.
“My selection as a Fellow is recognition of the contribution that operations research has made to the advanced analytics portfolio at SAS in recent years,” Kulkarni says. “It is also a recognition of SAS support for INFORMS, especially in its efforts to become a premier organization for analytics professionals.”
Adds Armistead Sapp, SAS executive vice president and chief technology officer: “Radhika is a passionate voice for the critical role of operations research and advanced analytics at our company. The many hours she spends promoting the value of advanced analytics to business leaders have notably benefited our customers who continue to increase valuable insight into their enterprises. The analytics industry as a whole is better for her strong advocacy.”
Kulkarni’s ongoing advocacy of advanced analytics is grounded in building cutting-edge research into solutions that really matter: “We are not creating algorithms in an ivory tower and then throwing them over the fence and expecting that somebody will use them someday,” she says. “We ask our customers what problems they are working on, and we drive to solve them like no other company.”
Kathy Lange is senior director, SAS Americas Technology Practice. She is a member of INFORMS.
- 41Ducati Corse, the racing department of Ducati Motor Holding, a world leader in sports motorcycle manufacturing, is working with Accenture to integrate the Internet of Things (IoT) and artificial intelligence technologies into the testing of its MotoGP racing bikes. Ducati Corse wants to make testing its race bikes faster, cheaper…
- 39Nate Silver, author of “The Signal and the Noise” and founder of FiveThirtyEight.com, will discuss using data to make predictions and decisions at Discovery Summit 2013. The annual analytics conference, hosted by the JMP business unit of SAS, convenes Sept. 9-12 at the Grand Hyatt San Antonio.
- 38FEATURES Fulfilling the promise of analytics By Chris Mazzei Strategy, leadership and consumption: The keys to getting the most from big data and analytics focus on the human element. How to get the most out of data lakes By Sean Martin A handful of requisite business skills that facilitate self-service…
- 36May/June 2012 Statistical modelers urged to embrace machine learning, open-source tools for the road ahead. By Sameer Chopra My thesis below addresses the following points: While statistical modeling is not going away, analytics groups are advised to leverage machine-learning approaches as well. While traditional statistical modeling software packages are not…