#1: Why is SAS Software after +40 years of successful market presence still a key building block of Analytics architecture for many organizations?

I am working in the Analytics field for nearly 40 years. I started my career as a statistician, working for various public and private organizations and moved to the Data & Analytics software industry in the second quarter of my career. I joined SAS 20 years ago focusing my expertise on Data & Analytics for various industries.

I have the pleasure to work with many large organizations using SAS Software. To better plan their future Data and Analytics roadmap, we organize workshops where we assess their current Analytics usage from the business side and the architecture in place. The aim of those workshops is to deliver our best recommendations in terms of Data & Analytics architecture and business solution based on the collected insights. The targeted outcome is a roadmap of SAS Modernization according the IT and business strategy from a Data & Analytics perspective.

Through this long experience in the field working on SAS modernization for SAS customers, I’ll try to answer the questions of why and how many organizations are reinvesting on SAS in 2021 and beyond to modernize their Data & Analytics environment through the prism of the business transformation principles: people, process, technology and data.

1 – The first question I would like to answer is: Why is SAS Software after +40 years of successful market presence still a key building block of Analytics architecture for many organizations?

SAS is still a key building block of Data & Analytics

SAS Software has been applied these last decades to almost any kind of business problem in every industry. There are no limits given the large choice of analytics techniques provided by SAS from descriptive to predictive and prescriptive analytics. Business users are autonomous with SAS in the process of building and automating their analytics process with an end-to-end software solution. They also can collaborate and share common development and data very easily. This approach shortens the analytics life cycle avoiding complex IT project management, increasing business productivity overall. I observe today that SAS is still a key building block of Data & Analytics strategy for many organizations.

Business users trust SAS for their efficiency

Due to the market presence of SAS these last decades, there are many SAS users relying on SAS Software in the industries where SAS is strong like Banking, Insurance, Public Sector, Health Science. In these industries, SAS users trust SAS Software to deliver the expected insights from their enterprise data. They have the skills to prepare and analyze any kind of structured data though SAS code or SAS User interface. SAS code is simple and concise and answers their business needs with accuracy. SAS Platform is a proven and reliable solution for their daily work supporting their business processes.

It is true that nowadays, the new analytics worker generation is learning Python or R at School and less SAS. But today in an organization when I ask the question of the number of SAS users versus the number of Python/R users, I often get an answer of hundreds of SAS users versus tens of Python/R users. The trend of the numbers of skilled Analytics talents is moving to high growth from year to year. Therefore, there is a demand to democratize Analytics from the business to scale up the Analytics potential. Specific coding languages like Python, R are sometimes complex, have a limited scale and therefore slow down the positive impact on business. I don’t expect a rapid swap between these skilled analytics groups, SAS will remain one of them for sure in those organizations for a certain period of time.

Enterprise by design

SAS® 9 technology is based on a portable language and platform by design running on large enterprise servers and sharing infrastructure resources for thousands of SAS users or SAS jobs. When I speak to SAS business users, they appreciate the ability to manage their Analytics with one software from data access, data transformation, advanced analytics to reporting for two main purposes: ad hoc study or shared business application.

IT on their side appreciates the light administration, just providing data access, servers and disks. IT acknowledges that SAS technology is simple, robust and need very few administration tasks. The integration by design principle ensures SAS can be used to operationalize analytics by normal business users without IT effort.

In a few organizations, I see sometimes IT involved in the business process preparing SAS datamart or reviewing the SAS code to better automate and put the recurrent jobs in production. This is certainly a very good best practice avoiding data redundancy, decreasing cost and improving business efficiency.

SAS is doing the job

With SAS, any amount of enterprise data can be extracted, loaded, transformed and stored in a business data format ready to deliver insights for the business on a regular basis. When I assess the SAS usage on customer site, it is common to observe that the business has created tens and sometimes hundreds of terabytes of SAS files representing a lot of business value but also a heavy burden for IT from a safety, security and compliance perspective. The autonomy given to the business to access and transform any kind of data without a lengthy IT development process or a limited business intelligence tool is probably the most appealing capability of SAS. For the business, SAS is reliable and scalable from a data processing perspective and never fail even with large table and always deliver the results at the end. Business users trust SAS.

My personal view based on what I observe at the customer sites is that SAS Software today still “does the Analytics job” in many organizations running in production millions of SAS critical workloads for the business.

Despite a high-cost perception from customers

During the last decade, other analytics technologies mainly based on open-source technology like Python and R, promised a new Analytics age. Indeed, they delivered quickly excellent results in term of innovation but surprisingly are expanding much slowly. When I ask the question on the number of Python applications in production, often just a few are in production compare to hundreds/thousands of SAS batch jobs running nightly on servers.

The question of replacing SAS with other Analytics technologies is often at stake in these organizations given the total cost of ownership of Data & Analytics processed by SAS in term of infrastructure both compute and storage. In my humble opinion, the explanation of the high total cost of ownership of SAS is more due to the cost of processing large volume of business data stored in raw formats with poor data quality and very few IT governance.

Creating two Analytics silos

After many years of investing into Open-Source software, it becomes obvious that it will take years to support all the Data & Analytics scope of the organization. Customers and especially the IT department realize the central role played by SAS in their business processes. Avoiding the dogmatic attitude, these customers are taking the decision to keep SAS for certain business scope and develop Open-Source technologies for new use cases based on new variations on analytics algorithms like machine learning, text mining or image recognition. But this strategy creates two analytics silos not sharing data and analytics. I will come back on this issue in a next blog.

So, If you consider keeping SAS in your Data & Analytics strategy to support your business, should you decide to keep SAS at his current version or modernize SAS technologies with new options?

I will try to answer these questions in the next blog post of this series.