As executive vice president, chief operating officer and chief technology officer of SAS Institute Inc., a North Carolina, US-based analytics software provider, Oliver Schabenberger is tasked with setting the technology direction and executing the company’s strategic objectives. An author of statistical textbooks with a PhD from Virginia Tech, Schabenberger used to teach at Michigan State University and Virginia Tech before joining SAS in 2002. He will be speaking at EmTech India 2018—an emerging technology conference organized by Mint and MIT Technology Review—on 9 March in Gurgaon. In an email interview, he talks about the evolution of analytics, and how machine learning and other cutting-edge advances can help companies become data-driven organizations. Edited excerpts:
How is data analytics evolving from being descriptive to predictive in nature?
Descriptive analytics help you understand what happened in the past; predictive analytics is the use of data, statistical algorithms and machine learning techniques to identify the likelihood of future outcomes based on historical data. Today, we want to go beyond knowing and understanding what has already happened to provide a good assessment of what will happen in the future. Increasingly, organizations are turning to predictive analytics to gain a competitive advantage. Why now? Growing volumes and types of data, and more interest in using data to produce valuable insights; faster, cheaper computers; easier-to-use software; and a need for competitive differentiation.
With interactive and easy-to-use software becoming more prevalent, predictive analytics is no longer just the domain of mathematicians and statisticians. Business analysts and line-of-business experts are using these technologies as well.
What are the key challenges faced by CIOs and other CXOs in deriving more practical insights from the piles of data they have?
The challenges fall into three areas. First, it’s about resources. Organizations simply don’t have enough people who can manage, manipulate and analyse the data that they accumulate.
The next challenge comes from the types of data that organizations are dealing with. This goes well beyond a “Big Data problem” to a “bigger than Big Data problem”. Data is now stored in a variety of architectures, both on- and off-premises. Some of that data (but not all) is under the direct control of the organization. Data volumes continue to accelerate, driven by streaming data... All of this adds up to a complex environment from an information technology (IT) perspective that is getting more complex all the time.
Finally, many organizations face the internal obstacle of territory, where different business units view applications, data or other IT assets as “theirs” and not a shared resource. When data exists in silos, one version of the truth is harder to come by.
How are the new developments in Big Data and artificial intelligence (AI) affecting the traditional mainstay of business intelligence (BI)?
We live in the era of Big Data. Data volumes will continue to increase, amplified by growing connectivity between us and inanimate objects. And in this era, automation has become a necessity. Customers are expecting BI vendors to provide advanced “data ingestion and preparation” tools that can run close to the data. Customers also want to see AI and machine learning (ML) employed to streamline data selection and data combination ahead of BI activities.
Analytics is key to deriving insight and value from the data deluge. Automation of analytics has become a necessity in order to tackle the deluge of data-driven problems. It enables us to focus our attention on higher-value tasks, and gain operational efficiency and repeatability. ML and AI are the technologies through which this automation is achieved. Advances in cognitive computing and AI are leading to new capabilities in BI: self-service data preparation that applies AI and ML to cleanse data, correct entries, and merge data sources; guided business analytics that assist a user to apply analytic tech; and natural language generation to explain the results of analytics. AI and ML are making BI more suggestive, augmenting our skills.
There are dozens of analytics tools available in the market. How do you think should CXOs make a purchase decision?
CXOs should consider the entire analytic life cycle, from data ingestion to exploration to modelling to deployment. Many organizations struggle with “operationalizing analytics”, with generating value by deploying it in business processes. If a single tool cannot take you through the entire data-analytics process, CXOs should ask how the tools can be stitched together, who will do the stitching and what additional investment is required to operationalize analytics.
What in your opinion is keeping most companies from becoming truly data-driven organizations?
Most companies struggle to become data-driven organizations because they have not been able to operationalize how they manage and analyse their data to produce outcomes and insight. They lack a complete data-analytics strategy that puts in place the right infrastructure and physical assets, the right domain expertise, analytics expertise and human assets—and the right processes and operational assets that integrate the strategy across the organization.
As a result, data and analytics are frequently confined to pockets of the organization and good work being done there is not operationalized for the greater good of the company. Too often, data management is thought of as the domain of the IT department and analytics is thought of as the domain of the business. To become a data-driven organization both sides need to work together and understand the goals, challenges and benefits of a data-driven organization.