Big hype about big data has played a significant role in driving awareness about the value of analytics. SAS welcomes the interest in big data, since it highlights our ability to work with huge volumes of complex and diverse data. Since this is such a critical topic, we have formulated a set of tenets that drive our big data analytics approach. Although big data is often seen as a new topic dominated by discussions involving emerging technologies and specialized use cases, SAS takes a broader, business value perspective.
Scale matters but big data provides promise for organizations of any size - Although the scale may differ, any organization in any industry can run into big data problems. More importantly, any organization in any industry can derive significant business value and differentiation if they can effective exploit big data analytics. Given that big data is defined by scenarios that exceed an organizations ability to manage, and analyze in a timely fashion, it doesn’t matter if the breaking point is hundreds of gigabytes or tens or hundreds of terabytes. The principles that an organization should apply to leverage big data, big data analytics are similar and the SAS approach can provide value in all customer scenarios.
Effective Information Management should leverage analytics – Most data integration and analytic vendors are focused on providing data management capabilities that prepare data for business intelligence, data mining and predictive analytics. Although SAS is recognized by Gartner as a leader in this area dominated by ETL/ELT capabilities, SAS provides a complete information management solution by applying analytics to the data preparation process. This is critical when it comes to big data because it is not always feasible to “store then score” - basically it’s not always possible to put data to disk before analysis is done. In addition to this traditional pattern, SAS supports a “stream, score and store” approach that leverages analytics as part of the data preparation stage. This approach can provide real-time analytics and can also be used to help make decisions about what to do with the data.
Big data is not about building bigger databases - Just as there is no one size fits all definition for big data, there is no one size fits all approach for addressing or leveraging big data. It’s not simply a matter of leveraging new storage techniques that provide the ability to store massive amounts of data. More importantly, it’s about providing a flexible infrastructure and trusted guidance based on best in class technology, pre-packaged business solutions, and years of analytics leadership that will result in a cost effective, low-risk solution that will drive business success and differentiation. Just as SAS provides a graduated approach to data management and analytics, SAS provides a big data, big data analytics approach that not only meets your individual needs but scales based on your business analytics requirements as well as for planned or unforeseen business or technical factors. SAS High Performance Analytics offering can run in both an SMP environment as well as a high performance grid environment. SAS also provides the flexibility to deploy the analytics solution in an on-premise environment, cloud environment (SAS hosted or otherwise) as well as an appliance option.
Big data analytics requires a flexible approach - Depending on your business goal, data landscape and technical requirements, SAS can assess, provide guidance and deliver solutions that deliver optimal analytical results for any big data, big data analytics scenario including:
- Complete data scenarios – SAS can help assess and deliver solutions that allow entire datasets to be properly managed and factored into the analytical processing. SAS has been at the forefront of leveraging in-DB, in-memory, and grid technologies that have reduced batch-based decision management analytics to the point that they can be integrated into real-time operational systems.
- Targeted data scenarios – In situations where it is technically infeasible to leverage the entire data set, or where the amount of data that is included in the analytical processing reaches diminishing returns, SAS provides analytics and data management capabilities to determine the data that needs to be factored into the analytical models to ensure optimal analysis. This capability is based on our ability to properly prioritize, categorize and normalize while also effectively determining relevance and providing enrichment capability at the point of data preparation or ingesting vs. analytical or query processing. This capability can also be leveraged to help manage data retention policies so that organizations can leverage analytics to make decisions about retention vs. using a rudimentary approach that purges away analytical value.
SAS proven approach to move processing to the data provides big data dividends – SAS was one of the first vendors to move data preparation tasks and analytical processing to the actual data source. This eliminates the need to move the data, which is costly in terms of processing and network resources and greatly adds to the latency of the data management and analytical tasks. SAS leverages the power of the MPP capabilities provided by the database vendors, which greatly reduces the processing times and allows organizations to fully leverage their MPP investment. This approach will pay additional dividends as the big data adds to the data volume that must be processed today and in the future.
SAS understands that risk mitigation and cost effectiveness is key – To date, only a limited number of organizations have fully capitalized on the true promise of big data, big data analytics. As organizations begin to understand the true business value and differentiation that can be realized, it will still be necessary for organizations to justify the investment in big data analytics projects. Leading organizations will also do everything within their power to minimize project risk. SAS unrivaled success in big data analytics implementations based on our combination of business solutions, technology, professional services, technical support, etc., virtually guarantees project success in a cost effective manner.
Big data, big data analytics solutions should not be hard – Although emerging technologies have helped drive interest in leveraging big data through big data analytics, the side effect has been that these technologies are highly complex, and that they require specialized skills that are in short supply. In other words, perception equates to high cost and risk. Unfortunately this perception has the potential to spread to other big data, big data analytics technologies and solutions. SAS provides an extensive array of pre-configured business solutions and business analytic solutions greatly simplify the most complex analytical problems, including those that are based on big data.
Hadoop and emerging technologies are part of an overall big data analytics arsenal – Not only has Hadoop and similar technologies enabled organizations to more effectively capture and store massive amounts of data, these technologies have also served to broaden the appeal of analytics to the masses. Hadoop provides an efficient storage mechanism and processing framework for large volumes of data that may not have been captured previously. Hadoop can be used to complement existing data sources and processing approaches.