• Votes for this article no votes for this yet
  • Dashboard Insight Newsletter Sign Up

One-on-One with Anne H. Milley
Senior Director, Technology Product Marketing, SAS

by Karly Gaffney, Media Relations, Dashboard InsightWednesday, May 27, 2009

Dashboard Insight recently spoke with Anne Milley about business analytics software, the SAS and Teradata partnership - plus a few thoughts on why SMBs choose SAS vs. lower-priced solutions.


 

Dashboard Insight: Tell us about the history of SAS.

Anne Milley: SAS once stood for "statistical analysis system" and was created in the early 1970s by Jim Goodnight, John Sall and other North Carolina State University colleagues to analyze agricultural-research data.  When demand mushroomed, SAS Institute was founded in 1976 to develop and sell the software.  After evolving into the world's leading provider of software and services for business analytics, the company dropped "Institute" from its name and became SAS, which is no longer an acronym.

DI: Where does SAS sit on the BI stack?

AM: SAS addresses the strategic and ongoing “intelligence” part of the business intelligence stack through analytics.

The term “business intelligence” has been stretched and widened to encapsulate a lot of different techniques, tools and technologies since it was first coined decades ago.  Essentially, business intelligence software has always been about information delivery, be it in static rows and columns, graphical representations of information, or the modern and hyper-interactive dashboard with dials and widgets.   Business intelligence technologies have also evolved to include intuitive ad hoc query and analysis with the ability to drill down into the details within context.  All of these capabilities are great for reacting to business problems as they occur.

Traditional and rapidly commoditizing query-and-reporting BI can’t deliver the full intelligence needed for businesses to effectively compete.  Businesses require robust data integration, data quality, exploratory data analysis, predictive analytics, text analytics, forecasting, experimental design and optimization technologies to best anticipate and plan for the future, avoid undesired outcomes and course-correct.

DI: With the current economic environment being the way it is, how does SAS business analytics software assist companies with identifying opportunities and positioning themselves for the eventual economic recovery?

AM: Analytics is the art and science of better – better processes, better decisions and better performance.  To survive today and thrive tomorrow, leading organizations are relying on analytical software and services from SAS.  With the right analytical approach, decision makers can understand not only what’s already happened but, “Why is this happening now?  What if these trends continue?  What will happen next?  What’s the best that can happen?”

DI: Will you elaborate on the SAS and Teradata partnership and what it means to your customers?

AM: SAS and Teradata entered into a strategic partnership in October 2007.  Substantial investment and staffing resources have been committed to joint product integration, development and marketing of packaged product offerings, as well as joint selling and consulting services.

Existing customers can maximize the utilization of their investments in SAS and Teradata through joint product integration efforts and new products and services.  For example, the Optimization Services Advantage Program can help joint customers to implement effective analytic data architecture and improve model development and deployment steps/processes.  This is accomplished by recommending techniques and best practices to better utilize the strengths of both environments, while eliminating unnecessary steps that would drive up the costs of deploying analytics.  This reduces the time to results. 

Organizations are always looking to maximize value from data stored in their enterprise data warehouse using data integration, reporting and analytic applications.  If they use their database and analytic technologies in a competing fashion, it can require extensive data movement, causing data replication and data latency issues.  These inefficiencies affect the model development and deployment cycles, which ultimately reduces their ability to leverage analytics across lines of business.  New products and services from SAS and Teradata will help organizations increase productivity, improve performance and make decisions confidently.  For example, the SAS and Teradata Analytic Advantage Program offers a comprehensive set of software, hardware and services to accelerate model development, deployment and management leveraging our in-database capabilities.  This integrated set of offerings meets diverse and growing analytic and data warehousing requirements.  The result?  Both IT and analysts are more productive and responsive to business needs.  Better analytic models ultimately help line-of-business managers make better business decisions to ensure more effective business outcomes.

DI: Tell us about SAS Analytics and its components.

AM: SAS provides an extensive and extendable analytic workbench that is unmatched in the industry – offering all of the necessary elements to help measure, define, model, deploy, learn and improve on a continuous basis. 

Many people think analytics starts with data, but it really starts with problem formulation.  What problem needs to be solved and what opportunities could be revealed?  Just because data has been generated and stored, it doesn’t mean that value follows “automagically” from blindly mining that data.  As Deming proposed, the object of taking data is to provide a basis for action.  First, organizations must formulate a question, a hypothesis, an idea or direction to guide in the use of the data. 

This isn’t just about opportunistically analyzing transactional data, but also about considering what an organization needs to measure – often through experiments – and evolving the data-collection strategy over time.  Thinking about what data is available – and what data is still needed – will help ensure that organizations are measuring what matters, that they have a plan to derive value from their data, and that insights from the data can continuously refine their data-collection strategy. SAS provides the ability to access virtually any data source – DBMS’s (natively), PC file formats, text, etc.  Accessing and leveraging data, regardless of source, supports mission-critical business decisions.

The data may be small in scope or large and complex.  Regardless of how much data and how many sources of data are involved, the data still has to be fit to provide value as the data progresses through the life cycle.  This often requires assembling and cleansing the data, providing standard and ad hoc reports and further preparing and analyzing the data, deploying results, managing and monitoring models.  Data integration, data quality and data-manipulation capabilities are required throughout the information-management life cycle.  There is an important distinction, though, when it comes to data preparation for common use as in a data warehouse – which often serves standard and ad hoc reporting needs, versus analysis needs.  Even if the data warehouse is populated with relatively clean data, data manipulation and transformation capabilities are still required on top of this for successful analysis.

Exploratory abilities are very important in the initial phase of any analysis.  Exploratory data analysis (EDA) is key in helping the analyst quickly identify anomalies, relationships, patterns and trends to further refine the best approach.  Visualization greatly facilitates EDA and helps you identify important attributes of your data faster, zeroing in on the issues critical to the organization, to reveal logical next steps gaining valuable insight along the way.  EDA sets the stage for more effective model-building for things like customer profitability, network utilization, forecasting customer demand, capacity planning, etc.  Model-building encompasses many flavors of analytics including predictive models, descriptive models, time series models/forecasts, econometric models, optimization models, simulation models and more – often these are used in combination depending on the complexity of the problem.  Capabilities to manage models and design experiments are also critical elements in a complete analytic workbench.  Closed-loop learning is essential for efficient and effective learning to continuously improve.

Tweet article    Stumble article    Digg article    Buzz article    Delicious bookmark      Dashboard Insight RSS Feed
 
 Next Page
1 2
Other articles by this author

Discussion:

No comments have been posted yet.

Site Map | Contribute | Privacy Policy | Contact Us | Dashboard Insight © 2017