October 11, 2012
By Kelly Liyakasa,

A Move from Big Data to Big Analytics?

LAS VEGAS—Deciphering big data relies heavily on using advanced or "big" analytics to turn patterns into real-time business decisions. It's not so much about the ever-growing volume as it is real-time processing in a time-sensitive manner. And time-sensitive means minutes or seconds.

This is a prominent theme at the 2012 SAS Premier Business Leadership Series in Las Vegas, where 700-plus executives are gathered this week to discuss the challenges of globalization and business competition, and how companies can use advanced data analytics to maintain an edge and increase their customer base.

One such company, the popular flash sale site Gilt Groupe, credited with digitizing the sample sale, relies on SAS Analytics to gain a deeper understanding of its customers and their propensity to purchase in the future.

"We try to figure out how to be personalized and relevant for all of our different segments," comments Tamara Gruzbarg, senior director of analytics and research at Gilt Groupe. Gilt's divisions include such wide-ranging categories as menswear, home decor, and baby. "[SAS] provides very robust statistical data mining and a very flexible tool for segmentation and to build predictive models….We recently developed a comprehensive customer lifetime value and opt-out risk score model to help identify members who would benefit from decreased email [campaign] frequency."

Like Amazon, Gilt Groupe develops its own technology as well, which Gruzbarg says includes its own internal data warehousing system. When Gilt Groupe mines customer data, it not only looks at what the customer actually purchased, but what other items or brands the customer may purchase, along with preferences.

During a keynote presentation, "The High-Performance Organization," Jim Davis, SAS's senior vice president and chief marketing officer, reiterated that organizations need and rely on speedy responses from analytics engines. But certain challenges, such as the complexity of data usage, an expansion of the user base, demands on data itself, and the need for immediate answers and insights, can complicate things.

Davis said it's "technically impossible" to perform real-time processing using database analytics. Instead, he said, it's necessary to use proactive processes, like statistical analysis, forecasting what goods and services will be needed in the future, and optimization like pricing and markdowns.

During the keynote, Davis and Randy Guard, SAS vice president of product management, touched on several new products that SAS is developing to combat some of the analytics demands on companies today.

Within its High-Performance Analytics architecture is SAS High-Performance Marketing Optimization, a solution Guard said helps marketers analyze things like the number of times a customer has been contacted, or how to maximize the response rate. "It has a business interface designed [to accomplish] business goals." SAS has also increased the breadth of the High-Performance Analytics Server to run on the Hadoop Distributed File System; a new release is expected in the first half of the new year.

Also announced Wednesday was an enterprise SAS Financial Crimes Suite, which helps companies detect fraud early in its stages and to minimize the monetary impact of such an instance. Guard said the system allows users to run "What if?" scenarios and utilize case management tools to see an incident through litigation.

SAS also debuted a new solution, SAS Revenue Management and Price Optimization Analytics, geared specifically for hospitality and entertainment organizations to be able to forecast demand, and come up with an optimal price for a certain customer set, factoring in considerations like competitor pricing.