KXEN’s Analytic Framework Allows Online Retailer to Build and Improve Predictive Analytical Models More Quickly
SAN FRANCISCO, June 13, 2006 - The highly competitive world of online retailing is always looking for ways to increase its customer base by boosting the conversion of visitors to customers and the amount spent per customer. Overstock.com® understands this business challenge and recently began using KXEN Analytic Framework to help provide additional “lift” or conversion of visitors to customers and increase the dollar amount per order.
“We’ve seen lift in some of the models we’ve run,” said Scott Burk, Ph.D., Senior Statistician and Technical Lead for Marketing Analytics at Overstock.com. “What’s different about KXEN’s Analytic Framework from other predictive tool sets is we can quickly build models and work on improving the model versus spending time and energy on the construction of each model.”
One of Overstock.com’s strategies this year is to do more analytical modeling. KXEN’s Analytic Framework builds models more rapidly; thereby allowing Overstock.com to run many different models and deliver on the strategy. For instance, when developing marketing campaigns, Overstock.com can use KXEN to build models that maximize the performance of each campaign.
“We knew we’d be doing more and more modeling,” said Burk, “That’s the nature of our business. Our models will go stale quickly because user behavior changes quickly. With KXEN we can focus on augmentation of the data set to build a better model. With other tools, most of our time is spent building the model, leaving less time to focus on the data that we have and on what additional data we can bring in to improve the model.” Burk estimates his team is producing four to six times the number of models using KXEN versus the previous predictive analytic toolsets used. Burk foresees using KXEN for predictive analytic work in a variety of applications within Overstock.
KXEN Analytic Framework 3.3
KXEN recently announced the release of its newest version: KXEN Analytic Framework 3.3. New features include:
Among many new features, KXEN has included new functionality that further increases modeling automation. For instance, version 3.3 provides automated script generation from the graphical user interface. This means a model can be created interactively, and with the click of a button a script can be produced for integration into an automated production environment to repeat the model creation and deployment process.
Version 3.3 provides greater productivity for analysts through features like flexible scoring output, an enhanced reporting system, and the option to automatically refine a model based on production and business constraints. The custom aggregation components have also been improved for greater flexibility and scalability during deployment. These features reduce the amount of time and effort spent to validate, refine, and deploy models.
Overstock.com, Inc. is an online "closeout" retailer, offering discount, brand-name merchandise for sale over the Internet. The company offers its customers an opportunity to shop for bargains conveniently, while offering its suppliers an alternative inventory liquidation distribution channel. Overstock.com, headquartered in Salt Lake City, is a publicly traded company listed on the NASDAQ National Market System and can be found online at http://www.overstock.com.
KXEN provides next generation business analytics software to drive better corporate decisions. KXEN’s unmatched speed, ease of use and scalability enable leading companies around the world to expand the use of predictive analytics and enhance corporate performance. Based on breakthrough mathematical theory, KXEN’s products offer reliable predictions and deep insight for achieving critical business goals. The company partners with leading systems integrators and software vendors to integrate advanced analytics into enterprise applications and business processes. Founded in 1998, KXEN is headquartered in San Francisco, California, with offices in the USA, UK, and France, and distributors throughout the world. Visit the KXEN Web site at www.kxen.com.
WRITE IMAGE for KXEN
Tel: +1 425-636-0270
WRITE IMAGE for KXEN
Tel: +1 425-636-0296