Page 1 of 2
By Tom Groenfeldt
In a recent special report on information management, the Economist provided some stunning statistics on data-generation leaders. Wal-Mart, for example, does 1 million transactions per hour and feeds a database of more than 2.5 petabytes, which the author helpfully points out is 167 times the number of books in the Library of Congress.
So what? Well, companies that tackle huge data stores can have a high ROI.
According to the Economist, Wal-Mart found that ahead of hurricanes people buy batteries and flashlights -- pretty predictable -- but also Pop-Tarts, which apparently are good for eating when the power goes out.
A Swiss mobile phone company cut annual churn from 20 percent to 5 percent after seeing dissatisfaction reflected in service calls rises in the ninth month, while cancellation doesn’t happen until the thirteenth. Now it offers special deals after seven months.
Nestle saved $30 million a year in purchasing vanilla -- and operational improvements worth over $1 billion -- by using data in its SAP system.
Nearly two-third of Netflix’s sales are based on computer-generated recommendations.
No wonder business intelligence is growing nearly twice as fast as other software areas.
The biggest non-scientific generator of data comes on the Internet, especially on consumer Web sites. “You can’t imagine how much you can catch and analyze to understand true customer behavior in the Web channel,” said Bill Franks, Teradata’s chief analytics officer for the company’s alliance with SAS. The two companies formed a partnership two years ago to combine SAS expertise with Teradata’s skills in data warehousing.
Companies with sophisticated Web sites see not just what a customer purchases, but also what they looked at and how long they looked at it. They can also develop a pretty good idea of how they made their decision. Did they take an item out of the shopping cart after reading customer reviews? Or did they did they put it in their cart after reading the short description and then remove it after reading the full product details?
Crude data capture would aggregate points of customer action, such as purchasing, and keep it in a silo. Now Teradata has customers who are capturing each move of a customer on a Web site and maintaining the details so business analysts can go back and run queries based on an individual’s actions, or the company can use the data to generate offers in real time.
“Technology has been advancing rapidly, so our sweet spot in the SAS partnership is capturing large volumes of detail which even five years ago we might have struggled to handle,” said Franks. With individual records from the Web site, stores can combine that with call center data and in-store purchases linked to loyalty cards.
“We are capturing the individual customer behavior record across time,” added Retha Keyser, SAS business development manager for customer intelligence practice. “We use the behavior to analyze the key drivers to understand what is most likely to be the future behavior of this particular client so we can do more relevant offers in real time.”