Putting Data to Work in the Real World

By 2020, Volvo (IW 1000/100) is shooting for a perfect record: zero accidents, zero deaths and zero injuries in any of its cars. To help it accomplish that modest goal, the company has turned to data. 

Every time a car pulls into a Volvo dealership for any reason, a technician downloads all of its sensors and all of the information it has been accumulating since its last service. That information is then pinged to the Volvo headquarters and added to its expansive data warehouse containing all of the diagnostic data from all of its cars for the past six years -- adding up to somewhere over 1.7 terabytes of data to date. With that data, the company can spot a serious problem, be it a manufacturing or driver-based issue, much sooner, thus reducing the detect-and-correct timeframe for defective, problem or damaged components from as much as eight months to as little as three weeks. 

Western Digital (IW 500/109) is obsessed with quality. To serve that obsession, the company has transformed its manufacturing process to allow scanning, recoding, testing and tracking of every disc drive it produces while they are still on the line. Running analytics in real time on the floor, the company can now locate and pull bad discs long before they reach the customer. Even if a disc passes initial review, if further analysis reveals a problem, it can be located and pulled even from inventory bins. This has resulted in a 1.9% defect per million rate and the lowest warranty return rate in the industry in 2010.

Along with its general transition toward a more digitally oriented structure, Dell sought to increase the efficiency of its online ordering system. To do so, Dell's business intelligence team ran analytics on its clickstream data, tracing every move, every click, every path taken by every visitor to its site. Digging into those results showed that the configure-to-order system required users to navigate through a staggering 30 to 40 clicks to place an order -- not exactly a lean setup. The team then used that information to help optimize the site and reduce navigation down to just five clicks. That project yielded a positive revenue impact of about $35 million.