When Nick Mason of the legendary '70s rock group Pink Floyd showed up for a signing at Virgin Entertainment Group's Megastore on Sunset Strip in Los Angeles, store managers fed the famed drummer updated sales figures for his latest book every 15 minutes by extracting data from the company's business intelligence system. That's just one example of a major change at Virgin made possible by BI. Specifically, Sunset Strip employees can use instantly available BI data to make on-the-spot decisions -- for example, strategically shifting stacks of Mason's books and CDs and quickly placing orders for dwindling supplies.
Local store managers are able to make tactical decisions like that because BI data is streamed continuously to Virgin stores across the country. The data goes far beyond statistics on customer traffic. Included in each store's near-real-time analysis is trending information, including current and historical sales and inventory figures.
Corporate executives at the entertainment giant were looking for more than a mere BI dashboard to serve up analyses on revenue and generate other data points to feed quarterly reports. Instead, they gave Virgin's IT staff the green light to design Crescendo, a system named for its ability to track dynamic waves in sales and identify other operational trends.
Marching orders, however, called for a BI workhorse tool that would provide analytical data that Virgin store managers could use in real situations. For instance, Crescendo often drives the physical placement of hot CDs or T-shirts in local stores. The system also churns out analyses of sales data that store managers use to schedule associates who walk the floors. Sometimes, BI even dictates when these workers should take their lunch breaks.
"We started with distinct store profiles that indicated heavy customer-traffic times so that stores could adjust their schedules accordingly," says Robert Fort, IT director at Los Angeles-based Virgin. "I remember that our store in Vancouver historically and consistently saw a blip around six o'clock each night, as people headed for the train station. Extra sales associates were scheduled to work at this time.
"However, traffic data doesn't mean anything at all on its own," Fort says. The company quickly moved from using BI just to compose snapshots of traffic patterns and began shoveling more trending data continuously to those directly in charge of each store.
"With our current system, we are able to track sales by the hour," says Charlie Brasby, a regional director at Virgin. "We are able to pick up on trends more quickly by evaluating any hours during the day where we fall behind last year's figures. We can also see spikes in traffic that we are not capitalizing on," he says.
Fort began spearheading the move toward increased use of BI in 2004 and was determined from the start to make its use more widespread and tactical. "We wanted a tool that would filter deep into the organization, not just be used by a select few," he says.
The first step was settling on a vendor. After wavering among providers such as Hyperion Solutions and Cognos, Virgin decided to pursue a system built around Microsoft products. Specifically, Virgin uses Microsoft BizTalk business process management servers with SQL servers.
Fort and his team have relied extensively on the analysis capabilities available with SQL services, particularly the use of cubes, or online analytical processing data sets, that provide quick access to Virgin's data warehouses. "Our use of cubes has helped us launch a number of reporting services that let us dive deeper and find information down to the SKU level," he says, referring to stock keeping units, which are identifiers that track products to the particular merchant.
In terms of data, Virgin relies almost exclusively on internal figures. However, it wraps in trending information from Nielsen SoundScan, a system operated by VNU Business Media, a New York-based company that tracks music and music video sales in the U.S. and Canada.
Among other things, this combined information on industry sales trends has bolstered Virgin's efforts to leverage BI in tricky purchasing decisions. For instance, regional sales directors and store managers must forge strategies for stocking "right now" products that are candidates for a sales surge whose size or duration can be hard to gauge. "We can work with our replenishment team to place a second order quickly and really take advantage of weekend business," says Brasby.
Store managers usually run through such a drill once a week. "Tuesday is considered 'new release day' in our market," says Rick Smith, Virgin's business analyst for store systems. "Throughout the day, we are able to easily see the best-sellers and know what is currently on hand in order to be able to make quick decisions on reorders if necessary."
Virgin isn't alone in its scramble to harness BI, notes Dan Vesset, an analyst at IDC. "Most large retailers and many entertainment companies are heavy users of business intelligence," he says. "Some of the most sophisticated users of advanced analytics rely on data mining and statistical analysis of customer behavior and purchase trends."
Virgin counts itself among the industry's more sophisticated users of BI. By better understanding customer behavior, store managers have realized better sales and created a better work environment for employees pounding the floors to sell more merchandise, says Brasby. "BI has not only improved our results but created more of a sales culture throughout our stores," he says, "where our associates enjoy hearing that we had a great hour."