In-Depth

CASE STUDY: BRIGGS & STRATTON MOWS DOWN DATA WITH VISUALIZATION

• By Kathleen Ohlson
• August 1, 2005

Decades of data has always been close at hand for Briggs & Stratton employees, but at one time it didn’t seem close enough. “It was a general consensus that there was more to be understood than what could be understood in rows and columns,” says Grant Felsing, Briggs’ decision support manager. Business intelligence’s importance is growing, he says, “I knew how this [was] changing over time, and the only way to understand it is if you see it visually, not through numbers.”

Briggs & Stratton manufactures gasoline engines for lawn equipment, pressure washers and generators for OEMs such as Campbell Hausfeld, John Deere, Craftsman and Toro. The company creates daily and monthly standard reporting for users, builds queries for specific parameters and grazes through this information to create reports. At one point, Briggs created more than 4,000 annual reports.

Briggs built a data warehouse and then moved from a legacy mainframe environment to Unix-based servers. An SAP R/3 ERP system was also incorporated into the infrastructure. R/3 changed the environment and removed the bulk of its legacy systems, essentially wiping out Briggs’ entire reporting infrastructure. The legacy systems housed most of the sources for its advanced data warehouse and accumulated specific programming layers for Briggs. To accommodate the change, Briggs added SAS Institute software for business reporting capability on top of R/3.

SAS Enterprise BI Server, which includes Information Delivery Portal, Web Report Studio and Information Map Studio, takes data from multiple sources, such as production and inventory control, and creates approximately 1,000 customizable reports. Data visualization software allows Briggs to watch changes over time, monitor assets, inventory levels, regional sales and other important areas, Felsing says. The company can resolve issues and understand why certain characteristics appear.

Operational managers can view data by customer, product line or time period across Briggs’ major product lines. Manufacturing supervisors use the information to optimize production levels based on customer inventory and demand, while supply chain managers compare inventory levels with sales and shipment levels to see whether all areas are up to snuff.

Briggs’ IT staff also developed a system for tracking historical information and monitoring quality trends, facilities operations and failure rates for its engines. This quality improvement application automatically flags potential issues, notifying managers and executives via e-mail when problems emerge. According to Felsing, Briggs has saved $4 million per year in warranty payouts through assets analytics and data visualization software.

Claims for engines and other parts come in and “we might not know about it for months,” he says. “We could be in months of production with glitches. We can take corrective action 4 to 5 months [sooner].” Early-warning alerts help managers address potential concerns before they affect customers. For example, the manufacturing team was able to identify and fix a million-dollar quality issue after implementing Enterprise BI Server. The problem was identified in the early stage of production, saving Briggs more than $1 million. Without Enterprise BI Server, managers wouldn’t have identified the issue for at least 4 more months.

“[Problem] rankings occur over time; at 180, you may totally ignore it, but you may notice it move from 140 to 110,” Felsing says. “But over three months, it moved from 180 to 110, you’re looking at it graphically, and it’s screaming, ‘I’m coming.’ You can pick it off that much earlier, or it continues to grow [and then you notice it] when it breaks the financial barrier.” Fewer complaints and warranty claims were important results, and addressing these issues led further credibility to data visualization.

“Some of the stuff is defined, so you can’t get blindsided,” he says. “You can find it dramatically sooner, weeks or months earlier, so you have an opportunity to react to an issue.”

Back to Feature: A View to the Thrill of Data