Business intelligence goes small: It's not just for the biggest shops anymore

IT pros in midsize businesses share their challenges and triumphs.

1 2 3 4 5 Page 2
Page 2 of 5

He also was aware that other critical components would most likely be out of sync. "Traditional BI software assumes the ETL [extract, transform and load], data warehousing and cubing stages all will occur sequentially and without glitches," Grosz says, something he could not guarantee at the 900-employee company.

Dan Grosz
Dan Grosz, VP of information systems at VIP Parts, Tires and Service, was attracted to BI for its ability to analyze data across sales, inventory, financial and payroll systems.

Grosz chose QlikTech's QlikView enterprise BI software because it supports an "iterative, incremental development cycle." He was able to connect the software to his company's point-of-sale, supply chain and inventory systems first. That allowed executives, business analysts and store operations managers to drill down into the merchandising hierarchy and extract detailed reports about how products were performing.

Once users were comfortable, they started requesting access to other reports such as the gross margin return on inventory at the SKU level. Those requests led to the tie-in of finance and payroll systems to BI.

Had Grosz tried to implement BI in a more conventional manner using the waterfall approach, the project "might never have been as successful as it has been," he says.

A hybrid approach

For Cubist's Murabito, flexibility has come from choosing multiple BI tools. Almost since the time of its founding, Cubist, which develops a drug called Cubicin that's used to treat blood and skin infections, has considered BI critical to its success. In 2003, IT deployed IBM Cognos software for on-demand reporting, to develop metrics for key performance indicators (KPI) and to help identify sales leads and understand the supply chain.

In 2008, reports from the Cognos system showed that the company's main distribution hub in Louisville, Ky., had consistently experienced weather delays over the years due to ice and snow storms. That was unacceptable for an acute care product that had to be shipped immediately during outbreaks of certain kinds of infections. "Based on this data, we opened a second site in Reno, Nev., to keep the supply chain full and flowing," Murabito says.

In addition to generating reports and KPIs, Murabito wanted to offer deeper analytics for a wide base of users at the 700-employee company. He felt Tibco's Spotfire analytics software, a stand-alone package that's not part of Cognos, was a good match in terms of its data query and visualization capabilities. He liked its features and felt it was a good fit. Moreover, he liked the idea of using two different products so Cubist wouldn't be locked in with any one BI vendor.

During the aftermath of Hurricane Katrina, Spotfire detected that demand for Cubicin was shifting from New Orleans to medical centers in Houston and other areas as Gulf Coast residents evacuated and relocated. Knowing this in real time made it possible to quickly reroute shipments to the appropriate facilities.

Both Spotfire and Cognos draw data from Cubist's data warehouse. It's important to Murabito that the reporting and analytics remain independent from one another so he can switch vendors and tools as needed. "A hybrid approach leaves you open" for flexibility in terms of vendor selection, he says.

Ready to scale

For one thing, Murabito is concerned that if Cubist expands its workforce much more, his software licensing fees will skyrocket and he'll have to find another product. As more and more employees start using BI, "that is driving up licensing costs," he says.

Leo Collins, vice president and CIO at Lionsgate, says he feels prepared for growth because he has already deployed SAP BusinessObjects' software as a large-scale implementation for all users in the Santa Monica, Calif.-based entertainment company, not just those in one department. His only regret: He focused more on integration and tools for strong analytics users instead of considering the end-user experience for all 500 employees. "It's clear [now] that easy-to-use end-user tools are a significant success factor for us," he says.

1 2 3 4 5 Page 2
Page 2 of 5
It’s time to break the ChatGPT habit
Shop Tech Products at Amazon