Using a software suite from Cognos Inc., Peet's Coffee & Tea Inc. found a way to move beyond the basics of business...
intelligence (BI). Today the company is creating highly granular analyses of its retail organization in more than 100 West Coast outlets.
The new BI solutions can measure the relationship between labor, productivity, and costs at various locations.
"We've come up with metrics for each of our individual stores to see where the 'outliers' are and then take action," said Michael Cloutier, CIO of the Emeryville, Calif.-headquartered company. Outliers, he explained, are employees that are not as productive as their peers.
"If a store is using more hours than somebody else who has a similar sales volume...than we would investigate what is causing that problem," Cloutier said.
Building the right BI systems
Analysts and experts point out that most companies today have only basic BI capabilities in place. That may include using software to produce simple and static inventory reports and analyses. But many CIOs are figuring out how to move BI to a more strategic level.
Experts say that an increasing number of companies today are beginning to wrestle with the question of how to grow BI into an application that puts out real-time numbers and helps them succeed in their goals of streamlining the supply chain, lowering inventory costs, raising profits -- and generally becoming a better business partner or supplier.
"There are very simple kinds of basic reporting and analytics, and then there is high-end BI that includes strategic analytics, dashboards and alerts. It's a more real-time kind of business intelligence that drives new opportunities and really high end business processes," said Joshua Greenbaum, principal consultant at Enterprise Applications Consulting of Daly City, Calif. "Not enough companies are doing that version of BI."
Greenbaum explained that the major obstacle to the adoption of strategic, enterprise-wide BI initiatives has more to do with ignorance than with price or implementation concerns.
"The standard modus operandi [in corporate America today] is to proceed without comprehensive views," Greenbaum said. "The other obstacle is that there is a diffusion of capability across the enterprise that makes it hard to synchronize and come up with accurate business intelligence metrics."
There is no shortage of high-end BI software available. The major players in the market include Cognos, Business Objects SA, Informatica Corp., and Ascential Software Corp. But high-end software is basically useless if a company isn't absolutely sure of the problems it wants to solve through BI technology.
"Companies have to start by identifying key business processes that need to be closely monitored -- and identify where more information and more analysis can actually make a strategic difference," Greenbaum said.
Keith Gile, a principal analyst at Cambridge, Mass.-based Forrester Research, agrees that smart BI implementations begin with a thorough examination of business processes.
"Consider more advanced business intelligence technologies that can be embedded within your operational applications," Gile said. "Look for ways to improve processes, whether it's the definition, the modeling, the automation of these processes, or the general decision making that goes along with these processes."
Gile said that many of Forrester's clients are trying to get more out their enterprise applications by using BI in more places. He said those clients have found that many BI vendors today are weak in their understanding of business processes and that potential buyers should beware.
"This is changing [however] and I would expect that we'll be seeing more activity in this area over the next 12 months," Gile said.
More advice for BI investors
Greenbaum suggests that companies, especially smaller ones, consider outsourcing to a third party company that offers hosted BI. If that isn't feasible, consider implementing an in-house training program.
"Very often, the people who are doing BI have a tremendous amount of knowledge about procurement, management, about manufacturing, about a lot of things, but they may not be very strong on complex data analysis," Greenbaum said.
Companies should be sure to allow room for initial mistakes and changes that inevitably arise during the process, Cloutier said.
"As you start to delve into BI and develop your [reports,] you're going to learn that you need more information in some areas and less in others," he said.
Allow more time than you think you're going to need to develop the tool and deploy it," he said. "Also, make absolutely certain that you've got someone from your customer community who owns the project so that they've got some skin in the game."