CIOIN

If 1-800-Flowers.com CIO Steve Bozzo had his druthers, even the online retailer's mailroom clerks would have access to business intelligence. "There's valuable information at every level of the organization," he says.

Clearly, Bozzo sees the power of pervasive BI. "Business intelligence needs to be part of the business fabric: Not an afterthought layered on top of a business initiative, but part and parcel of the overall process from the get-go," Bozzo says. "And that's what it is for us—it's a part of our culture."

But pervasive BI doesn't mean everybody in the company has sophisticated analytics tools to use as they wish, cautions Dan Vesset, an analyst at IDC. Rather, he says, pervasive BI is about ensuring that everybody—front-line employees, middle managers and executives—can make decisions using the right information at the right time.

Achieving BI ubiquity takes considerable time and effort—10 years and counting in the case of 1-800-Flowers.com. "Pervasive business intelligence is something that we have and continue to work very hard at—and we think we're really successful at it," Bozzo says.

Over time, the company has learned the imperative of having a BI/analytics practice within IT as well as having corresponding liaison groups in each business unit. "These liaisons are experts in BI, but they major in business and minor in IT, whereas the analytics group in IT majors in IT and minors in business," Bozzo says. "The groups complement each other perfectly, and this has made a huge difference in the way we roll out BI."

At 1-800-Flowers.com, a family of 20 brands, IT asks each business group to identify who needs access to BI and to classify each designated individual as either a basic, intermediate or super user. A basic user can generate basic queries and pull ad hoc reports, while a super user can write macros and generate his own reports; the capabilities of an intermediate user fall in between the two, Bozzo explains.

These are not static designations, he adds. "Our goal is to turn basic users into intermediate users and intermediate users into super users. Ultimately, someone defined by the business as somebody who has a need for BI information will go through that process, with IT taking the responsibility and accountability for facilitating the training," he says. Thirty-two 1-800-Flowers.com employees recently attended a training class run by its BI vendor, SAS Institute, he adds.

"Our basic goal is that we understand everything we can about our customers, so it's important to get increased numbers of people involved in business intelligence. That effort cannot hurt as long as they have the appropriate training and can use the tools that we give them," Bozzo says.

Aberdeen Group has seen the correlation between training and the success of pervasive BI programs, says David White, an analyst at the research firm. "Best-in-class companies on pervasive BI are making sure users understand not only the capabilities of the BI tools, but also the data, statistics if necessary, and analytical techniques, and how these help in decision-making. They have broad educational efforts around pervasive BI," he says.

Democratizing Data
Training is a critical success factor in achieving pervasive BI, which is, in turn, essential for better business excellence, agrees Bobby Nix, director of business intelligence and analytics at Allconnect, an Atlanta-based consumer services company. "We want to be a data-driven company, so we are democratizing data and making sure everybody has access to it," he says.

Toward that end, Nix's team has equipped all corporate employees with SAP BI tools so they can conduct their own day-to-day analysis of how the business is performing. "That means we spend a lot of time training and mentoring them on how to use reports and pull analysis together, as well as on analysis techniques—but we don't create the analysis for them unless it requires really complex analytics," he says.

For this effort to succeed, the BI team has to fully understand the company's business needs, Nix says. "This isn't just me making decisions about what they need. It's an exchange, a gathering of business requirements and a coming to an understanding of how they run their business and what their biggest obstacles are," he says.

First and foremost, companies with a commitment to pervasive BI need to look at the types of decisions being made and determine whether they are strategic, corporate or tactical, Vesset says. "Each type of decision has a different requirement for the type of technology a company needs to apply to support it, and they're different in the way that people interact with the data," he adds.

Businesses can typically handle tactical decisions with rules-based automated systems that kick out exceptions for more in-depth human analysis, for example. Corporate decisions typically entail collaborative BI, so users involved in those decisions will require more than analytics capabilities—they'll also require tools that enable effective communication with colleagues should they need advice on the intelligence, Vesset explains. At the strategic level, where users are making decisions for the longer term, tools are less important than experience.

"One of the primary reasons for BI failure is that IT never really understood why business users needed the information they requested. All it heard was, 'We need this data point,' and that's it," Vesset says.

Everybody from the top down must understand the importance of the data—even individuals who never use an analytics tool or see a business report, agrees John Lucas, director of park operations at the Cincinnati Zoo & Botanical Garden.

"In achieving success, the key has been allowing our project to be steered by the people who are influential in the organization—those who can make budgetary decisions and set strategy and vision, as well as people who are directly responsible for the success or failure of the business, specifically revenue," says Lucas. "But the pervasiveness is core." He spearheaded the decision to foster an enterprisewide culture of BI at the zoo.

The zoo was one of the first visitor attractions to take on such a deep BI project, and Lucas frequently shares his experiences in speaking engagements around the country. "The No. 1 thing I tell people is, if you don't succeed on making everybody understand, embrace and participate in the process, you really shouldn't do analytics," he says. "The cultural buy-in is key to reaching your full potential with analytics."

Lucas says he likes to tell zoo employees, "If you can imagine it, we can measure it," and he has hosted companywide meetings to drive home that message. Even front-line cashiers, who are typically college kids working summer jobs, have to know that the data they gather—patrons' ZIP codes, for example—is critical to the zoo's success. And because the BI team has taken the time to convey that message, he says, "almost literally every day, somebody comes up with an idea on how we can use business analytics to drive the needle for us."

Ongoing efforts to make BI pervasive do pay off, White agrees.

"At the companies we survey, we see pervasive use of business intelligence being tied to better business performance. So those business functions that have access to business intelligence are able to make better-quality and timelier decisions in a way that impacts the business in a positive way," he says.
But you have to remember that making BI pervasive involves constant learning, says 1-800-Flowers.com's Bozzo.

"We're always learning and discovering new tricks," he says. "That's not to say we're not already getting enormous benefit from our BI efforts, because we are—day in and day out. We're just greedy—we want even more."

Send feedback on this feature to editor@cio.in