Companies of any size and from any industry today have one thing atop their business agendas: business intelligence.
MORE ON CIO.com
Part 1: A Technology Category in Tumult
Part 2: Nine BI Vendors to Watch
Part 3: BI and On-Demand: The Perfect Marriage?
Opinion: Don’t Make BI Suck for Users
Two pieces of research from Aberdeen Group illustrate business’s growing and insatiable need for actionable reporting and analytics data. A 2008 survey of 4,300 companies found that the number-one technology that could have the greatest impact on businesses during the next two to five years was BI and analytics. That’s not surprising because Aberdeen’s 2007 research showed that the number-one technology spending item for companies was “reporting and analytics.”
IT departments, in particular, are feeling the heat to give line-of-business executives and their demanding end users exactly what they want and need—even if it’s clear that IT doesn’t have the necessary application development and integration abilities.
And in the rush to achieve BI bliss, many critical questions surrounding the ultimate costs (as in, “What’s our total cost of ownership going to be with this BI system?”) are left unanswered by many companies at the outset. Perhaps, it’s for a good reason. “That’s a difficult thing to determine,” says David Hatch, a BI research director at Aberdeen Group. “There is no standard for a BI scenario.”
Hatch’s research shows that those companies that have recently rolled out a BI system notice financial pressures coming from several areas. The first, and most important, Hatch says, is the pressure to improve data integration from multiple applications (noted by 42 percent of the respondents). From interviews with those who took part in the survey, Hatch notes that “this single factor alone can dictate the success or failure of a BI initiative.”
“What I’m finding is that companies don’t really understand the extent of the data management challenges to BI until they get into it,” Hatch says. And if internal IT does not have the skill sets or resources to solve these data challenges, that can become pricey for many companies. “You have to invest internally or look outside for help,” Hatch says, “and both are expensive options.”
IT departments also report feeling BI implementation pressures related to the need to: deliver BI to more end users (29 percent); improve ease-of-use of BI for nontechnical users (27 percent); and speed the development of customized BI capabilities (19 percent), according to Aberdeen’s survey.
Unresolved and combined, those three problems can lead to substantial decreases in user adoption—a pervasive problem with BI applications, and one that makes determining TCO a “cloudy exercise,” Hatch notes.
Even if user adoption is high and BI projects swell in scope, any lingering issues like those revealed in the Aberdeen survey will become only more intensified, Hatch warns. “Most companies start small with this,” Hatch says. “And it either grows, or it doesn’t. But once it gets to the enterprise-size level, some of these issues can get ugly.”
David O’Connell, a senior analyst at Nucleus Research who covers BI, says that while overall cost is important, he sees a different type of problem right now. “I think [companies] are spending too much time on TCO and not enough on identifying and estimating attainable benefits,” O’Connell says. “If you strive to minimize costs, you’ll minimize functionality, adoption, benefits and ultimately ROI.”
While O’Connell concedes that not all BI vendors’ pricing models are “always flexible,” he does note that some vendors, such as Cognos (now owned by IBM), have pricing structures that offer a “price point that can accommodate just about any project size.” In addition, the recent growth in on-demand and SaaS BI options, which offer quicker implementation times and lower costs to get started, is another route for companies to go.
“Flexibility like this is important,” he says. “Without it, potentially high-ROI projects that should be pursued die on the white board during the cost analysis.”
That’s the real hidden cost of BI implementations, O’Connell says: “Non adoption.”
Secrets of the BI Masters
In the course of analyzing the Aberdeen survey data, Hatch identified the top three strategies in which “best in class” companies managed the TCO of BI. (Aberdeen’s best-in-class performers scored the highest compared with other surveyed companies in these three categories: time-to-completion of BI projects; on-budget completion of BI projects; and cost-per-user of BI applications.)
These companies understood what the end-user requirements were for the BI applications. They were able to identify data sources for BI applications. And they defined business rules and calculations required for reports and analytic views.
Not surprisingly, understanding what the end users want is a critical factor, Hatch notes. “This highlights the importance of planning BI implementations based on user requirements as a primary cost-management factor,” he writes in the survey report. “If the end-user requirements are not well-understood up front, a lot of time and effort ends up being wasted.”
O’Connell suggests that companies pursue BI projects that they know will result in direct benefits. “By this, I mean tactical benefits such as improved productivity, tightening of the value chain or sales improvements,” he says. “Use these benefits to pay for the larger more indirect benefits such as improved visibility and better ability to detect and respond to changes in the business environment.”
(For more on CIO‘s special BI series, see Part 1 “BI: A Technology Category in Tumult“; Part 2 “Nine BI Vendors to Watch“; Part 3 “BI and On-Demand: The Perfect Marriage?“; and Part 5 “Opinion: Don’t Make BI Suck for Users.”)