Seven secrets to business intelligence (BI) success

At the fall 2009 TDWI World Conference, keynote speaker Cindi Howson shared her seven secrets to business intelligence success. Read these BI tips to find out whether your company is doing all it can to get the most out of its BI system.

Not much has changed over the past few years with regard to how successful companies rank their business intelligence (BI) projects, according to Cindi Howson, founder of the technology evaluation website BIScorecard.

In 2007, 8% of companies surveyed said that their BI initiatives were failures, while 68% said that the projects were slightly or moderately successful and 24% reported that their BI efforts had been very successful, Howson said during a keynote speech last week at the fall 2009 TDWI World Conference in Orlando. Two years later, those numbers have stayed well within the survey's margin of error, according to Howson. This year, 3% of the respondents declared their BI projects to be failures, 73% said they were slightly or moderately successful and 21% called their initiatives very successful.

But there are ways for organizations to increase their chances of achieving BI success, Howson said. Here are her top seven secrets to successful BI deployments:

For more on BI success
Get advice on whether you should use your ERP vendor for BI or go with third-party software

Read some best practices for using BI in financial services
Measure success. For a project to be successful, it's important to first define what success would mean for the project, she said. Will it provide a business contribution (increased revenue, better processing efficiency, improved customer satisfaction, etc.), faster access to data, a quantifiable return on investment (ROI), a higher number of active users, or something else?

Then you should track how well your goals for BI success are met. "Don't think of BI as an ad hoc tool, but as a way to help people do their jobs better in the context they need it," Howson said.

Consider the "LOFT effect." BI initiatives can be boiled down to four different motivational factors, collectively referred to as LOFT, she said. That stands for luck (i.e., a change of leadership), opportunity (such as new business opportunities), frustration (for example, a company "flying blind" or dealing with data chaos), and threat (competition, the threat of bankruptcy, etc.).

Howson recommended that IT professionals look at the impact the so-called LOFT effect can have on their organizations, especially the opportunity and threat components.

"A certain level of pain can force you to work smarter," she said. "Where are your opportunities? Where are your pain points?" Answering those questions can help lead BI projects in the right direction, she added.

Howson cited the example of Emergency Medical Associates (EMA), a Livingston, N.J.-based provider of emergency-room services to hospitals, which had a lot of data at its disposal and saw an opportunity to leverage the information for better service. Doing so has led to improved patient care and reduced emergency room waiting times, she said. And because of its BI system, EMA can report swine flu cases on a daily basis while other organizations can do so only weekly.

Get executive support. Howson admitted that the idea of "getting executive support" has become almost a cliché among consultants nowadays, but that doesn't mean it's any less applicable than in the past. And, she said, it's particularly apt for BI, which has become more important than ever because of the economic recession that has financially crippled many organizations around the globe. "The downturn has forced people to work smarter," she said.

But which executives should IT concentrate on for getting support? According to Howson's research, CEOs have the biggest impact on getting new BI projects up and running, while CIOs and other IT managers have the least impact on that.

So if the IT department needs support from upper management, excluding its own CIO, how does it attract that support? Howson said IT needs to demonstrate continuous success on other projects and show the "hard business benefits" of a BI deployment, such as how a new tool or initiative could make a positive impact on the company. Of course, a little luck helps, too.

Ensure a solid data foundation. Sometimes, it seems that businesses are in a Catch 22 situation when it comes to their data and BI tools. BI applications are only as good as the underlying data fed into them, Howson said -- and yet data is nothing without powerful BI tools to put the information in context for users.

To achieve BI success, companies need to recognize that problems typically start at the source because of factors such as disparate systems or a lack of data ownership. Howson said that master data management (MDM) software and other technologies can help alleviate data problems, as can the concept of "right-time data", which involves balancing the cost of timely updates against an assessment of where the delivery of real-time data can make the biggest impact in your organization.

Foster an IT-business partnership. Howson said that while some people might compare the relationship between IT and business users with that of Rocky Balboa and Apollo Creed, the differences in working styles and personalities between the two groups are much more Yin-Yang in their nature. IT professionals tend to be more introverted and risk-assessing, and they like to document everything, she said, whereas business users often are more outgoing and more carefree with regard to risk and don't understand why new tools and software need to be tested for several months before being rolled out.

Communication is key to bridging this gap between IT and the business, Howson said, adding that IT should stay away from "techno-babble" and emphasize that life would be better with the new BI application than without it.

Promote relevance. "Instead of waiting for the business to come to you, you need to be out there and studying what drives the business and deliver ways [that] BI and data can help them," she said.

She gave several examples, such as flight attendants and gate agents. They aren't going to go to IT and ask for data, she said, but some airlines are providing their frontline employees with data on customers so they can make decisions such as who should be upgraded to first-class. IT needs to provide "data in the context that [business users] need it," she said.

Choose appropriate tools. A BI tool that might be perfect for a business employee in the trenches might not be the right one for a high-ranking executive, Howson cautioned. The key, she said, is "matching the right BI tool capabilities to the right user segment."

Dig Deeper on Business intelligence best practices