Midmarket business intelligence problems and pitfalls to avoid

There are lots of ways that BI deployments in small and midsize organizations can go wrong, according to analysts who offer advice on how to sidestep the roadblocks to a successful deployment.

This Content Component encountered an error

If companies of all stripes aren’t careful, business intelligence (BI) projects can turn into cases of death by a thousand cuts: There are plenty of ways to go wrong, some large and some small. But midmarket organizations face their own set of business intelligence problems, challenges and pitfalls, according to analysts who follow the BI market.

For instance, too many small and medium-sized businesses (SMBs) try to use BI software as a giant Band-Aid to “fix” disparate systems, said Jeanne Johnson, global head of KPMG LLP’s BI consulting group. “Most people assume that you can put something on top of your existing systems and make sense of all the variability in the source systems underneath,” Johnson said.

She added that some companies try to “ETL through the process” -- using extract, transform and load tools to translate and reconcile data in order to get the BI output that they want. But such approaches are “a slippery slope” that can end up being nothing more than exercises in building interfaces, Johnson cautioned.

Instead, she advised midmarket companies to look holistically at their IT architecture to see if they’re doing the right things in terms of capturing, consolidating and distributing data. Sometimes, she said, “IT renovations” are needed before a BI system can be successfully deployed.

Possible cause of BI problems: Striving for perfection
Another common problem, Johnson said, is SMBs trying to “perfect their requirements” to the point where they have a seemingly airtight BI project plan. That sounds like a great idea -- but in practice, both BI tools and business needs are likely to evolve faster than such plans envision, according to Johnson.

“There is a huge value in working with a few metrics at a foundational level and then building additional capabilities as you develop,” via an iterative process, she said. Doing so can help ensure that business decision makers and other workers actually use a BI system, although Johnson acknowledged that there tends to be “a little bit of an art to that process.”

John Lucker, a principal at Deloitte Consulting LLC who leads the firm’s advanced analytics and modeling practice, said one of the common BI problems he sees involves setting expectations as part of a BI business case and then failing to delivering on them. Even the question of whether to pursue a BI implementation needs to be examined honestly, Lucker advised. “People read articles that say they have to have BI, but companies, especially at the midsize range, may not really have the resources,” he said.

For organizations that do decide to pursue business intelligence strategies, Lucker said it’s important to articulate a clear vision and develop a BI roadmap that includes a mix of short-, medium- and long-term deliverables; otherwise, the BI effort could lose focus and internal support.

Three-year plan could end in business intelligence problems
Perhaps just as important, he added, is the need to achieve some initial successes. “Don’t tell people that in three years it will all be worthwhile,” he said. “You want to be able to show something that will begin to throw off benefits in three months.” Such accomplishments can then be used as a down payment of sorts on getting approval for additional BI capabilities, according to Lucker.

Not having a good handle on data can also lead to BI problems. Lucker said midmarket companies sometimes fail to put the required effort into developing proper information management and data governance processes to support their BI systems. That doesn’t necessarily mean SMBs need to build data warehouses, but they do have to make sure that business users can effectively utilize available data, whether it’s from internal or external sources, he noted.

Yet another misstep stems from companies not understanding that doing BI well isn’t just a matter of hiring the right technical people. The development of BI processes should be viewed as a business project with technical components, Lucker said. “What makes for success,” he added, “is staying focused on the business and organizational aspects, where you not only deliver information but also change the management, organization, structure and training of people.”

But the ultimate pitfall, according to Gartner Inc. analyst Kurt Schlegel, may be failing to grasp that BI requirements evolve over time and that internal BI best practices need to be regularly revised and updated. That can be exacerbated by allowing a vendor-customer relationship to develop between IT or a BI team and the business, Schlegel said. A cross-functional BI project management approach that makes participants equal partners offers a better chance of avoiding problems, he added.

Alan R. Earls is a Boston-area freelance writer focused on business and technology.

Dig deeper on Business intelligence project management

Pro+

Features

Enjoy the benefits of Pro+ membership, learn more and join.

0 comments

Oldest 

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to:

-ADS BY GOOGLE

SearchDataManagement

SearchAWS

SearchContentManagement

SearchCRM

SearchOracle

SearchSAP

SearchSQLServer

Close