Companies engage in Business Intelligence projects with brilliant strategic intentions. A 2018 study by Dresner Advisory Services shows that while better decision-making and improved operational efficiencies and cost savings still top the list of BI objectives, expectations of enhanced customer service, revenue growth, and gains in competitive advantage are growing rapidly. With such high expectations, funds are readily allocated to BI projects. Still, a staggering 70 to 80 percent of Business Intelligence projects fail according to analysts at Gartner.
There are many factors that play into BI project failure, but if each is anticipated and mitigated proactively, the chances of project success are far greater, and the company will fully realize a return on their investment.
Executive Support: Many BI projects are doomed to fail from the get-go because executives are not sufficiently engaged in the project. They may expect too much too soon, only passively track project progress, or are mostly focused on budgets and deadlines. Executives must ensure that project expectations are clear and reasonable, and they need to champion the project throughout its design and execution to ensure that their staff has the support it needs to complete the project successfully. Executives can even identify a specific problem—one they assign high priority to—to gauge if the solution can be achieved and yield benefits (e.g., time saving, new critical information) quickly (see below, under Too Large a Scope).
Reliance on Old Technologies: Businesses often get far too attached to their “legacy” systems because it is cost-prohibitive to upgrade, and users don’t want to learn a new way of doing things. But limitations of old technology must be acknowledged and dealt with for a BI project to be successful. This can be accomplished by either upgrading old systems or working with a BI firm specializing in integrating legacy systems with BI technologies. In the case of favored technologies—especially products like Excel or Access, currently being used to address some BI needs—find a vendor that can “knit” these tools into the overall solution. There should be no reason that a preferred front end, for example, can’t be included seamlessly into the overall BI project. (Put another way: be way of solution providers that advocate a wholesale technology overhaul.)
Too Large a Scope: It’s far better to start small with BI. You might think the smartest way forward is to implement business intelligence for inventory or human resources or some other specific department to test it out first—but it’s better to go to an even more granular level: pick one particular knotty task/situation that will yield immediate benefit, assuming a solution can be created relatively quickly by a vendor. Use this introductory project to identify lessons learned and best practices before trying to implement BI fully within a department, or even company-wide across multiple functions.
Insufficient Collaboration Between IT and the Rest of the Business: Too often, BI initiatives are treated solely as IT projects. But, for a corporate-wide project, it is impossible for IT to have a complete understanding of HR, accounting, operations, customer service, and other departments without working with them very closely throughout the project design phase. Even within department-wide projects in larger companies, there are complex and competing needs. Multiple iterations should be expected, with users offering inputs along the way. If all parties impacted by the project are engaged in its design, the chances for success are far greater.
Low Adoption Rates: Even after a company invests significant time and resources into a BI project, low adoption rates by staff will negate the benefits. 80% of companies with over 10,000 employees have BI adoption rates below 40%, according to Dresner. This often results from the lack of cross-functional collaboration during the project design phase. Other considerations that help boost adoption rates are to keep the results of BI – dashboards, etc. – simple to use and understand and to use information already in the company’s existing systems as much as possible. If users find the BI results too cumbersome to learn and use, or if the BI relies on new data that needs to be input by users, the project can be doomed to fail.
Training and Maintenance: BI is not a one-time project that provides a static solution for years to come. Companies grow, new metrics are added, and the competitive landscape evolves. While it makes sense to ensure that a company has IT and other functional personnel trained on the BI solution there is another, possibly safer way forward: engage a vendor to provide BI as a service, so that major part of updating or reworking the system is handled by experts. After all, the strength of BI lies in its contribution to business intelligence and strategic adaptation not only to present needs, but also to future circumstances.