Why do Business Intelligence Projects Fail?

Share this Post

According to Gartner, the famous technology research group, 70% to 80% of Business Intelligence (BI) and data warehousing (DW) projects end up in the trash bin. BI projects often fail, incurring colossal losses to organizations. Why do DW/BI projects fail? What can organizations do to avoid pitfalls that cause DW/BI projects to fail?

Influenced by my experience salvaging failed BI/DW projects, this blog post shares few insights on what organizations can do to clear roadblocks and anxieties which prevents them from implementing DW/BI solutions with confidence.

Should you do it on your own or get help?

Decision makers, overestimating the skills and competencies of their IT departments (or to cut costs), may tend to assign the task of implementing a DW/BI project to their IT departments. However, the competencies of a seasoned DW/BI expert include much more than the knowledge of dimension tables, concepts, ETL, OLAP, and reporting tools. To be eligible to confer the responsibility of implementing a DW/BI tool, a team of BI professionals must have expertise in a plethora of BI technologies, tools, architectures, best practices, and delivery experience across diverse industries.

Begin with the end in mind

Organizations should thoroughly research what they expect from a BI/DW solution before embarking on initiating a BI/DW project. This ensures that organizations reap the maximum returns on investment and avoid the frustrating scenario where organizations find no value in the expensive solution they implemented.

Predictable schedule & predictable delivery

A BI/DW project cannot be implemented at a touch of a magic wand. In the initial implementation, the goal is to prioritize the delivery of mission-critical data and to make data available for specific analyses. The time taken to deliver a comprehensive BI/DW solution which is equipped with forecasting models, advance reports and dashboard requirements may vary from three months to one year, depending on the complexity and the nature of the business.

When implemented using the Agile methodology – designed to adjust to changing requirements – BI/DW project development activities are carried out iteratively, and incrementally releasing solution builds periodically. This method—as opposed to Waterfall methodology—ensures that the BI/DW implementation is continuously improved based on the end-user feedback.

New is good, but I like the old ways

BI/DW implementation teams often encounter end-users who may be reluctant to get onboard with the new solution. This can happen due to several reasons:

Lack of genuine interest shown by the leadership
It is of paramount importance that the leadership shows a genuine interest in adopting the new system. The leadership must set an example to the rest of the company by embracing the new system with faith.

Inadequate training
Enforced by the interest shown by the leadership, the implementation team should provide adequate training for end users who have been habituated to old ways, especially during the transition to the new system.

Author:
Asanka Padmakumara
Consultant – Business Intelligence

Asanka has more than eight years’ experience in the software industry with much of his latter years focused on business intelligence. His skill repertoire includes across-the-board experience working with Microsoft BI technologies, developing dashboards and delivering solutions in the banking, telco, apparel FMCG, and healthcare domains.