Why Business Intelligence projects fail
There is much talk about Big Data and Business Intelligence (BI). A miracle machine to easily increase revenue or save costs. Business Intelligence projects are, however, are often flown wrong. Therefore, the benefits are not converted into hard cash. This has several causes.
First of all, a big group of managers and administrators take decisions based on intuition and rules. This is fine, but it carries risks. Decisions with (big) data are fundamentally different. An analysis that is based on the facts doesn’t care as much about the existing images. Fact-based decision-making means that the decider and the employee have to adjust their way of making decisions.
Another big problem is that there are huge differences between departments, even though they use the same system. Differences in vision, capturing data and the working process. A BI- system that is meant to solve these differences is doomed to fail. In order to do something with Big Data is regularly opportunistically startswith the implementation of BI tools based on slick presentations and success stories. The focus isn’t on the end user, but in reading and linking of all possible data that could be interesting.
This makes data inconsistent and it has missing links. To make measurements and not bother the user, alfa rhythms are often used to clean the data automatically. These alfa rhythms are becoming more and more complex and give the user perfect excuses not to use the BI-system: this is incorrect. Users that don’t want to participate, mail lists with evidential mistakes of the automatic cleansing. It doesn’t take long before an image is created whereby the results aren’t trusted.
Finally, there are great BI tools on the market that make the organization into a data-driven power station. Often decision makers are in love with the tool, buy it, and underestimate the costs associated with the training of their employees. In practice, many employees seem to be sold to their own reports.
An implementation of a BI is no system implementation. Een implementatie van BI is geen systeemimplementatie. It intervenes in the way of making decisions. Furthermore, the tactics play a big role: involving existing users and taking the existing reports into the solution. You must also keep the updates of key data in high priority. This way your profits will be much bigger than the benefit of 1 BI-tool compared to another.