How to prepare for a business intelligence implementation project?
Each project has its own scope. A well-established approach would be to document the needs of the analytics, align them internally, and present them to the supplier as a functional requirements document (FRD). In practice, however, this does not work. If a company is implementing business analytics for the first time, it is almost impossible to professionally prepare detailed requirements. Without using the tool, it is difficult to fully understand its capabilities and requirements are often similar to the same reports that a company already uses.
​
Therefore, in practice, more successful projects start with a prototype or smaller scale project. The requirements for a prototype are much lower because the project itself is smaller. It is better to choose an area for the prototype for which the selected installer already has a workpiece or at least experience in other projects.
​
Before implementing a prototype, it is necessary to answer the question "who in the company will be responsible for analytics?". It is the employee who is responsible for ensuring that everyone receives the required data as efficiently as possible and that the data is as meaningful as possible. He understands how analytics systems work, what can be expected from them, but also understands the processes going on in the company. Without this person, the implementation of analytics within the company becomes very risky.
​
When implementing a prototype, the company begins to better understand the implementation process itself, responsibilities, how changes in the process and data are reflected in the analytics. How certain insights begin to change the process itself. After the implementation of the prototype, it is possible to professionally prepare the requirements for the next stage / project. Trying to install "everything at once" is a very wrong path. Within a company, it is usually not possible to focus as much on analytics deployment as would be required for a full deployment in all areas at once. Therefore, it is best to choose the next area and move on.