October 24, 2015

Self-service BI - three flavours

Self-service BI - three flavours

As information management consultants we often get the question to integrate self-service BI in our solutions. This is question we welcome with open arms, not just because you can do technically beautiful things with this, but first and foremost because it means the organisation realizes the value that a well organized business intelligence platform can bring.

Three levels of self service

  1. Guided analytics. A well built QlikView dashboard already allows the users to interact in an intuitive way with the data. The associative logic and in-memory technology supports business discovery at the speed of thought. This is already one step beyond classic BI tools, forcing a certain hierarchy upon the drill down possibilities. This is a minimal form of self-service, but for some type of users already enough to discover a lot of value.

Guided analytics

  1. Ad hoc analysis. This allows you to select your own dimensions and metrics from a list and show them in a table or graph. This list of dimensions and metrics is predefined, but can span the whole breadth of data available in the source system. This is still a relatively easy way to set up self service capabilities in the organisation, with a maximum of control on the data being offered for the user. In QlikView a proven method exists to make this happen.

Ad hoc analysis

  1. Build your own dashboard. This is the maximal form of self-service BI. A certain familiarity with technology has to be present in the people being given access to the systems to develop their own dashboards. All data in the model is available, and even transformations in the model can be allowed. The power user has to be aware what kind of combinations of the data make sense in a visualization. This flavor of business intelligence leans more toward the way Sense - rather than QlikView - is conceived.

Build your own dashboard