We now know that there are very good reasons for adopting an ETL approach to loading data in QlikView. Now we need to learn how we should go about implementing the approach.

Each part—Extract, Transform, and Load—has its own set of recommendations because each part has a very different function.

Essentially, the approach looks like this:

The approach can be explained as follows:

The final two layers, the transformed QVDs and the final applications, become potential sources for a user's self-service. We can have confidence that users who load data from these layers will be getting access to clean, governed data.