Qlik implementation

More than building dashboards. You set up the layer beneath them: load scripts that retrieve data from sources, QVD files as an intermediate layer, a data model with master items, section access for permissions, and spaces and reload tasks in Qlik Cloud or on your own servers.

We do this as part of the data and AI platform, so the data behind Qlik remains available for other reporting and applications.

They design the model, write the load script, define calculations as master measures, and set up permissions and publishing. When building Qlik dashboards, they start with the user's question, not the chart.

Master items are saved dimensions, measures and visualisations. Change one, and it changes everywhere it is used in the app. That way, every sheet uses the same definition.

Qlik Sense filters according to your selections and assigns each value a colour: green for selected, white for possible, light grey for alternative and dark grey for excluded. This also lets you see what falls outside your selection, such as customers without orders.

This only works well when tables are linked by a single key field. If two tables share more fields, Qlik creates a synthetic key. According to Qlik, multiple synthetic keys often point to a flawed model.

By not reading the entire source every time. With incremental loading, you retrieve only new or changed records from the database and get the rest from a QVD file containing previously loaded data. You then write a new QVD file for the next reload.

Reading a QVD is fastest in optimised mode. A where clause or a transformation of fields during loading disables that mode.

With section access: a part of the load script containing a security table that defines what each user can see. When someone opens the app, Qlik hides the rest. You can restrict rows, columns or both.

Watch the capitalisation: Qlik always converts field names and values in section access to uppercase. The fields used to restrict the data must therefore also be uppercase in your data.

Qlik provides the Qlik Analytics Migration Tool for this. It moves content from Qlik Sense Client-Managed, QlikView and Qlik NPrinting to Qlik Cloud according to a migration plan. Qlik itself says it is not a one-click solution.

Reload tasks are not transferred when you publish to a managed space, so you need to set them up again. You can access sources behind your firewall through Qlik Data Gateway Direct Access, which creates only an outbound, encrypted connection.

That depends on what you already have. If you have used Qlik for years and your definitions are captured in sound load scripts and master items, cleaning things up often makes more sense than switching. If your organisation is bringing its reporting together in Microsoft Fabric, Power BI is worth considering.

We do not decide in advance. Read about our Power BI consultant and business intelligence with other tools.

If you have a few apps that work well, reload on time and are managed by a dedicated colleague. Or if the source is the problem: you cannot fix duplicate customers and missing keys in a load script.

Start instead with data quality or data governance. If a simpler solution is a better fit, we will say so.