Creating your own dashboards

Zetaly offers basic dashboards suited to most needs, but there may come a time when you will want a specific view on your metrics. This page will help you design your own dashboards.

If you are just starting, it can be easier to reuse existing content

Any dashboard or cube can be easily duplicated using the 3-dots menu on top right of their card. You can rename and modify the copy without affecting the original.

Gathering data into a cube

Add tables to the cube

The first step in designing a dashboard is to gather the necessary data. To do so, create a new cube in the "Data" tab within Sisense, then click the "+ Data" button : you will be prompted with a multi-pages menu to select your server, database and table to insert (the last connections will be displayed on the first line).

To find the tables most suited to your need, you can check the currently collected SMF data by Zetaly, which will help you link SMF fields and database tables.

Creating custom tables

Custom tables are table created directly inside your data model. These table doesn't refered to an existing database table. Their content is based on data already included inside the data model.

To add custom tables used the "+ Custom" button : this opens a field on the bottom of the screen that lets you create sub-tables within the cube, for example to add a filter on distinct CPCs.

Linking tables

 You can also link your tables togther using the "relationships" field in any table's menu :

Once the columns that act as links are selected, the relationship is created.

Designing a dashboard

With the cube ready, create a new dashboard in the "Analytics" tab with your cube as datasource.

Once on the dashboard page, add widgets by following the prompted instructions. You can find more details on available widgets on the Sisense Documentation.

Once added, any widget can be altered using the pen icon on the top right.

You will enter in the widget customization window that allows you to select data to display.

Optimizing the cube

Once you are content with your dashboard, you might find that it loads slowly. This is expected when working on raw tables that can contains hundreds of columns. You should determine which are effectively used, and design an appropriate view gathering them.
Afterwards, you can add the dimensions BiFilters, to enable Zetaly Filters on the dashboards, and Dim_Calendar, to handle timezones on hourly data (useless if your tables/views are grouped by day or more).