Data Discovery

Use the data discovery function to construct pivot tables.

Adding data to a pivot allows you to create a cross-reference display to summarise trends, customer segments and transaction hotspots.

Tip: you can also add rules to the pivot domain to refine your data.

Adding data to a pivot

On the table tab:

  1. Drag-and-drop the banding that will form your columns into the columns drop zone.

  2. Drag-and-drop the banding that will form your rows into the rows drop zone.

    Tip: to switch axes, drag the banding name from one axis to the other within the table.

  3. Continue adding bandings as required.

  4. Filter any unwanted values/entries in your table using the cog icon next to the banding name.

  5. Set your view styles in the properties pane.

  6. When you are happy with your data discovery configuration, press the Save or Save as button in the function menu.

  7. Enter a name for your data discovery.

  8. Enter a description.

    Tip: you can also enter a description in the properties pane. This description will pre-populate the field when saving.

  9. Select a location to save the data discovery within Audiences.

    Tip: create folders and subfolders to organise your saved elements.

  10. Press the Save button.

    Tip: you can also export your data discovery results to a new audience, an intersection, or a label.

Filtering your data discovery values

When you add a banding to a row or column, you may not want to include all values. To filter your data:

  1. Press the cog icon next to the banding name.

  2. In the dialog box, scroll the table and deselect the values you wish to exclude from your table.

    Tip: use the AZ and ZA buttons to sort your fields by ascending or descending order.

  3. If you have a label filter to apply:

    • Press the Labels button.

    • Select an operator, e.g. Between.

    • Enter the label names/values.

  4. If you want to apply a value filter, e.g. count the top 100:

    • Press the Values button.

    • Select an operator, e.g. Between.

    • Enter the values to use for your results.

  5. When you have configured your filter settings, press the Apply button.

Tip: you can also set a domain to limit the records available for your data discovery.

The database table your pivot operates at is automatically updated based on your applied bandings.

If you need the pivot to function against a different table:

  1. In the properties pane, change the table the field is calculated against.

Note: calculating a field at the wrong table level can result in errors.

Setting data discovery view properties

Change the view settings for your table in the properties pane.

Under the Display menu:

  1. Select a table pattern from the grid type dropdown.

    • Table - a simple patterned row table.

    • Heatmap - displays cells in different colours according to cell value to make outliers more visually distinct.

    • Row heatmap - a heatmap that bases its colour gradient on row data variation.

    • Column heatmap - a heatmap that bases its colour gradient on column data variation.

  2. Select a number format from the value dropdown.

    • Measures - an integer display of the cell values.

    • Row percent - the integer value of the cell converted to a percentage of the row total.

    • Column percent - the integer value of the cell converted to a percentage of the column total.

    • Grid percent - the integer value of the cell converted to a percentage of the grid total.

If you wish to view your data as a chart, select the chart tab at the top of the workspace.

When viewing as a chart, the grid type dropdown is replaced with a chart type dropdown to change your chart type, e.g. horizontal bar, vertical bar, etc.

To view your table and chart simultaneously, select the mixed tab at the top of the workspace.