Custom dimensions are available as attributes of the dimension itself and can be used for filtering and creating charts. You can create custom dimensions using dimension attributes and values. 

Based on the datatype, custom dimensions are classified into two types.

  • Dimension banding: It is the classification of numerical values and date ranges into buckets. For example, you can have age 12-16 as the Teenage dimension, 17-25 as the Adults dimension; you can create recency with user dimension by applying the banding user.last_transaction_date.
  • Dimension grouping: It is the classification of dimensions into groups. This is applicable to all String and Enum type dimensions. For example, to see the split of transactions across registered customers and non-registered customers, you can classify non-loyalty and not-interested customers into a group say non-registered.

 Based on the attributes of the dimension, the creation flow changes automatically.

 To create custom dimensions, follow these steps.

  1. On the Insights+ main menu, click Library > Dimensions.

  1. Click Create custom dimension

  1. Fill in the fields based on the description provided in the table below.
Dimension NameSpecify a name for the dimension.
DescriptionProvide a brief description of the dimension.
Based onChoose the parameter on which you want to create a dimension. For example, Loyalty.
AttributeYou will see attributes based on the option selected parameter (in Based on).
  1. Click Next. Based on the datatype of the dimension attributes, you will see the corresponding grouping or banding option.
  2. Configure the dimension group or band, for the dimension. The options will appear automatically based on the dimension type.
  • Dimension Band (for numeric or date type)

No. of BandsSpecify the number of bands you want to have for the dimension.
Band 1, Band 2, ...Specify a name for each band.
Break 1, Break 2, ...Configure breakpoints for each band. The condition differs based on the Attribute selected.
For example, Recent (band name) if user.last_transaction_date < 30 days (breakpoint).

  • The number of bands you see depends on the value you set in No. of bands.
  • Band names appear as dimension values for that dimension.
  • Dimension Group (for String or Enum values)

Using existing attribute valuesTo select dimension values from the preselected attribute values for each group. This is selected automatically if attribute values are less than 100.
By uploading a CSV fileTo import all groups along with attribute values from a CSV file. This is selected automatically if attribute values are more than 100.
To upload a CSV file, click Upload File.
You need to have the following in the CSV file.
  • The first row should be group names or identifiers.
  • Other rows should be the values of each group.
  • You can create a maximum of 15 groups.
Group ValueSpecify a name for the current group.
ValuesSelect one or more dimension values from the drop-down list.
A value used in a group will not be available in another group.
Add GroupClick to add more groups to the dimension. You can add up to 15 groups.
  1. Select Using existing attribute values to select each dimension value from the existing values.
  2. In Group Value, specify a name for the current group.
  3. In Values, select one or more values from the drop-down list.
  4. To add another group to the dimension, click Add Group.
  5. Click Create

You will see the configured details as shown below

Modify custom dimension

You can only modify the name, and group or band details of a dimension.

To modify a custom dimension, navigate to the dimension and click Edit, do the necessary changes, and click Update.

You cannot update Based on parameter and Attribute of a dimension once created.