Narrow report data with Browse and filters

Use filters when a report block should include only part of your data. In the query editor, these controls live under Include only.

Content of this page

This page focuses on practical filtering. For the full query editor overview, see Design a report block.

What Include only means

An Include only field restricts the data used by the report block. If a field is empty, the block is usually not restricted by that field. If you select values, only matching data is included.

Example:

  • If Devices only is empty, the block can include all device types.
  • If Devices only contains mobile, the block includes only matching mobile data.

Use Browse when it is available

Most filter fields have a Browse action. Browse opens a list of known values from your Keytiles data, such as event source types, campaigns, content types, devices, visitor types, or tags.

Use Browse instead of typing whenever possible. It helps avoid spelling mistakes, wrong IDs, and filters that accidentally match nothing.

[screenshot: Include only section with Browse-backed selected chips]

Selected values

Selected filter values appear as chips or compact rows. You can remove individual values, and many fields also provide a Clear all action when values are selected.

For unordered filters, the visible order of selected values is only for readability. It does not change the meaning of the filter.

Browse-only fields

Many fields are Browse-only because Keytiles already knows the valid values. For these fields, choosing from Browse is safer than typing manually.

Browse-only fields include common filters such as:

  • Content (Tile) types only.
  • Event source types only.
  • Event source names only.
  • Campaigns only.
  • Campaign mediums only.
  • Campaign contents only.
  • Devices only.
  • Visitor types only.
  • Primary tags only.
  • Secondary tags only.

Manual-entry exceptions

Some filters still need manual entry because not every useful value can be safely selected from Browse.

  • Languages only uses manual entry.
  • Content structure path matching only supports manual entry because wildcard or path-pattern matching can be useful.

For manual values, enter one meaningful value or pattern at a time. Keep values consistent with the way your tracked data is stored.

Content structure path matching

Content structure paths are useful when you want to focus a report block on one part of a website or content tree. This field supports Browse, but manual path patterns may also be useful when you need broader matching.

Use this filter carefully. A path pattern that is too narrow can remove all rows from the report block.

[screenshot: Content structure path matching selector with Browse and manual add]

Referrer filters

Referrer performance blocks can include only matching referrer data. Typical filters are source type and source name.

Keep source type and source name consistent. For example, selecting a source name that does not belong to the selected source type can produce an empty or incomplete report block.

Campaign filters

Campaign performance blocks can include only selected campaign values. Typical filters are campaign, campaign medium, and campaign content.

These filters are helpful when one setup should report on a specific campaign family, channel, or content variant.

Additional filters

Additional filters are less common Include only filters. They include device type, visitor type, language, primary tags, and secondary tags.

Use them when they are part of the question you want to answer. For example:

  • Use device type when the report should focus only on mobile or desktop behavior.
  • Use visitor type when the report should focus on new or returning visitors.
  • Use language when content language matters.
  • Use tags when your content tagging model is part of the analysis.

Avoid empty report blocks

Filters are powerful, but each filter narrows the data. If a report block returns no rows, check these things first:

  1. Remove the newest filter and generate the report again.
  2. Check whether selected Browse values still exist in the selected report period.
  3. Check whether two filters contradict each other.
  4. Check whether the selected events exist for the filtered data.
  5. Check whether the breakdown creates rows that are too sparse for the selected period.

When building a new report, start broad. Add filters one by one only after the basic result looks correct.