Separate branded and non-branded performance

Use this guide to split recorded search queries into branded and non-branded groups. The assistant checks the brand terms and returns performance for both groups by engine.

When to use this use case

Use this guide to review branded demand separately from other search visibility. The split describes returned query data, not all traffic recorded for the property.

What to ask

Replace the domain, dates, and brand names before using this prompt.

Review branded and non-branded web-search performance for example.com during the 28 days ending September 30, 2026, compared with the previous 28 days. Check the effective brand terms against these names and aliases: [brand names]. Show example branded queries so I can check the classification. Keep Google and Bing separate. Explain what the reported shares measure and flag limited query coverage. Do not change brand settings without asking me.

How the assistant runs it

The assistant checks classification before comparing the groups:

Call get_server_status, list_sites, and get_connection_status to confirm the site and resolved workspace.
Call get_workspace_brand_terms with that workspace ID. Read the effective brandTerms, confirmed aliases, and unconfirmed candidates.
Compare the effective terms with the user's supplied names. Stop for a classification correction if the terms do not represent the brand. Do not silently treat an internal workspace title as a complete brand definition.
Call analyze_branded_vs_nonbranded for the current window. Read the applied terms and sample branded queries.
Call the tool for the preceding window. Compare each group's metrics by engine.
Report classification and coverage limits alongside the changes.

Example tool inputs

These inputs compare two adjacent 28-day windows.

json

{  "tool": "analyze_branded_vs_nonbranded",  "arguments": {    "site": "example.com",    "to": "2026-09-30",    "windowDays": 28,    "searchVertical": "web"  }}

json

{  "tool": "analyze_branded_vs_nonbranded",  "arguments": {    "site": "example.com",    "to": "2026-09-02",    "windowDays": 28,    "searchVertical": "web"  }}

What the result includes

The comparison contains the following information:

The effective brand terms and sample branded queries.
Branded and non-branded clicks, impressions, CTR, and position by engine.
Changes in each group's metrics across the selected periods.
Each group's share of the returned query clicks and impressions.
Classification concerns and query-coverage limits.

How to read the result

The tool uses the workspace title and confirmed aliases. Candidates do not affect matching until they are confirmed. The same effective terms also affect branded-query exclusions in other tools.

Shares use the returned query rows as their denominator. They are not complete shares of property traffic. If Google truncates the query list, it can omit non-branded queries with fewer impressions and overstate the branded share.

A change in brand aliases can change the split without a change in total traffic. Keep the classification consistent across the compared periods. Describe share changes in percentage points, and keep average position separate by engine.

This use case does not measure brand mentions in AI answers, competitor visibility, or conversions.