Sign in Start free

Separate Branded From Non-Branded Properly

Branded search largely reflects demand created elsewhere - ads, PR, word of mouth - and mixing it into SEO reporting hides whether the work is doing anything. A simple filter on the brand name is not enough: misspellings, product names, staff names and domain-string searches are all branded and all leak through.

CATEGORY
Keyword research
FORMAT
separate-branded-from-nonbranded.md
STEPS
6
PRICE
Free - no account
WHEN TO REACH FOR THIS

Use when reporting SEO performance or sizing opportunity and branded search is inflating the numbers.

The skill file

separate-branded-from-nonbranded.md 
---
name: separate-branded-from-nonbranded
description: Use when reporting SEO performance or sizing opportunity and branded search is inflating the numbers.
---

# Separate Branded From Non-Branded Properly

Branded search largely reflects demand created elsewhere - ads, PR, word of mouth - and mixing it into SEO reporting hides whether the work is doing anything. A simple filter on the brand name is not enough: misspellings, product names, staff names and domain-string searches are all branded and all leak through.

## What you need first

- full Search Console query export
- a list of brand names, product names, common misspellings and any acquired brands
- agreement with stakeholders on what counts as branded

## Method

1. Build a regex covering the brand, plural and possessive forms, spacing variants, the domain with and without the TLD, and every misspelling that appears in the export with more than a handful of impressions.
2. Add trademarked product names and any founder or spokesperson name, which behave as branded even though they do not contain the company name.
3. Create a third bucket for hybrid queries such as brand plus a category term. These are neither pure brand nor pure discovery and should be reported separately, because they indicate consideration not awareness.
4. Recompute all headline metrics with branded excluded, and expect a large drop - a fall from 60 percent to 25 percent branded share is common once misspellings are captured.
5. Track branded impressions separately as a proxy for brand demand, so a marketing campaign is not misread as an SEO win or a paid pause misread as an SEO loss.
6. Re-audit the regex quarterly, because new products and new misspellings appear continuously and a stale filter silently reintroduces brand traffic.

## What this produces

A branded, non-branded and hybrid segmentation of query data with a documented regex anyone can reproduce.

## Where this goes wrong

- Filtering on the exact brand string only, which misses the misspellings that often make up a fifth of branded volume.
- Excluding hybrid queries as branded, thereby hiding the terms with the best conversion rates on the site.
- Comparing non-branded performance across a period when the brand filter definition changed, producing a fake trend.

---

From the QuQi skill library - https://www.quqi.io/skills/separate-branded-from-nonbranded
Free to download · no account, no email

What you need first

  • full Search Console query export
  • a list of brand names, product names, common misspellings and any acquired brands
  • agreement with stakeholders on what counts as branded

Method

  1. 01 Build a regex covering the brand, plural and possessive forms, spacing variants, the domain with and without the TLD, and every misspelling that appears in the export with more than a handful of impressions.
  2. 02 Add trademarked product names and any founder or spokesperson name, which behave as branded even though they do not contain the company name.
  3. 03 Create a third bucket for hybrid queries such as brand plus a category term. These are neither pure brand nor pure discovery and should be reported separately, because they indicate consideration not awareness.
  4. 04 Recompute all headline metrics with branded excluded, and expect a large drop - a fall from 60 percent to 25 percent branded share is common once misspellings are captured.
  5. 05 Track branded impressions separately as a proxy for brand demand, so a marketing campaign is not misread as an SEO win or a paid pause misread as an SEO loss.
  6. 06 Re-audit the regex quarterly, because new products and new misspellings appear continuously and a stale filter silently reintroduces brand traffic.

What this produces

A branded, non-branded and hybrid segmentation of query data with a documented regex anyone can reproduce.

Where this goes wrong

  • Filtering on the exact brand string only, which misses the misspellings that often make up a fifth of branded volume.
  • Excluding hybrid queries as branded, thereby hiding the terms with the best conversion rates on the site.
  • Comparing non-branded performance across a period when the brand filter definition changed, producing a fake trend.

Use this skill in your own AI

The download is a plain markdown file with the name and trigger in its frontmatter. Where an assistant supports skills it can load itself, that frontmatter is what it reads to decide this one applies.

Claude Code Save it as ~/.claude/skills/separate-branded-from-nonbranded/SKILL.md and Claude loads it on its own when what you are doing matches the trigger line. Put it in .claude/skills inside a project instead if the whole team should have it.
Claude Upload the file in the skills section of your settings. Once it is there it applies itself in any conversation where the trigger fits, so you do not have to remember it exists.
ChatGPT There is no skills format to install into, so paste the file contents into a Project instruction or a Custom GPT instead. It then applies to every chat in that project rather than only the one you paste it into.
Anything else Paste the markdown into the chat before your question. It works in any assistant, it just has to be pasted again each time.

Questions about this skill

When is a proper segmentation worth the effort rather than a filter on the brand name?

Whenever branded search is large enough to move the headline number, which is most established sites. Filtering the exact brand string leaves misspellings, product names, staff names and domain-string searches sitting in the non-branded bucket. Those regularly account for a fifth of branded volume, which is enough for an SEO report to describe the results of a PR campaign instead.

What do I need in hand before starting?

The full query export, a list of brand names, product names, acquired brands, founder or spokesperson names and the misspellings that actually appear in the data, and agreement with stakeholders on the definition before you publish any number. Without that agreement the first person who dislikes the drop relitigates the filter instead of the finding, and the reporting change never lands.

What do I end up with, and which part gets used?

Three buckets, branded, non-branded and hybrid, plus a regex anyone can rerun. The regex is the real deliverable, because an undocumented filter produces a number nobody can reproduce or challenge. The hybrid bucket, brand plus a category term, is the one worth watching separately: it signals consideration rather than awareness and often carries the best conversion rate on the site.

What is the mistake that ruins this, and what does it cost?

Changing the definition mid-series and then comparing periods. A regex that caught more misspellings this quarter manufactures a non-branded decline that did not happen, and somebody acts on it. Re-audit quarterly because new products and misspellings appear continuously, but restate the earlier periods with the new filter before showing any comparison.

More in Keyword research

Read Intent From The Live SERP

Use when you have a keyword list and need to decide what kind of page, if any, can actually rank...

Striking Distance Prioritisation

Use when you need traffic in weeks rather than months and already have a site with Search Consol...

Map Queries To Pages From Real Data

Use when several pages compete for the same terms or you cannot say which page owns which query.

Zero Volume Terms That Convert

Use when tool volume says zero or ten but you suspect real buyers are searching in those words.