Separate brand and non-brand searches to distinguish people looking for your business from people searching for a service or answer. The split gives useful context to search reporting, but it is not a clean division between existing and new customers. Define the brand terms, record the rules and review ambiguous searches manually.
What counts as a brand search?
A brand search includes a recognisable reference to your business or one of its distinct brands. It may use the full name, a common abbreviation, a misspelling or a combination of the name and a service.
For an imaginary business called Northfield Workshop, “Northfield Workshop opening hours” is clearly branded. “Commercial joinery” is not. “Northfield joinery” may need checking if Northfield is also a place name.
That ambiguity is why a simple contains-the-name rule can misclassify some searches. Start with a useful working definition rather than claiming perfect accuracy.
Why the split changes the conversation
A rise in branded searches may follow recommendations, publicity, advertising or repeat customers looking for your contact details. It can be valuable without proving that your website has become more visible for a wider service market.
Non-brand searches can show how the site appears for questions that do not name the business. They can include people comparing suppliers, but also students, jobseekers and readers with no intention of buying.
Both groups matter. Reporting them separately prevents a strong month for one from hiding a weakness in the other.
Build and document the brand list
List the company name, trading names, distinctive product names and genuine common variations. Keep the scope sensible. Do not classify an ordinary industry word as branded merely because it appears in your company name.
Review the actual queries visible in Search Console to find variations you missed. Ask someone who knows the business to check the list.
Google’s filtering and comparison guidance explains ways to filter query data. Choose an approach that you can reproduce, then record the rule and the date you last checked it.
If you use a regular expression, which is a pattern for matching text, test it against clear brand, clear non-brand and ambiguous examples. A technically valid pattern can still describe the wrong group.
Search Console also offers a branded and non-branded query filter for eligible properties. Google notes that it is unavailable for sites with low impressions and that classifications can be wrong. If you use it, inspect representative queries and state that the grouping comes from Google rather than your own matching rule.
Keep an ambiguous group where needed
Not every query needs to be forced into a confident label. Short names, initials and words with other meanings can be difficult to interpret.
For the imaginary Northfield business, you might keep queries containing only “Northfield” under review rather than treating every one as recognition of the company.
A small manual check can improve the report more than a complicated classification that nobody understands. Record why the ambiguous group exists and whether it is large enough to affect your conclusions.
Compare the groups using the same conditions
Use the same period, country, search type and device scope for both groups. Look at clicks, impressions and the pages receiving traffic, rather than comparing only average position.
Also show the business relevance of the pages. A non-brand increase reaching an unrelated old article may not support the current service priorities.
For each group, ask:
- Which searches changed most?
- Which pages benefited or lost visibility?
- Does the content match the searcher’s likely question?
- Is there a useful next step for an interested buyer?
- What evidence would justify a change to the site?
Explain the limits of the figures
Search Console does not expose every individual query. Google’s performance-data documentation explains anonymised queries and related reporting limitations. Query-based groups therefore need careful labelling and may not reconcile neatly with every headline total.
More fundamentally, a query does not reveal whether someone has bought from you before. A new prospect can search your name after a recommendation. An existing customer can search a generic service term.
Call the groups brand and non-brand, rather than new and returning customers. That keeps the conclusion within what the evidence can support.
Use the split to choose the next action
If brand searches dominate, review whether the site clearly explains the services and answers relevant non-brand questions. If non-brand traffic is growing but enquiries are unsuitable, investigate the topic and page fit.
Our guide to impressions and clicks explains the underlying measures. The monthly report guide shows how to turn them into decisions. Our SEO service can help connect that analysis with specific page improvements.