Split your Search Console queries in two. Filter the Performance report for queries containing your brand name, then for queries not containing it, and compare each against the same period before. If only branded clicks grew, people who already know you are searching for you more, which usually follows other marketing. Non-branded growth is the SEO signal.
Most founders look at the total clicks line, see it go up, and credit the SEO work. Sometimes that is right. Often a large share of the rise is people typing your company name, and those people found you somewhere else first. The split takes about fifteen minutes, needs nothing but Search Console, and gives you two lines that mean different things.
What is the difference between branded and non-branded search?
Branded search is people looking for you. Non-branded search is people looking for what you do.
A branded query contains your company name, a product name, your domain, or a misspelling of any of them. Someone typing your name followed by “pricing” or “login” is branded. So is someone typing your founder’s name, if that is how people find you.
A non-branded query describes a problem, a category or a task without naming you. “How to reduce churn in a B2B SaaS”, “invoice software for small agencies”, “what is answer engine optimization”. The person typing it may never have heard of you. If they click through to your site, your page earned that visit on its merits.
The distinction matters because the two grow for different reasons. Branded search follows awareness. Non-branded search follows the pages you publish and how well search engines rank them. A single total clicks line mixes them, and the mix can hide a flat SEO programme behind a successful podcast, or a working SEO programme behind a quiet quarter for everything else.
Why does branded search usually follow other marketing?
Nobody searches for a name they have not heard. Something else put it in their head.
Think about how someone comes to type your company name. They heard it on a podcast, saw a post, were told by a colleague, read it in a newsletter, met you at an event, or saw it named in an answer from an AI assistant. Then, later, they searched for you to find the site. The search engine is where the visit is recorded, but not where the interest began.
That is why a rise in branded clicks is good news for the business and weak evidence for the SEO. It says your name is travelling. It does not say your pages are winning new searches. SEO work can help branded search at the edges, for example by making sure your site, not a directory or a review site, is the first result for your name. It rarely creates the demand.
The practical consequence: when branded search jumps, look at what else happened in the weeks before. A launch, a round of press, a campaign, a talk. The timing will often line up. Lining up is not proof, but it tells you where to look for the cause, and it is usually not the blog.
Nobody searches for a name they have not heard. Something else put it in their head.
How do I split branded and non-branded in Search Console?
The Performance report, the Query filter, and a date comparison. Nothing else is needed.
First, list your brand terms. Do this on paper before you open anything.
- Your company name, with and without spaces or hyphens.
- Common misspellings you have seen in emails or in the Queries list itself.
- Product names that only you use.
- Your domain name, since some people type it into the search box.
- The founder’s name, if people search for it to find the company.
Then run the split.
- Open Search Console, choose your property, and go to Performance, then Search results.
- Set the date range. Last three months is a sensible default. Tick all four metric boxes at the top: total clicks, total impressions, average CTR and average position.
- Add a filter from the filter bar above the chart, and choose Query.
- For a single brand word, choose “Queries containing” and type it. For several spellings at once, choose Custom (regex), select “Matches regex”, and enter your terms separated by a vertical bar, for example acme|acmehq|akme. Search Console uses RE2 regular expressions; starting the pattern with (?i) makes it ignore capital letters.
- Apply the filter and note the four totals. That is your branded view.
- Edit the same filter and switch it to “Queries not containing”, or to “Doesn’t match regex” with the same pattern. Note the four totals again. That is your non-branded view.
Now open the date picker again and use its Compare tab. Compare the last three months to the previous period, or to the same period last year if your business is seasonal. Run the comparison once with the branded filter on and once with the non-branded filter on. You now have four numbers per metric instead of one, and the question in the title has an answer.
One caveat before you read them. Search Console hides some rare queries to protect searchers’ privacy. Those anonymised queries count in the unfiltered totals but cannot be matched by a query filter, so your branded and non-branded totals will usually add up to less than the total. That gap is normal. Compare each half with itself over time, not with the grand total.
Search Console also has a built-in branded queries filter in the Performance report, introduced in November 2025 and available to all eligible sites since March 2026. Google decides what counts as branded, including misspellings and product names, so check its split against your own list of brand terms before relying on it. The regex method works on every property and leaves the definition of your brand to you.
How do I read the two halves?
The same metric means different things on each side of the split.
Branded queries tend to have a high click-through rate and an average position close to the top, because people searching for you usually want you and you usually rank first for your own name. Non-branded queries tend to have lower click-through rates and positions that vary widely. Do not compare the two sides against each other and conclude that non-branded is “performing worse”. They are different kinds of search.
| Signal | Branded | Non-branded | What it means |
|---|---|---|---|
| Clicks rising | More people who know you are coming to find you | Your pages are earning visits from people who did not know you | Only the non-branded rise is evidence the SEO is working |
| Impressions rising, clicks flat | More people searching your name but choosing another result, such as a review site | You are appearing for more searches but not yet high enough, or not compelling enough, to be clicked | Branded: check what outranks you for your name. Non-branded: look at position and titles |
| Average position | Should sit near the top; if not, something else owns your name | Improving positions on relevant queries usually come before clicks do | Watch the trend for non-branded; watch for problems on branded |
| Average CTR | Typically high; a drop can mean a confusing result or a competitor bidding on your name | Typically lower and varies by query; improving titles and descriptions can help | Compare each side with its own history, not with the other side |
| New queries appearing | New products or campaigns people now search for by name | Your pages are being shown for problems you had not ranked for before | Non-branded newcomers are the early sign of SEO reach widening |
| Pages earning clicks | Mostly the homepage, pricing and login | Articles, guides and product pages that answer a specific question | If the homepage earns most non-branded clicks, check which queries; they may be near-branded |
If non-branded clicks and impressions are both rising over equal periods, and the rise holds when you compare year over year, that is the clearest sign in Search Console that SEO is growing. If only branded is rising, the business may be doing well, but the SEO is standing still.
Which pages are actually earning non-branded traffic?
Keep the non-branded filter on and switch from Queries to Pages.
With the non-branded filter still applied, click the Pages tab below the chart. Now you see which URLs earn clicks from people who did not search for you by name. For a company doing SEO seriously, this list should be mostly the pages written to answer buyer questions: guides, comparisons, explainers, product pages built around a use case.
Click any page in that list and then switch back to the Queries tab. Search Console now shows the queries that page appeared for. This is where you see whether a page is winning the searches you wrote it for, or different ones, and whether it sits on page one or just below it.
Two things often turn up. One is a homepage that earns a lot of “non-branded” clicks, which on inspection turn out to be brand misspellings your pattern missed. Add them to the regex and run the split again. The other is a small number of pages doing nearly all of the non-branded work. That is useful to know: those pages are what your SEO actually is, whatever the content calendar says.
What usually goes wrong with the split?
Most mistakes come from the brand list, the dates, or asking the data for a cause.
Check these before you share the numbers with anyone.
- A brand name that is also an ordinary word. If your company is named after a common noun, “Queries containing” will sweep in unrelated searches. Use a regex that matches your name only alongside words you know are yours, and inspect the matched queries by eye.
- Unequal or uneven periods. Compare equal lengths of complete days. A range that includes a holiday, a launch or a site migration on one side and not the other will mislead you.
- Competitor names. Queries that name a competitor are non-branded for you, but they behave differently from category searches. If there are many, split them into a third bucket.
- Reading a rise as a result. Non-branded clicks growing in the months after you published new pages is consistent with the pages working. It is not proof. Algorithm updates, seasonality and competitors changing their sites can all move the line.
None of these is a reason to skip the split. They are reasons to write the brand pattern down, keep it with the numbers, and use the same pattern and the same kind of comparison every time. A split you can repeat next month is worth more than a clever one you cannot.
Total clicks tell you that people arrived. The split tells you whether they already knew your name. Only one of those is your SEO.
Questions people ask next
How do I filter out branded queries in Google Search Console?
In Performance, then Search results, add a Query filter and choose “Queries not containing” with your brand name, or Custom (regex) with “Doesn’t match regex” and your brand spellings separated by a vertical bar.
Why don’t my branded and non-branded totals add up to the total clicks?
Search Console hides some rare queries for privacy. They count in the unfiltered totals but cannot be matched by a query filter, so the two filtered halves usually add up to less. Compare each half with its own history instead.
Is growing branded search a sign that SEO is working?
Not on its own. Branded search grows when more people already know your name, which usually follows other marketing, press or word of mouth. Non-branded growth over equal periods is the better SEO signal.
How far back can I compare in Search Console?
The Performance report keeps about sixteen months of data, which is enough for a year-over-year comparison of the most recent months. Export the numbers if you want a longer record.
Can Throughline show me whether my search growth is branded or not?
It reads Search Console every morning and reports which queries and pages moved, which shows whether a rise sits in your name or in searches that never mention it. It then sets that movement beside bookings and deals over the same weeks, as correlation.