The short version
- Branded search measures demand you already created. Non-branded measures demand you are capturing from strangers. Averaging them tells you nothing.
- Branded traffic almost always converts better, which inflates blended organic conversion rate and flatters your SEO reporting.
- Rising branded search with flat non-branded is the signature of brand marketing working — including AI assistants recommending you.
- Splitting them is a free five-minute filter in Search Console and almost nobody does it.
Almost every organic search report averages two populations that have nothing in common. People who typed your company name already knew about you; something else in your marketing did that work. People who typed a generic query describing their problem did not know you existed. Reporting these as one number produces a figure that describes neither.
The split is trivially easy to make and the insight is immediate, which makes its rarity slightly baffling. It is the single highest-value five minutes available in most analytics setups.
The short answer
Fit for your workflow
Report them separately, always. They answer different questions.
Non-branded performance tells you whether your SEO is reaching new people. Branded performance tells you whether your brand, product and everything else you do is creating demand. Both matter; neither substitutes for the other. In Search Console, filter queries containing your brand name and its common misspellings, then look at both halves.
- Non-branded rising
- SEO is working — you are reaching people who did not know you.
- Branded rising, non-branded flat
- Brand marketing or word of mouth is working. Possibly AI recommendations.
- Both flat
- Nothing is compounding. Diagnose before spending more.
- Branded falling
- Investigate urgently — this often precedes broader decline.
What each one actually tells you
The conversion-rate row is why this matters commercially rather than just analytically. If branded traffic converts at several times the rate of non-branded and your branded share grows, your blended organic conversion rate improves without your SEO having done anything at all. Teams routinely report that as an SEO improvement. It is not — and the mistake leads to funding the wrong thing.
How to split it
- Open the Search Console performance report and add a query filter.
- Filter for queries containing your brand name. Include common misspellings and any product names people search for directly.
- Note impressions, clicks and average position. That is your branded baseline.
- Invert the filter to queries not containing your brand. That is your non-branded baseline.
- Track both monthly. The ratio between them is as informative as either number on its own.
Defining your brand filter properly
The split is only as good as the filter, and most filters are too narrow. Include common misspellings, spacing variants, your product names, your founder’s name if people search it, and combinations like "brand + reviews" or "brand + pricing" that are unambiguously branded even though they contain other terms.
Two traps are worth knowing. If your brand name is also a common word, a naive contains-filter will sweep in genuinely non-branded queries and overstate your branded share. And Search Console withholds some long-tail queries for privacy reasons, so neither bucket is complete — the ratio between them is more reliable than either absolute number. Write the filter down and keep using the same one, because a definition that drifts makes your trend line meaningless.
Branded search as an AI-visibility proxy
There is no first-party data on how often assistants recommend you. But if they are, some of those people will search your name afterwards. Branded search impressions rising with no corresponding campaign is one of the few free signals that assistant-driven discovery is happening. It is indirect and confounded by everything else you do — and it is better than nothing, which is the alternative. See ranking vs being cited.