Branded traffic is the visits that come from searches that include the name of your company or your products, and in many reports it is mixed in with everything else. I bet you have seen an organic traffic chart going up month after month, with applause in the meeting, and nobody asking how many of those visits were people who were already looking for you by name.
So we looked at Search Console data from 131 company sites in the US, Latin America and other markets. In September 2026, the median share of organic clicks that came from searches with the brand name was 18%. But if you divide only by the clicks where Search Console shows the query, the same number goes up to 41%. And in 1 out of 7 sites, the brand took more than half of everything that comes from Google.
My take is that separating the brand from the rest is the bare minimum to evaluate SEO, and that you also have to say what you are dividing by, because the same site can report 18% or 41% without lying in either case.
Let's start with the basics and then get to the numbers.
What is branded traffic?
Branded traffic is the traffic that comes from queries containing the brand name, the name of one of its products or any of its variants: with typos, with spaces where they don't belong or next to another word ("name + login", "name + reviews", "name + branch"). Whoever types that already knows the company and is using Google as a shortcut.
Non-branded traffic is everything else: generic searches about what you sell or what your customer needs, without your name. In other words, people who haven't chosen a company yet.
A few examples with a made-up insurer, Acme Insurance:
| Search | Type | What it says about the searcher |
|---|---|---|
| acme insurance | Brand | Already knows the company |
| acme insurance login | Brand | Is already a customer |
| acme insurance reviews | Brand | Is choosing among options they already know |
| car insurance for young drivers | Non-brand | Hasn't chosen a company yet |
| how much does car insurance cost | Non-brand | Is learning |
The queries at the top are also called branded keywords. If you want to review the other types, there is a guide on what keywords are.
A practical detail: in theory, the brand includes variants and named products. In practice, the most common filter only looks for the domain name, so it falls short. That will matter in the methodology.
Why separate branded from non-branded traffic?
Because they respond to different things. Branded searches depend mostly on brand awareness: TV, physical stores, word of mouth, the end-of-year campaign, the customers who log in to pay their bill. SEO helps that visit land on the right page, but the demand already existed.
Non-branded traffic, on the other hand, is new demand. That is where you see whether the site shows up when someone looks for a solution without knowing who is going to provide it.
If you report them together, two things happen. SEO takes credit for other teams' campaigns (and also the blame, when marketing cuts media spend and the brand drops). And a good month in brand can hide a bad quarter in non-brand, with nobody noticing until it's too late. It's the same logic as the metrics a CMO should look at in organic: the aggregate number looks good on a slide and helps little with business decisions.
The guides ranking for this topic today agree on all of that. Where they stop agreeing is on how much the brand should weigh: some say there is no ideal ratio (true) and none of them gives a number. So we measured it.
How much of a company's organic traffic is branded?
On the typical site in the sample, the brand was 1.3% of impressions and 18.2% of organic clicks (medians of 131 sites, September 2026). This is the full table from the study:
| Metric (Sep 2026) | Median | p25 | p75 | Sites |
|---|---|---|---|---|
| % brand over total clicks | 18.2% | 5.5% | 39.5% | 131 |
| % brand over visible (query not hidden) | 41.2% | 13.2% | 73.7% | 131 |
| % brand over total impressions | 1.3% | 0.4% | 3.6% | 131 |
| % with hidden query | 47.8% | 37.3% | 57.2% | 131 |
| CTR, brand | 15.4% | 8.7% | 32.2% | 131 |
| CTR, non-brand (visible) | 0.5% | 0.2% | 1.6% | 131 |
| Brand / non-brand ratio (previous two rows) | 22.3 | 8.0 | 91.7 | 131 |
| Average position, brand | 2.7 | 1.8 | 5.4 | 131 |
| % brand, US | 17.9% | 6.0% | 35.3% | 74 |
| % brand, Latin America | 21.9% | 6.8% | 48.4% | 30 |
| % brand, June 2026 | 15.5% | 6.3% | 36.5% | 129 |
| % brand, September 2026 | 18.2% | 5.6% | 38.1% | 129 |
| Sep - Jun difference per site (points) | 0.0 | -1.6 | +1.4 | 129 |
| Change in total clicks, Jun to Sep | -8.2% | -23.3% | +5.6% | 129 |
(p25 and p75 are the 25th and 75th percentiles: half of the sites fall between those two values.)
The range is huge. A quarter of the sites have less than 5.5% brand and another quarter, more than 39.5%. By bracket, over the total:
| Brand share | Sites |
|---|---|
| Less than 10% | 47 |
| 10% to 25% | 29 |
| 25% to 50% | 36 |
| More than 50% | 19 |
So 19 of 131 sites (1 in 7) depend on their brand for more than half of the organic channel. For those companies, a drop in brand searches is felt across the whole channel even if the SEO team is doing its job well.
Is there a good ratio? I wouldn't use it as a target. A very well-known brand with little content will have a lot of brand traffic, a new site with a big blog will have little, and both can be fine. It works as a reference: if your site is at 60%, you know your SEO depends quite a bit on what the rest of marketing does.
Why does Search Console change the number from 18% to 41%?
Because Search Console hides infrequent queries for privacy reasons. On the typical site, 47.8% of clicks came from searches that don't show up in the report (median of 131 sites). Those clicks are in the property's total, but you can't tell whether they were branded or not.
That leaves two ways to calculate the percentage:
-
Brand divided by all the property's clicks, including hidden ones: 18.2%.
-
Brand divided only by clicks with a visible query: 41.2%.
The first figure is a floor and the second, a ceiling. Hidden queries tend to be rare and long, so most of them are probably non-branded and the real value is likely closer to 18%. But this data can't prove it.
In practice, two agencies can look at the same site, report 18% and 41%, and both be right. So when someone shows you a brand percentage, the first question is what they divided by. Long searches are the ones that get lost the most in that hole (I covered it in the long-tail keywords study).
What CTR do branded searches get?
Branded searches get few impressions and lots of clicks. Their median CTR was 15.4%, versus 0.5% for visible non-branded queries. On the typical site, the brand gets 22 times more clicks per impression, and its CTR was higher on 93% of the sites.
Part of the explanation is position: the brand's average position was 2.7 (median), while for everything else the site often shows up far from the first page. The other part is intent: whoever searches for your name has already chosen you and only needs to click.
That has an effect on reports. The average of a site with a lot of brand traffic looks great and that of a site with little brand traffic looks bad, even if the second one competes much better on generic searches. If you are going to compare click-through rates across sites or months, separate the brand first. More detail in what CTR is and how to improve it.
Does it change across countries or over time?
A little across regions and almost nothing over time.
In Latin America the brand weighs a bit more than in the US: 21.9% versus 17.9% (medians of 30 and 74 sites). In Latin America, 20% of the sites get more than half their clicks from the brand, versus 14% in the US, and the brand shows up higher (position 2.1 versus 3.4). The ranges overlap a lot, so it's an indicative difference. Also, the US has more hidden queries (52% of clicks versus 43%), which lowers the figure calculated over the total.
Between June and September 2026 the median went from 15.5% to 18.2% (129 sites), which looks like an increase. But site by site, the median difference was 0 points and the share rose in only 46% of the sites. In the same period total clicks fell 8% (median), with similar drops in brand (6%) and non-brand (7.5%). That's three months with possible seasonality, so there is no clear trend.
How do you measure branded traffic in Search Console?
With a query filter in the Performance report. This is the study's method, adapted for a single site:
-
In Search Console, go to Performance, choose web search and the last three months.
-
Add a query filter with "Custom (regex)" and paste your brand name with its usual forms, for example
acme|ac me|ac-me|acmé. Search Console uses RE2 syntax and is not case sensitive. -
Write down clicks, impressions, CTR and average position for that group. Those are your branded searches.
-
Switch the filter to "Doesn't match regex" with the same text. That group is your visible non-brand traffic.
-
Calculate the brand percentage twice: over the property's total clicks (no filter) and over the sum of the two visible groups. Report both or say which one you use.
Add your named products, common typos and the abbreviations your customers use. To find them, check the queries that land on your homepage: that's where the variants you would never have thought of show up.
Careful with Google Analytics 4. GA4 doesn't show organic search queries, so on its own it can't separate brand from non-brand (the Search Console integration brings the same data as above). What you can see in GA4 is traffic coming from assistants like ChatGPT, which is largely people an assistant mentioned your brand to. I explained it in ChatGPT traffic in Google Analytics.
Can you grow branded traffic?
Yes, but almost never from Google alone. The brand grows when more people know the company: advertising, PR, events, partnerships, service good enough to create word of mouth. That's why I wouldn't give the SEO team a brand target, just like I wouldn't give PR a rankings target.
What SEO is responsible for is making sure that demand isn't lost:
-
Every search with your name should land on the right page. If someone searches "acme login" and lands on the homepage, you lost the easiest click to serve.
-
Defend the result. If your competitors bid on your brand in Google Ads, check how many clicks they are taking from you.
-
Get named by AI assistants. More and more people learn about a brand because an assistant mentioned it and then search for it on Google. I wrote about that in AI already has favorite brands and in brand positioning.
Study methodology and limitations
The data comes from Google Search Console (web search, Search Analytics API) for company sites in the US, Latin America and other markets we have access to. The criteria were set before looking at the results: one property per domain, only domain or root-prefix properties, excluding unverified properties, our own sites, domains whose name is a generic search term (53) and names shorter than 4 letters (5). That left 204 candidate domains and 146 analyzed; sites with 1,000 or more impressions and 100 or more clicks in the main period were included: 131 sites (74 from the US, 30 from Latin America and 27 from other markets, based on the country with the most impressions).
The main period runs from September 6 to October 3, 2026, and the comparison period from June 7 to July 4, 2026 (28 days each, both after the num=100 change of September 2025 and after the impressions counting error that Search Console carried until April 2026). We counted as branded any query containing the domain name, allowing a space, period or hyphen between letters, with a case-insensitive RE2 regular expression. Per site we calculated the share of branded clicks over the total and over the visible, the share of impressions, branded and non-branded CTR, branded average position and the share of clicks with a hidden query; then medians and 25th and 75th percentiles across sites. The June versus September comparison uses the 129 sites that meet the threshold in both periods.
The limitations, no fine print:
-
The brand filter is an approximation based on the domain name: it misses abbreviations and brands different from the domain, so it falls short. Six sites ended up at 0% because of this and stayed in the sample; without them, the median rises to 21%.
-
Search Console hides infrequent queries (48% of clicks, median): the figure over the total is a floor and over the visible, a ceiling.
-
Traffic from AI Overviews and AI Mode is included in web search and can't be separated.
-
This is correlation, not causation, with two 28-day windows and possible seasonality.
-
It's a convenience sample and doesn't represent the market. The region is the country with the most impressions, not the company's headquarters.
Frequently asked questions about branded traffic
Is direct traffic branded traffic?
They're similar, but they're measured differently. Direct traffic comes from typing the URL or from a bookmark, and it has its own channel in GA4. The brand traffic in this study comes from Google searches that include the name. They tend to rise and fall together because the same awareness drives them.
Do branded searches convert better?
Guides usually say yes, because whoever searches for your name has already chosen you. Our study didn't measure conversion, so I don't have our own data on that. What we did measure is a CTR 22 times higher than non-branded searches on the typical site, which fits with more decided intent.
What I would do this week
Open the Performance report, build the filter from step 2 with every variant you can think of and calculate your two percentages: over the total and over the visible. It takes half an hour.
In your next organic report, put brand and non-brand in separate columns, with clicks and average position for each, and add a footnote saying what you divided by. If non-brand is stable or growing, the channel is contributing to the business. If all the growth comes from the brand, congratulate whoever ran the campaign (and talk to the SEO team about the other column).