Brand vs. Non-Brand Search Traffic

TL;DR — A branded search includes your brand name (or a close misspelling, or a product only you sell). A non-branded search is everything else — generic “best running shoes” type queries from people who don’t know you yet. Splitting the two matters because branded traffic mostly reflects how well-known you already are, while non-branded traffic is the honest signal of whether your SEO work is actually earning new visitors.

What the two types are

Every organic visitor arrived by typing something into search. That something was either:

• Branded — it mentions you. “Nike shoes,” “gmial” (a misspelling of Gmail), or a product name only you have. The person already knew you and went looking.
• Non-branded — it doesn’t mention you. “waterproof running shoes,” “how to fix a slow website.” The person had a need but hadn’t picked a company yet, and foundA 302 (\"Found\") is a temporary redirect: it forwards users to a new URL while telling search engines the original URL should stay in the index. It's a weak canonicalization signal, not the zero-equity dead end of SEO folklore. you in the results.

Both are real visitors. But they mean very different things.

Why you’d bother splitting them

Imagine your boss sees “organic trafficVisitors from unpaid search results — it compounds without ad spend. is up 20% this quarter” and congratulates the SEO team. Sounds great — until you realize the company also ran a big TV campaign and got mentioned on a popular podcast that quarter. A lot of those extra visitors just Googled the brand name because they heard it somewhere. That’s branded traffic, and the SEO team didn’t create it.

The traffic that does tell you SEO is working is the non-branded traffic — new people, who’d never heard of you, finding you through search. If you never separate the two, a big jump in brand awareness can make SEO look like it’s crushing it when non-branded is actually flat, and a real non-branded decline can hide behind a branded bump.

The simple version of how to do it

• In Google Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results., Google now has a built-in “branded queries” filter that automatically splits the two for you (as long as your site is big enough and set up as a full-domain property).
• The DIY way is to make a list of your branded terms — your name, common misspellings, your product names — and filter your query and traffic reports to either include or exclude them.
• The rule for reporting it: show your boss two lines, not one. “Branded traffic” and “non-branded traffic” as separate numbers, so nobody mistakes brand demand for SEO results.

Want the enterprise version — the new Google filter’s limits, regex and Looker Studio methods, rank-level tagging, and how to report this to a CFO? Switch to the Advanced tab.

TL;DR — Segmenting organic into branded vs. non-branded is a reporting methodology problem, not a “which is better” debate. Branded demand at an enterprise comes largely from non-SEO sources (PR, ads, word of mouth, existing customers), so a blended organic number systematically flatters — or hides — what SEO actually did. Non-branded is the cleaner content/technical signal. Three ways to split it: Google’s native GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. branded queries filter (AI-assisted, not regex, launched Nov 2025, top-level properties only, no full historical backfill, not customizable), a maintained branded-term list applied via regex in GSC/GA4 or a Looker Studio dashboard, and rank-level tagging in Ahrefs Rank Tracker. Report the two as separate series to execs, pair with revenue, and don’t reflexively credit SEO for branded upticks — they can be brand-marketing or AI-citation driven.

What “branded” actually means (and why the line is fuzzy)

A branded query is one that names you. Google’s own definition is precise: it includes “your brand name (for example, Google), variations or misspellings of the brand name (for example, Gogle), and brand-related products or services: (for example, Gmail).” Non-branded is everything else — the generic, category, and problem-based queries where the searcher hasn’t decided on you.

The trap is treating this as a clean binary. It isn’t. A company named after a common word (“Apple,” “Square,” “Monday”) collides with generic usage. A product name that’s also a generic term (“Photoshop” as a verb, “Kleenex” for any tissue) blurs the edge. Google itself acknowledges the ambiguity — its classifier warns that “some queries may occasionally be misidentified.” Whatever method you use, budget for a gray zone and be consistent about how you resolve it, because a definition that drifts quarter to quarter makes your trend lines meaningless.

Why this matters more at enterprise scale

For a brand-new site with no awareness, almost everything is non-branded — there’s barely anything to split. The bigger and more established the brand, the more the problem bites:

• Large brands have branded demand baked in from non-SEO sources. TV, PR, an existing customer base, other marketing channels, conference talks, a founder’s podcast tour. None of it is the SEO team’s work, and all of it shows up in organic as branded search.
• So a blended headline number is mostly a brand-demand story. “Organic sessions up 20%” at a big company can be almost entirely branded — which means the SEO team’s real, incremental contribution is invisible inside it. This is exactly the point I made in Enterprise SEO Strategies for Maximum Growth: “a quick pie chart showing branded vs. unbranded traffic is likely to show that most organic visitors come from branded terms, and you should be more focused on unbranded terms.”
• Branded traffic isn’t bad — it’s just not yours to claim. As I put it in the same piece: “Branded traffic is a good thing. It’s high-quality and converts well, but you should be getting it even without SEO help.” The point of segmenting isn’t to dismiss branded traffic, it’s to stop it from masking the number that measures your work.

Non-branded growth is the cleaner signal. It reflects whether new people — who didn’t already know you — are finding you through search, which is the closest thing to a pure read on content relevance plus crawlingCrawling is how search engines use automated bots (like Googlebot and Bingbot) to discover URLs and download pages. A page has to be crawlable to be indexed, but crawling on its own isn't a ranking factor., indexingStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed., and technical health working as intended. It’s the same throughline as enterprise SEO ROIEnterprise SEO ROI is the financial return an organic-search program generates relative to its total cost — staff, tools, content, and agency fees. The formula is simple ((Revenue − Cost) ÷ Cost); measuring the revenue side cleanly at enterprise scale is the hard part.: the metrics you report have to isolate what SEO actually caused.

How Google now surfaces this natively

For years the only way to do this was manual. That changed in late 2025. Google added a branded queries filter to the Search Console PerformanceThe Google Search Console report that shows how your site actually performed in Google Search, built from real impressions and clicks. It reports four metrics — clicks, impressions, average CTR, and average position — and keeps the most recent 16 months of data. report — the feature is “designed to help analyze the queries driving traffic to your site by automatically differentiating between branded and non-branded queries.” It shipped a month after Query groups (Google’s AI-clustered similar-query feature) and was expanded to all eligible sites in March 2026.

Four things you have to understand before you rely on it:

1. It’s AI-assisted, not regex. Google is explicit: “The classification of branded versus non-branded is NOT based on a regular expression method of including or excluding keywords… It is determined by an internal, AI-assisted system.” It covers your brand name in all languages, typos, and queries that refer to a unique product without naming the brand. So it won’t match a hand-built regex, by design.
2. You can’t customize it. John Mueller confirmed (via Search Engine Journal coverage) that site owners can’t add their own branded terms or teach the AI: “At the moment, I’m not aware of plans to provide customization, but feedback in the tool is always welcome!” Asked directly whether you can educate the AI to treat specific terms as branded, he answered no, not at this time.
3. There’s no full historical backfill. Per Mueller, “there’s a point when the data starts being tracked, and you’ll see that in the report if you look far enough back.” You can’t retroactively split years of old data.
4. Eligibility is limited. “This is only available for top level properties (and not for URL path properties… or subdomain properties),” and only for “sites with a sufficient volume of queries and impressions.” Sub-properties and low-volume sites are out.

One more caveat worth stating to stakeholders: the filter “has no effect on how Google Search ranking works.” It’s an analysis lens, nothing more.

Bing has no equivalent. Bing Webmaster ToolsMicrosoft's free portal for monitoring and improving how a site appears in Bing search — the peer to Google Search Console, plus IndexNow instant indexing, richer backlink data, and keyword volumes. Because Bing's index also feeds Microsoft Copilot, it doubles as a window into AI-search visibility.’ Search Performance reportThe Google Search Console report that shows how your site actually performed in Google Search, built from real impressions and clicks. It reports four metrics — clicks, impressions, average CTR, and average position — and keeps the most recent 16 months of data. gives you the same raw building blocks — queries, impressions, clicks, CTR, position — but no native branded/non-branded split. On Bing data you’re still doing it manually downstream (Excel, Looker Studio, a regex filter). Don’t assume the two engines match here.

How to segment it yourself

The native filter is a great first pass, but a maintained internal method is still what you want when definitions have to be defensible and consistent across teams and over time. Here’s the practitioner stack, which is essentially what I described in Enterprise SEO Storytelling: “I typically split branded and non-branded terms with Looker Studio and a custom list of branded terms. You can use either GSC or Ahrefs data for it.”

1. Build and maintain a branded-term list. This is the foundation of every manual method. Include:

• Your exact brand name and any legal/DBA variants.
• Common misspellings and typos (think keyboard-adjacent slips and phonetic misspellings — “gogle,” “gmial”).
• Spacing and hyphenation variants (“wellsfargo,” “wells fargo”).
• Product and service names uniquely tied to you.
• For global brands, international spellings and transliterations.

Version it. Treat it like code — a definition that quietly changes breaks year-over-year comparison. This is the piece Google’s AI filter takes away from you, which is exactly why you keep your own for defensible reporting.

2. Apply it with regex in GSC and GA4. GSC’s Performance report has a “Filter by query” custom (regex) option, and GA4 lets you filter or build audiences/segments on the landing-page query. Match your branded terms to isolate branded traffic, then invert the match for non-branded. GA4 has no native branded/non-branded report at all, so regex against a maintained list is the only way to segment there — one reason your own list stays essential even after GSC’s native filter exists. (See the Scripts lens for ready-to-adapt regex and a Looker Studio field.)

Know the data’s blind spots before you report the split. Query-level GSC data — the table your regex filters run against, and the data the native filter classifies — omits anonymized queries: ones searched by too few people (roughly a few dozen users over a two-to-three-month window) to show individually without a privacy risk. Google’s own illustration makes the effect concrete: filter a query set to include a term and you might see 175 clicks; exclude the same term and you might see 275; the two sum to 450 rows — while the report’s chart total, which does count the anonymized clicks, shows 550. Branded clicks plus non-branded clicks will not add up to your site’s total organic clicks, and that gap is expected privacy filtering, not a broken term list or a miscount. Report the two shares against the classified total (the sum of itemized rows), not the site’s unfiltered total, and say so in your methodology note. Separately, the Performance report’s UI caps exports at 1,000 rows; the Search Analytics APIA set of REST APIs that let you programmatically read and manage Google Search Console data for properties you've verified — Performance data, URL index status, sitemaps, and properties — authorized with OAuth 2.0. returns up to 25,000 rows per request (paginate with startRow toward its 50,000-rows-per-day cap), so a large site segmenting its full query list needs the API or the BigQuery export, not the on-screen table, to get complete coverage.

3. Build it once in Looker Studio. Rather than re-filtering by hand every month, wire GSC (or Ahrefs) data into a Looker Studio dashboard with a calculated field that tags each query branded or non-branded off your list. Now the split is a saved view, not a chore, and it pulls from a source you control.

4. Segment at the rankings level with tagging. Traffic isn’t the only thing worth splitting. Ahrefs Rank Tracker lets you tag keywords so you can see branded vs. unbranded rankings and visibility, not just clicks. As I noted in the same article, “Rank Tracker allows for custom segmentation via a flexible tagging system” — you can tag branded/unbranded, product lines, business units, authors, whatever your reporting needs.

5. Reach for BigQuery only when you need long history. GSC’s UI retains roughly 16 months. If you need multi-year trend lines, bulk-export GSC data to BigQuery and run your regex there. It’s a heavier, engineering-level lift that most in-house teams won’t need — treat it as the “I need more than ~16 months of history” option, not a default. (See the GSC BigQuery exportGSC BigQuery export (bulk data export) is a Google Search Console feature that schedules a daily, unsampled export of a property's Performance data into a Google Cloud BigQuery project — bypassing the UI's ~1,000-row export cap and its rolling ~16-month retention window. It doesn't backfill and still excludes anonymized queries at the query level. deep dive.)

TIP Review ambiguous brand variants instead of trusting one regex forever

Brand classification is a maintained rule set. This result surfaces a query whose automated label disagrees with the reviewed label so the dictionary can be corrected.

Split and review a GSC export with my free Brand vs Non-Brand Splitter Free

1. Load a Search Console query export and seed the known brand, product, and misspelling variants.
2. Review disagreements and high-impact candidate variants instead of bulk-accepting every match.
3. Update the maintained rule set, rerun the same export, and retain the review notes with reporting.

Tap to enlarge

The result presents a brand-classification review case where an automated label and the reviewed decision disagree, making the query a candidate for updating the maintained brand-term rules.

Reporting it to executives

The segmentation is only worth doing if the report lands. Three rules:

• Two lines, never one blended number. Report branded and non-branded as separate series. The whole point evaporates the moment they’re summed back together on the slide.
• Pair it with revenue or competitive framing. Raw traffic counts don’t move executives. Tie non-branded growth to pipeline, revenue, or share of voice — the same discipline that runs through enterprise SEO reportingEnterprise SEO metrics are the KPIs used to measure and report SEO at large, complex organizations — tied to business outcomes like revenue and pipeline, not vanity metrics like raw rankings, and organized by audience from the C-suite down to engineering. and ROIEnterprise SEO ROI is the financial return an organic-search program generates relative to its total cost — staff, tools, content, and agency fees. The formula is simple ((Revenue − Cost) ÷ Cost); measuring the revenue side cleanly at enterprise scale is the hard part..
• Keep the visualization dead simple. My go-to is a plain pie chart showing branded vs. unbranded — it makes the “most of your organic is brand demand, so let’s push on non-branded” argument in one glance. A memorable chart beats a precise one nobody reads.

The caveat nobody puts on the slide: a branded uptick isn’t automatically a win

Rising branded search feels like success, but it doesn’t always trace to your work — and in 2026 it increasingly traces to something you can’t claim. Branded/direct traffic can climb because of a PR hit, an ad campaign, a conference talk — or because an AI OverviewAI Overviews are the AI-generated summary box Google shows above or within its regular search results, written by Gemini models from pages retrieved out of Google's normal Search index. It's a Search feature, not a separate platform or index. or AI-search answer cited you, building recall without a click, so the person searches your brand days later and lands as Direct. Before you credit SEO for a branded bump, check whether non-branded is moving too. If branded is up while non-branded is flat, that’s a brand-awareness story, not proof the SEO strategy is working — and it’s a reason segmentation is getting more important as AI answers reshape the funnel, not less. (This connects to AI traffic attributionAI traffic attribution is the practice of correctly identifying and measuring website visits that come from AI tools — ChatGPT, Perplexity, Gemini, Claude, AI Overviews, and AI browsers. It's hard because many of those tools strip the referrer header, so the visits land in your analytics as Direct traffic with no source. and the broader dark-funnel problem.)

AI summary

A condensed take on the Advanced version:

• Branded query = includes your brand name, its variants/misspellings, or a product uniquely tied to you. Non-branded = everything else. The line is fuzzy for brands named after common words; expect a gray zone.
• Why segment: at enterprise scale, branded demand comes mostly from non-SEO sources (PR, ads, word of mouth, existing customers), so a blended “organic up 20%” number flatters — or hides — what SEO actually did. Non-branded growth is the cleaner content/technical signal.
• Google’s native GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. branded queries filter (launched Nov 2025, expanded Mar 2026) auto-splits the two. It’s AI-assisted, not regex; not customizable; has no full historical backfill; and works only on top-level properties with sufficient volume. It doesn’t affect ranking.
• Bing has no equivalent — you segment Bing data manually.
• DIY methods: maintain a versioned branded-term list (name + misspellings + product names + international variants); apply it via regex in GSC/GA4 (GA4 has no native split, so this is the only option there); build it once in a Looker Studio dashboard off GSC/Ahrefs data; tag branded/unbranded at the rankings level in Ahrefs Rank Tracker; use BigQuery only for history beyond GSC’s ~16 months.
• Report branded and non-branded as two separate lines, never blended; pair with revenue or share of voice; keep the visual simple (a pie chart makes the point).
• Data mechanics to know: Search ConsoleGoogle's free tool for monitoring crawling, indexing, and search performance. omits anonymized queries (too few searchers to show without a privacy risk) from every query-level table and filtered view but still counts them in chart totals — so branded + non-branded clicks won’t sum to your unfiltered organic total; that’s expected, not a broken term list. Report shares against the classified total instead. The UI also caps exports at 1,000 rows; use the Search Analytics APIA set of REST APIs that let you programmatically read and manage Google Search Console data for properties you've verified — Performance data, URL index status, sitemaps, and properties — authorized with OAuth 2.0. or BigQuery for full coverage on large sites.
• Caveat: a branded uptick can be PR-, ad-, or AI-citation-driven, not an SEO win — check non-branded before claiming credit.
• Don’t fabricate benchmarks. There’s no universal “good” branded/non-branded ratio; it depends on brand size and maturity.

Official documentation

Primary-source documentation on segmenting queries and reading performance data.

Google

• Introducing the branded queries filter in Search Console — the launch post: what the filter is, how the AI classification works, and its limitations (Nov 2025, updated Mar 2026 to note availability to all eligible sites).
• Introducing Query groups in Search Console Insights — the AI-clustered-similar-queries feature that shipped one month before the branded filter; related but distinct.
• Search Console Performance report — clicks, impressions, CTR, position, and the “Filter by query” (including regex) control you use for manual segmentation.
• A deep dive into Search Console performance data filtering and limits — the primary source on anonymized queries (why branded + non-branded won’t sum to your total organic clicks) and the UI’s 1,000-row export cap versus the API’s per-day row limit.
• Bulk data export to BigQuery — how to export raw Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. data for longer history than the UI retains, so you can run your own branded/non-branded regex at scale.

Quotes from the source

On-the-record statements from Google. Each Google-blog link is a deep link that jumps to the quoted passage on the source page.

Google — what the branded queries filter is and how it classifies

• “This new feature is designed to help analyze the queries driving traffic to your site by automatically differentiating between branded and non-branded queries.” — Google Search Central Blog. Jump to quote
• “A branded query is a query that includes your brand name (for example, Google), variations or misspellings of the brand name (for example, Gogle), and brand-related products or services: (for example, Gmail).” Jump to quote
• “Branded queries typically lead to higher-ranking pages from your site and result in higher click-through rates, whereas non-branded queries offer organic growth, as they show how new users find your content without any initial intent to go to your site.” Jump to quote
• “The classification of branded versus non-branded is NOT based on a regular expression method of including or excluding keywords, which is already available in the ‘Filter by query’ section. It is determined by an internal, AI-assisted system.” Jump to quote

Google — the limitations to state to stakeholders

• “Due to the dynamic and contextual nature of brand classification, some queries may occasionally be misidentified. This filter is designed to make it easier for you to segment and analyze your data within Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results.; it has no effect on how Google Search ranking works.” Jump to quote
• “This is only available for top level properties (and not for URL path properties such as https://example.com/path or subdomain properties such as developers.google.com). This is only available for sites with a sufficient volume of queries and impressions for our signals.” Jump to quote

Google — why branded + non-branded won’t sum to your total organic clicks

• “Some queries (called anonymized queries) are not included in Search ConsoleGoogle's free tool for monitoring crawling, indexing, and search performance. data to protect the privacy of the user making the query. Anonymized queries are those that aren’t issued by more than a few dozen users over a two-to-three month period.” — Google Search Central Blog, “A deep dive into Search Console performanceThe Google Search Console report that shows how your site actually performed in Google Search, built from real impressions and clicks. It reports four metrics — clicks, impressions, average CTR, and average position — and keeps the most recent 16 months of data. data filtering and limits.” Jump to quote
• “While the actual anonymized queries are always omitted from the tables, they are included in chart totals, unless you filter by query.” — Google Search Central Blog, same post. Jump to quote

John Mueller, Google (relayed via Search Engine Journal’s coverage of Google’s answers — not a primary Google transcript)

• On adding your own branded terms: “At the moment, I’m not aware of plans to provide customization, but feedback in the tool is always welcome!”
• On historical data: “Yes, there’s a point when the data starts being tracked, and you’ll see that in the report if you look far enough back. As a smaller site starts to grow, they might also start seeing this data at some point.” Read the coverage

Brand vs. non-brand segmentation checklist

Run this before you present a branded/non-branded number to anyone.

Define it defensibly

• Branded-term list built: exact brand name + legal/DBA variants
• Common misspellings and typos included (keyboard-adjacent + phonetic)
• Spacing/hyphenation variants included (“wells fargo” / “wellsfargo”)
• Product/service names uniquely tied to the brand included
• International spellings/transliterations included (global brands)
• The list is versioned and the change history is tracked
• A rule exists for ambiguous terms (brand-name-is-a-common-word cases)

Pick and apply a method

• GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. native branded queries filter checked (top-level property, enough volume?)
• Regex filter applied in GSC “Filter by query” for the manual view
• GA4 segmented with the same regex (GA4 has no native split)
• Looker Studio dashboard built with a branded/non-branded calculated field
• Rank-level tags applied in Ahrefs Rank Tracker (branded vs. unbranded)
• BigQuery export set up only if you need history beyond ~16 months
• API or BigQuery export used (not the 1,000-row UI table) for a full-coverage query list on large sites

Report it right

• Branded and non-branded shown as two separate lines, never blended
• Paired with revenue, pipeline, or share of voice — not raw traffic alone
• Non-branded checked against branded before crediting SEO for any uptick
• No fabricated benchmark ratio quoted (there’s no universal “good” split)
• Branded + non-branded reported against the classified total, not the site’s unfiltered organic total — anonymized queries mean the two won’t reconcile

The mental models

1. Branded = demand you already have; non-branded = demand you’re earning. Branded search mostly reflects awareness built elsewhere (PR, ads, word of mouth, existing customers). Non-branded reflects new people finding you through search. When you report SEO’s impact, non-branded is the number that isolates your work.

2. Blended numbers lie by omission. One organic total can rise on brand demand alone, and it can hide a non-branded decline behind a branded bump. The split exists to stop a headline number from telling a story that isn’t true. The bigger the brand, the more this bites.

3. Native filter vs. maintained list — two different jobs. Google’s GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. filter is a fast, zero-setup first pass (AI-assisted, not customizable, not backfilled). Your own versioned term list is what you use when the definition has to be defensible and consistent across teams and over time. Use both; don’t assume they agree.

4. Segment traffic and rankings. Traffic-level segmentation (GSC/GA4/Looker Studio) answers “who’s showing up.” Rank-level tagging (Rank Tracker) answers “where we’re visible.” Different questions — report both when the audience needs the fuller picture.

5. A branded uptick is a hypothesis, not a win. Before crediting SEO, check non-branded. Branded up + non-branded flat = brand-awareness or AI-citation story, not proof your content/technical work moved the needle.

Brand vs. non-brand — cheat sheet

What counts as branded

• Brand name · legal/DBA variants · misspellings & typos · spacing/hyphenation variants · uniquely-yours product/service names · international spellings

The three segmentation methods

GSC native filter — the fine print

• AI-assisted, not regex · not customizable · no full historical backfill
• Top-level properties only (no path/subdomain properties) · needs sufficient query volume · no effect on ranking

Reporting rules

• Two separate lines, never one blended number
• Pair with revenue / pipeline / share of voice, not raw traffic
• Check non-branded before crediting SEO for a branded uptick

Don’t

• Don’t quote a universal “good ratio” — there isn’t one; it depends on brand size/maturity/market
• Don’t assume GSC’s filter and your regex list agree — they won’t, by design
• Don’t credit SEO for branded growth automatically (could be PR / ads / AI citationsAn AI citation is the visible source link an AI answer engine shows next to its generated text — the clickable reference that credits the web page it used. A citation's presence is a separate thing from whether the cited page actually supports the statement, and from being retrieved (read behind the scenes) or merely mentioned (named without a link); citation is driven more by brand mentions and being retrievable than by traditional ranking.)
• Don’t expect branded + non-branded clicks to sum to your total organic number — Search ConsoleGoogle's free tool for monitoring crawling, indexing, and search performance. omits anonymized queries from every query-level view

Which segmentation method should I use?

Start at the top and follow the first branch that fits.

Q1 — Do you just need a fast look, right now, with zero setup? → Use GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results.’s native branded queries filter — if your site is a top-level property with enough query volume. Good enough for a quick internal read. Stop here unless the definition has to be defensible.

Q2 — Does the definition have to be consistent and defensible across teams / over time (e.g., exec reporting, cross-team KPIs)? → Build and maintain your own versioned branded-term list. Google’s AI filter isn’t customizable and “some queries may occasionally be misidentified,” so you can’t guarantee it matches a definition your CMO signed off on.

Q3 — Are you segmenting GA4 (not just GSC)? → You have no native split — apply your term list via regex in GA4. This is the only way there.

Q4 — Do you report this every month / to a recurring audience? → Wire it into a Looker Studio dashboard with a branded/non-branded calculated field off your list (GSC or Ahrefs data), so it’s a saved view, not a monthly chore.

Q5 — Do you need the split at the rankings/visibility level, not just traffic? → Tag keywords in Ahrefs Rank Tracker (branded vs. unbranded) alongside the traffic-level view.

Q6 — Do you need more than ~16 months of history? → Export GSC data to BigQuery and run your regex there. Heavier lift — only if the UI’s retention window isn’t enough.

In practice: most enterprise teams run the GSC native filter and a maintained Looker Studio view side by side — the filter for speed, the list for defensibility.

Regex and formulas for branded/non-branded

The native GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. filter is AI-assisted and can’t be customized. Everywhere else — GA4, Looker Studio, BigQuery, Excel — you split with your own list. Here’s the reusable pattern. Escape regex metacharacters in brand names (dots, hyphens) and remember GSC/GA4’s regex filters run on RE2, which is case-sensitive by default — RE2’s own syntax reference lists its i flag as “case-insensitive (default false),” so you need the explicit (?i) prefix used below or a misspelled-case variant will slip through unmatched. RE2 also doesn’t support lookaheads or backreferences, so keep the pattern to plain alternation like the examples here.

Match branded queries (adapt the alternation to your term list)

(?i)brandname|brand name|brandnaem|bran dname|productname

Non-branded = everything the branded pattern does not match. In GSC’s “Filter by query,” set the custom (regex) filter to “Doesn’t match regex” with the same pattern. In GA4, invert the match in your segment/exploration filter.

GA4 exploration / segment — filter the landing-page query (or use a GSC-linked report) with the same regex, once to include (branded) and once to exclude (non-branded). GA4 has no native branded/non-branded dimension, so this is the only route.

Looker Studio calculated field — tag each row so you can pivot on it:

CASE
  WHEN REGEXP_MATCH(LOWER(Query), '(?i)brandname|brand name|brandnaem|productname')
  THEN 'Branded'
  ELSE 'Non-branded'
END

BigQuery (GSC bulk exportGSC BigQuery export (bulk data export) is a Google Search Console feature that schedules a daily, unsampled export of a property's Performance data into a Google Cloud BigQuery project — bypassing the UI's ~1,000-row export cap and its rolling ~16-month retention window. It doesn't backfill and still excludes anonymized queries at the query level.) — same idea, run over the full history the UI won’t show:

SELECT
  CASE
    WHEN REGEXP_CONTAINS(LOWER(query), r'brandname|brand name|brandnaem|productname')
    THEN 'Branded'
    ELSE 'Non-branded'
  END AS segment,
  SUM(clicks)      AS clicks,
  SUM(impressions) AS impressions
FROM `your_project.searchconsole.searchdata_site_impression`
WHERE query IS NOT NULL
GROUP BY segment;

Excel / Sheets — if you exported to a spreadsheet, a per-row flag:

=IF(REGEXMATCH(LOWER(A2), "brandname|brand name|brandnaem|productname"), "Branded", "Non-branded")

Keep the alternation list in one place (a named range, a variable, a tab) and reuse it across all four so every surface splits identically. That single shared list is your definition — version it.

Tools for segmenting brand vs. non-brand

• Google Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. — the native branded queries filter (fast, AI-assisted first pass on top-level properties) plus the manual “Filter by query” regex control for your own list. The Performance reportThe Google Search Console report that shows how your site actually performed in Google Search, built from real impressions and clicks. It reports four metrics — clicks, impressions, average CTR, and average position — and keeps the most recent 16 months of data. is the raw data source either way.
• GA4 — no native split; apply your branded-term regex in segments/explorations. Pairs GSC’s pre-click query data with on-site conversions.
• Bing Webmaster ToolsMicrosoft's free portal for monitoring and improving how a site appears in Bing search — the peer to Google Search Console, plus IndexNow instant indexing, richer backlink data, and keyword volumes. Because Bing's index also feeds Microsoft Copilot, it doubles as a window into AI-search visibility. — same raw query/impression/click/CTR data as GSC, but no native branded/non-branded split — you segment downstream.
• Ahrefs Rank Tracker — tag keywords branded vs. unbranded for a rank-level and visibility view, not just traffic. Ahrefs data can also feed the Looker Studio split.
• Looker Studio — build the branded/non-branded calculated field once off GSC or Ahrefs data so recurring reporting is a saved view, not a manual re-filter.
• BigQuery — for GSC data beyond the ~16-month UI window; run your regex over the bulk export. (See the GSC BigQuery exportGSC BigQuery export (bulk data export) is a Google Search Console feature that schedules a daily, unsampled export of a property's Performance data into a Google Cloud BigQuery project — bypassing the UI's ~1,000-row export cap and its rolling ~16-month retention window. It doesn't backfill and still excludes anonymized queries at the query level. deep dive.)

Brand/non-brand reporting mistakes

Reporting one blended organic total

Why it fails: brand demand from ads, PR, existing customers, or AI citationsAn AI citation is the visible source link an AI answer engine shows next to its generated text — the clickable reference that credits the web page it used. A citation's presence is a separate thing from whether the cited page actually supports the statement, and from being retrieved (read behind the scenes) or merely mentioned (named without a link); citation is driven more by brand mentions and being retrievable than by traditional ranking. can lift the total while non-branded discovery stays flat or declines.

Do instead: show branded and non-branded as separate series and connect each to conversions or revenue.

Treating every product name as branded

Why it fails: some product names are generic or shared with competitors, so the rule can move category demand into the branded bucket.

Do instead: document whether a term uniquely identifies the company and keep an explicit gray-zone rule for ambiguous names.

Changing the term list without versioning it

Why it fails: the apparent trend can move because the classification changed, not because search demand changed.

Do instead: version the term list, record additions and removals, and restate or annotate historical comparisons when the definition changes materially.

Treating Google’s native filter as ground truth

Why it fails: Google’s AI-assisted classifier is not customizable and Google acknowledges that some queries may be misidentified.

Do instead: use the native filter for a fast read and a maintained internal list for repeatable executive reporting.

Publishing a universal target ratio

Why it fails: brand maturity, category, market, and campaign mix determine the split. A borrowed percentage creates a target with no causal meaning.

Do instead: establish a company baseline and judge movement alongside revenue, share of voice, and known brand campaigns.

Common brand/non-brand segmentation issues

The native GSC branded filter is missing

Symptom: the Performance reportThe Google Search Console report that shows how your site actually performed in Google Search, built from real impressions and clicks. It reports four metrics — clicks, impressions, average CTR, and average position — and keeps the most recent 16 months of data. has no branded-query control.

Likely cause: the property is a URL-prefix path/subdomain property or the site does not have sufficient query and impression volume.

Fix and confirmation: check the top-level property and its eligibility. If the filter remains unavailable, apply the maintained branded-term regex to exported GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. data and confirm branded queries are tagged in the output.

A common-word brand pulls in unrelated searches

Symptom: the branded bucket contains generic queries whose context does not refer to the company.

Likely cause: a broad exact-name pattern matches every use of a word such as a normal noun or day of the week.

Fix and confirmation: add qualifying phrases or exclusions for known generic contexts, move unresolved cases into a documented gray-zone review, and inspect a sample of newly classified queries.

GSC, GA4, and the dashboard show different splits

Symptom: branded share changes materially depending on the reporting surface.

Likely cause: Google’s native classifier, regex lists, dimensions, dates, and data retention do not match across tools.

Fix and confirmation: align date ranges and filters, use one versioned term list for manual methods, and label the native GSC view as a separate methodology rather than forcing it to match.

A branded jump is mistaken for an SEO win

Symptom: branded traffic rises after a campaign or citation spike while non-branded traffic and visibility stay flat.

Likely cause: the report attributes existing or newly created brand demand to organic optimization simply because the final visit came through search.

Fix and confirmation: annotate PR, paid, launch, and AI-citation events; compare non-branded clicks, conversions, and visibility before assigning credit.

Branded + non-branded clicks don’t add up to total organic clicks

Symptom: summing the branded-filtered and non-branded-filtered click counts (from a regex filter or the native GSC filter) comes out lower than the site’s total organic clicks shown elsewhere in Search ConsoleGoogle's free tool for monitoring crawling, indexing, and search performance..

Likely cause: Search Console omits anonymized queries — ones with too few searchers to show individually without a privacy risk — from every query-level table and filtered view, while still counting their clicks in chart totals. An include-filter and its exclude-filter counterpart, run on the same query field, both leave those clicks out.

Fix and confirmation: treat the gap as expected privacy filtering, not a broken term list. Report branded/non-branded shares against the classified total (the sum of itemized rows), not the unfiltered site total, and note the methodology when you present it.

Metrics for brand vs. non-brand search

Non-branded organic clicks and conversions

• Metric: clicks and downstream conversions from queries classified as non-branded.
• What it tells you: whether new search demand is finding the site and producing useful outcomes without relying on prior brand awareness.
• How to pull it: apply the versioned term-list exclusion to GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. query data, then join or compare the resulting landing pages with GA4 conversions.
• Benchmark / realistic range: there is no universal target; establish the site’s own baseline by market, device, and season, then compare like-for-like periods.
• Cadence: monthly for operating reviews and quarterly for strategy, with campaign and algorithm-update annotations.

Branded and non-branded share of organic clicks

• Metric: each segment’s clicks divided by classified organic clicks.
• What it tells you: whether topline growth is being driven by existing brand demand or by broader search discovery.
• How to pull it: calculate both shares from the same GSC export and the same versioned classification rule; keep unclassified/gray-zone queries visible.
• Benchmark / realistic range: depends on brand maturity, category, and campaign mix. Use an internal historical range rather than a borrowed industry ratio.
• Watch out: compute both shares against the classified total (itemized rows), not the site’s unfiltered organic total — Search ConsoleGoogle's free tool for monitoring crawling, indexing, and search performance. omits anonymized queries from every query-level view, so branded + non-branded won’t sum to the unfiltered number.
• Cadence: monthly, and whenever the term list changes or a major brand campaign launches.

Non-branded search visibility by topic

• Metric: impressions, ranking distribution, or share of voice for tagged non-branded keyword groups.
• What it tells you: whether the site is earning category visibility before a searcher already knows the brand, including cases where clicks lag.
• How to pull it: tag keywords in a rank tracker and pair the view with non-branded GSC impressions by query group or landing-page topic.
• Benchmark / realistic range: set a baseline against the site’s own tracked topics and named competitors; no single visibility percentage fits every market.
• Cadence: weekly for monitoring and monthly for stakeholder reporting.

Resources worth your time

My related writing

• Enterprise SEO Storytelling: Metrics, Reports & Dashboards — has a dedicated “Brand vs. non-brand breakdown” section; the source for my Looker Studio + custom-list method and Rank Tracker tagging.
• Enterprise SEO Strategies for Maximum Growth — the “most organic visitors come from branded terms, focus on unbranded” pie-chart framing, and why branded traffic is “a good thing” you should get even without SEO.
• Enterprise SEO Challenges & Mistakes — building the business case for SEO work when brand demand muddies the topline.
• The Realities of In-House SEO — reporting to stakeholders and aligning SEO metrics with company goals.

My speaking

• Enterprise SEO Chaos (SMX) — including the “KPIs and Fuzzy Math” reality of enterprise measurement, of which the brand/non-brand split is a core example.

From around the industry

• Introducing the branded queries filter in Search Console (Google Search Central) — the primary source for the native filter, its AI classification, and its limits.
• Google Answers Questions About Search Console’s Branded Queries Filter (Search Engine Journal) — John Mueller’s follow-up answers on customization, eligibility, and historical data.
• How GSC’s branded query filter changes SEO reporting and analysis (Search Engine Land) — why manual regex was always imperfect (character limits, language variants, misspellings, no shared standard) and why non-branded is the clearest discovery proxy.
• How to Segment Branded vs Non-Branded Traffic for SEO (SEOmonitor) — the most thorough BigQuery + Apps Script + Looker Studio walkthrough of the manual approach, and the source for GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results.’s ~16-month UI retention detail.
• Branded vs Non-Branded Traffic: What’s More Important for SEO? (seoClarity) — definitions and the “does not contain brand name and variations” filter approach.
• Branded vs. Non-Branded Traffic: What’s the Difference? (Corey Morris, Search Engine Journal) — the case for splitting brand vs. non-brand across rankings, impressions, traffic, and conversions in reporting.
• Branded Search vs. Non-Branded Search (Mateusz Makosiewicz, Ahrefs) — a solid conceptual primer on the distinction (note: this one is a colleague’s, not mine).

Stats worth citing

A deliberately short list — most of the specific percentages floating around this topic (e.g., “45.7% of searches are branded,” “branded converts 2–3× better”) trace to no identifiable primary source, so I won’t repeat them. What’s actually verifiable:

• There is no universal “good” branded/non-branded ratio. It depends on brand size, maturity, and market — a new site is almost all non-branded; an established brand skews heavily branded. Any fixed benchmark you see quoted is almost certainly unsourced. If you want a number, pull it from your own GSCA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results./Ahrefs data, freshly, not from recycled competitor content.
• Google’s native split is AI-assisted, not regex, and imperfect by admission — “some queries may occasionally be misidentified” — so treat even the official numbers as approximate, not ground truth. Source
• GSC’s UI retains ~16 months of data — the reason multi-year branded/non-branded trend lines require a BigQuery export. Source

Test yourself: Brand vs. Non-Brand Search Traffic

Five quick questions on segmenting organic search and reporting it. Pick an answer for each, then check.

Brand vs. non-brand search traffic

Segmenting organic search performance into branded queries (your brand name, its variants and misspellings, and products uniquely tied to you) versus non-branded queries (everything else). Splitting the two stops brand demand from masking the SEO-driven growth in your reporting.

Related: Organic traffic, Enterprise SEO Metrics, Enterprise SEO Reporting, Performance report

Brand vs. non-brand search traffic

Brand vs. non-brand search traffic is the practice of splitting organic search performance into two segments so you can tell brand demand apart from SEO-driven demand. Branded queries include your brand name, its variations or misspellings, and products or services uniquely associated with you (Google’s own example is “Gmail” for google.com). Non-branded queries are everything else — generic, category, and problem-based searches from people who haven’t decided on you yet.

The reason to segment matters more at enterprise scale than anywhere else. A large, established brand already has significant branded search volume baked in from PR, ads, word of mouth, and its existing customer base — none of which the SEO team created. So a blended “organic sessions up 20%” headline can be almost entirely a branded-demand story with no SEO cause behind it, and can also hide a real non-branded decline. Non-branded growth is the cleaner signal of whether your content and technical SEOTechnical SEO is the practice of making a site easy for search engines to crawl, render, index, and (now) be eligible for AI answers. It's the foundation that lets your content and links rank — not a ranking trick of its own. are actually earning new visibility.

You can segment three ways: Google’s native branded queries filter in Search ConsoleA free Google service that reports how a site performs in Google Search and surfaces problems with how Google crawls, indexes, and serves it. It's first-party data straight from Google — but you don't need it to appear in results. (AI-assisted, launched November 2025, not customizable), manual regex or a maintained branded-term list applied in GSC/GA4 or a Looker Studio dashboard, and rank-level tagging in a tool like Ahrefs Rank Tracker. The right reporting discipline is to show branded and non-branded as two separate lines to executives rather than one blended organic number.

Related: Organic traffic, Enterprise SEO Metrics, Enterprise SEO Reporting, Performance report

Changelog