Branded vs Non-Branded Search: The Split Agencies Hide

The report says organic traffic is up 41% quarter over quarter. Your front desk says the phone rings about the same. Both are telling the truth.

This is the most common way an SEO report lies without lying. It shows you total organic clicks. It does not show you that most of the growth came from people typing your business name — people your van, your yard signs, your referral network, and your last five years of good work sent to Google. The agency collected those clicks. It did not earn them.

Splitting the two takes about two minutes and one line of regex. Here is the split, the filter, and what the answer actually means.

What is the difference between branded and non-branded search traffic?

A branded search contains your company, service, or product name; a non-branded search doesn't. That is the whole definition. 'riverbend plumbing' is branded. 'emergency plumber austin' is not. Everything else in this post follows from that one line.

The distinction matters because the two buckets answer completely different questions. Branded queries measure how many people already know you. Non-branded queries measure how many strangers you took from a competitor. Only one of those is a search-visibility program's job.

Verdict, so nobody has to guess: non-branded clicks are the scoreboard. Branded clicks are worth having — you should absolutely own your own name — but they are a lagging indicator of everything else you do, not a measure of the SEO you're paying for.

Why does branded traffic make a bad SEO program look good?

Because you already rank #1 for your own name on day one, so branded clicks cost an agency exactly zero effort — and Google Search Console dumps them into the same 'total clicks' number as the clicks they had to fight for. Search Console groups data by Queries, Pages, Countries, Devices, Search appearance, and Dates. There is no branded/non-branded dimension anywhere in the report.

So the default view — and the default agency screenshot — is a single blended line going up and to the right. If your business grew for any reason at all (a busy season, a truck wrap, a five-star run on Google, a radio spot, word of mouth from a job well done), branded searches rise. The line goes up. The invoice keeps clearing.

That is why we treat the blended organic chart as a vanity metric. If you want the full list of what a report should actually contain, our Search Console guide for service businesses walks through the reports that matter and the ones that don't.

How do you split them in Search Console in two minutes?

Four clicks and one regex: open Performance → Search results, go to the Queries tab, click + Add filter → Query → Custom (regex), paste your brand pattern, and read the clicks. Then flip the same filter to Doesn't match regex. That second number is your non-branded traffic.

Google's own documentation for the Performance report confirms the mechanics: regex filtering uses RE2 syntax, default matching is a partial match, and matching is not case-sensitive by default — so you don't need to worry about capitalization. For multiple values, Google's instruction is to separate them with a vertical bar and wrap the whole thing in parentheses.

Here is a copy-pasteable pattern. Swap in your own name, the way people actually spell it wrong, and your bare domain:

• The branded regex: (riverbend|river bend|riverbnd|rivebend|riverbendplumbing)
• Branded clicks: apply it with Matches regex on the Queries tab
• Non-branded clicks: same pattern, switch to Doesn't match regex
• Include: legal name, DBA, common misspellings, your domain without the .com, founder name if customers search it
• Exclude: any word that is also a real service term — a plumber named 'Drain King' must not put 'drain' in the pattern, or every 'clogged drain' search lands in the branded bucket

That last rule is where most DIY splits break. If your brand name contains a service word, a city, or a common noun, anchor it to the full phrase instead of the fragment. A pattern that eats your own money keywords will tell you your SEO is failing when it isn't.

And know the limits before you build a dashboard on this. Google documents all of them on its data discrepancies page:

• Branded + non-branded will not equal your total. Google states that when you filter by query, the 'match' and 'does not match' totals may not add up to the unfiltered total, due to omitted anonymized queries and data truncation.
• Anonymized queries are dropped. Rare queries — searched a very small number of times — are omitted from the query table to protect user privacy. On a small local site, that long tail is a real slice of your non-branded volume, and it is invisible.
• The table caps at 1,000 rows. Long-tail rows get left out of the table but stay in the chart totals.
• Data lags 2–3 days, and daily views are labeled in Pacific Time — so a same-day number is never a real number.

None of that breaks the exercise. It just means you read the trend in non-branded clicks, not a to-the-decimal percentage. Anyone who reports your branded/non-branded split to one decimal place is telling you they don't know how the tool works.

What's a healthy non-branded share for a local service business?

There is no trustworthy current benchmark for a US local service business, and you should be suspicious of anyone who hands you one. The figure that circulates most traces back to a single Uberall analysis of 22 global brands with 48,000 locations plus more than 450,000 SMBs, with the data collected between August 2018 and August 2019. In the SMB slice of that dataset, roughly 19% of queries featured a branded term and 81% did not.

Treat that as a museum piece, not a target. The data is seven years old, it predates AI Overviews entirely, and 'SMBs' in a global location-data platform's customer base is not your three-truck HVAC company. Most posts quoting it are quoting each other, and the number gets mangled in transit — the same article's separate 42% figure is a consumer survey about how people discover global brands, not a share of anybody's search volume. Do not let the two be blended into one benchmark.

So stop hunting for a target percentage. Use your own baseline instead. It is a better measurement anyway, because a share can improve while the actual number of new customers falls:

• Track non-branded clicks in absolute numbers, not as a share. If branded traffic doubles, your non-branded share drops — even though nothing bad happened.
• Compare trailing 90 days to the prior 90 days. Month-to-month is noise; a service business has seasons.
• Segment non-branded by intent. 'emergency plumber near me' and 'how does a water heater work' are both non-branded, and only one of them buys anything.
• Tie it to booked calls. Non-branded clicks that produce no booked calls are a ranking problem you traded for a conversion problem.

The number to hold an agency to is non-branded clicks on commercial-intent queries, and the calls they produced. Our SEO ROI guide shows how to run that math backward from closed revenue.

What does it mean if branded is the only thing growing?

It means your SEO is not working and something else in your marketing is. If non-branded clicks are flat or down across two consecutive 90-day windows while branded climbs, the program has not won a single new stranger — it has been riding demand that someone else created.

One quarter is noise. Two consecutive 90-day windows is a verdict. That is also roughly the horizon we hold ourselves to: a channel producing no qualified leads in 90 days gets cut, not defended with a new deck.

There is one honest exception. On a brand-new domain or a rebuild, branded traffic usually moves first because it is the easiest thing to fix — you claim your name, your Business Profile, your knowledge panel. If your program is under six months old and branded is leading, that's expected. If it's been eighteen months and branded is still the only line moving, you are paying a retainer to be a scorekeeper for your own word of mouth.

Does branded growth mean your SEO is working, or your van is?

Branded demand is created almost entirely off-search, so in most service businesses the answer is the van. People do not type 'riverbend plumbing' because a blog post ranked. They type it because they saw the truck in a neighbor's driveway, got a postcard, heard your name at a barbecue, or worked with you three years ago and forgot the number.

The tell is timing. Match your branded-click line against the things that create name recall — the month you wrapped two new trucks, the run of five-star reviews, the direct-mail drop, the sponsorship. If branded jumps two weeks after a real-world push and non-branded doesn't budge, the SEO did not cause it.

There is a paid-side version of this exact question — whether you're buying branded clicks you'd have gotten free — and it's a different test with a different method. That one belongs to paid search, and it needs a real holdout, not a regex.

How should this change what you ask your agency for?

Ask for one number at the top of every report: non-branded organic clicks, with the exact regex used to define 'branded' printed underneath it. If they can't produce that in the time it takes to run this meeting, they have never looked.

Then hold the line on four things:

• The regex is disclosed and stable. A pattern that quietly changes month to month can manufacture any trend you want.
• Branded and non-branded are reported separately, always. Not on request. Not in an appendix.
• Non-branded clicks map to booked calls. Clicks that don't turn into calls are a website problem, and it's still their job to say so.
• No lock-in. If a program can't show non-branded movement in two quarters, you should be able to leave. Twelve-month contracts protect the agency, not you.

An agency that reports blended organic totals is either not measuring or not telling. Neither one is worth $2,500 a month. If you want a second pair of eyes on the split before you make that call, our SEO program starts by pulling exactly this report — and you can see what it costs on our pricing page before you talk to anyone.

Run the regex yourself first. If the non-branded line has been flat for two quarters, Get my free audit — we'll show you which queries you should be winning and aren't.