Branded vs non-branded campaigns: why they are separated
Separating branded and non-branded campaigns means that traffic from searches with your company name is run in a separate campaign from that of people who do not know you yet. The people who search for you by name have already decided. When their conversions sit in one campaign with those of strangers, the average figure looks great and means nothing. Brand queries buy up the weak performance of everything else. The separation is done with a separate campaign for the brand and negative keywords for it in the others. After that each campaign is read on its own: cost per conversion, ROAS and volume. It is the simplest change that makes your report honest, and usually the first one I make in an account I take over.
My name is Hristiyan Panov — a performance marketing consultant with over 5 years’ experience and over €2 000 000 advertising budget managed for over 90 companies in Google and Meta. Below is why the separation changes decisions and how it is done.
What exactly does mixing them distort?
Branded searches have a low cost per click and a high conversion rate. Mixed with generic ones, they pull down the average cost per result and push up the average return.
The consequence is not cosmetic. Budget decisions are made on that average figure — and the money goes to campaigns that look profitable only because they bring people who would have bought anyway.
The distortion shows in the search terms report. You open the non-branded campaign, filter by the company name and the domain, and look at what share of its conversions comes from there. When one group or campaign has a conversion rate several times above the rest, the cause is almost always this, not better copy or smarter bidding.
The second sign is dependence on other channels. When an email campaign, a TV ad or a strong organic period starts, branded searches rise and the overall campaign “improves” with them, without anything in it being touched. That way an ad that has not brought in a single new customer credits itself with demand created elsewhere, and gets more budget because of it.
How is it separated in practice?
- A separate campaign for the branded terms only, with exact and phrase match.
- The same words as excluded in all other campaigns, including Performance Max.
- A separate budget and separate goals for the two.
Without the second step the separation exists only on paper — the generic campaigns keep collecting the branded traffic.
The order matters. First the list of branded terms is compiled: the company name, its misspellings, the domain and the names of your own product lines. Then the list goes in as a shared list of negative keywords in all non-branded campaigns. Only then is the branded campaign launched. The reverse order leaves days of overlap.
The type of exclusion matters too. Exact match catches only the spelling you entered, so the core of the name goes in as a phrase exclusion, and the Latin and Cyrillic variants are added separately. In Performance Max the exclusion goes through the brand exclusions at campaign or account level.
In the first days after launch the search terms report is opened again. If your name still appears there, the list is not complete.
Is it worth paying for your own brand at all?
It depends on three things: whether a competitor is bidding on your name, whether your organic result is in first position, and what the click costs.
When a competitor sits above you on your own name, the branded campaign is protection and is usually cheap. When there is no competition and the organic result dominates, part of the spend is for traffic you would have received for free.
The way to find out is a test: pause the branded campaign for two weeks and see what happens to the total number of orders, not to the figures inside the campaign.
What changes in the report after the separation?
The overall return of the non-branded campaigns will fall. This is not a deterioration — it is the first time you see the real cost of a new customer.
That is why the separation has to be explained before it is made. Otherwise the next report looks like a collapse.
It helps to say in advance exactly what will move. The cost per conversion in the non-branded campaigns goes up, their return falls, and the branded campaign comes out with a low cost and a high return. The total for the account stays where it was — the same conversions are simply redistributed between two rows in the table.
In practice this means three things: the date of the separation is recorded as a note in the account, period-over-period comparisons for the non-branded campaigns start from it, and the whole account is judged by MER, which is not affected by moving traffic between campaigns.
The old targets also lose their meaning. If until now you have tracked one overall cost per conversion, the bar for the non-branded campaigns is now a different number. It is agreed before the separation, not after the first report.
How is it measured afterwards?
- Non-branded campaigns — by the cost of acquiring a new customer.
- Branded campaign — by the share of protected searches and by whether a competitor is there.
- Total — by MER, because only that sees the two together.
Each of the three is read from a different place. The cost of a new customer requires conversions to be split into new and returning buyers — in Google Ads this is the new customers segment, and in GA4 the distinction between a first and a subsequent order. Without it you pay a premium for people who are already yours.
For the branded campaign, I look at the impression share on branded queries and the top-of-page impression rate. A competitor’s presence is checked in the auction insights — there you can see who else appears on the same queries and how often they show above you.
MER is calculated outside the platforms: total revenue divided by total ad spend across all channels. It is read on a monthly basis and is the only number that does not change when traffic moves between campaigns.
And what about Meta?
There the equivalent is retargeting. It also looks excellent and also often collects people who would have bought without the ad. The principle is the same — separate cold from warm audiences and do not judge growth by overall return.
The separation goes through the exclusions in the ad sets. Retargeting sits in a separate campaign with a clearly defined source: website visitors, people who watched the video, people engaged with the profile. In the cold ad sets the same audiences and the list of buyers are set as excluded, so that what is really a reminder is not counted as finding a new customer.
For broad-targeting campaigns the separation is not automatic. There I check what share of spend goes to existing customers and limit it if the share grows, because otherwise one campaign both discovers and collects at the same time, and the creative for the two is different.
The check is the same as for the brand: retargeting is paused and the total number of orders is watched, not the figures in the campaign. If the total does not move, you are paying for a click you would have received without the ad.
FAQ — Frequently asked questions about branded campaigns
Which words count as branded?
The company name, its common misspellings, the domain and the names of your own product lines.
Will my traffic drop if I pause the branded campaign?
Some of the clicks move to the organic result. Exactly how many can only be known through a test, not an assumption.
Can I bid on a competitor?
Technically yes, provided the trademark rules are followed in the ad copy. Effectiveness varies widely between industries and needs separate monitoring.
How do I separate them in Performance Max?
With branded terms excluded at campaign or account level. This is also one of the reasons PMax is launched after the Search campaigns have been cleaned up.
How much budget for a branded campaign?
As much as covers the real search for your name. More brings nothing — the volume is limited by how many people search for you.
Conclusion
Branded traffic is not a fraud, but it is easy. Mixed with the rest, it hides what a new customer really costs — and that is the number on which every growth decision is built.
If you do not know what share of your result comes from people who already know you, start with an independent review.
Request a quote for a digital audit →
Related articles: Performance Max · negative keywords · MER
