Brand searches are people who already typed your name. They convert more easily, they cost less, and they will make a blended campaign look healthy while the cold queries in the same campaign lose money. If those two kinds of demand share a budget, a bid strategy, and a ROAS number, you cannot tell whether you are buying new demand or paying for demand you already earned.

Split them. Then keep them split. A negative list is what stops the brand campaign's success from leaking back into the non-brand ad groups and flattering the average.

What goes in the brand campaign

The brand campaign gets your name, the obvious misspellings, and the brand-plus-product queries you are willing to pay to defend: "your brand duvet," "your brand promo code," "your brand returns" only if you have decided a paid click is really how that support query should be answered. Many of those support queries should be negatives and organic pages instead.

Give this campaign its own budget and its own target. It will look efficient. That efficiency is not a template for the rest of the account. Do not copy the brand ROAS target onto a cold collection and then wonder why the cold campaign "cannot hit the number."

Defend the name when other advertisers, or your own Shopping and Performance Max campaigns, are eligible to serve on it. The longer discussion of defense versus going after someone else's brand is in brand-term defense versus competitor conquesting. Separation is the prerequisite for both.

What must be excluded everywhere else

Non-brand product ad groups, problem and use-case tests, and competitor campaigns should carry your brand terms as negatives in the right match types. If they do not, a query for your name can match a generic keyword and get reported as a generic win.

Dynamic and expanded URL campaigns need the same exclusion. Otherwise the brand query is simply served from a different campaign type, and the Search split you were proud of is an accounting fiction. Write the owner down: brand Search owns the brand query, or a specific Shopping campaign does, but not both by accident.

Check the search term report for your name in campaigns that claim to be non-brand. The first week will show whether the negatives were too narrow. A misspelling you did not list is still your brand.

How mixed reporting misleads a catalog team

Consider a hypothetical. A blended Search campaign spends $20,000 in a month. Brand queries are $4,000 of that and return well above the merchandising target. Cold collection queries are $16,000 and return below the contribution you need after product cost. The campaign-level ROAS sits near the target, so nobody touches it. The cold spend keeps running because the brand spend is in the same cell of the spreadsheet.

Split the campaigns and the conversation changes. Brand is a defense cost you can judge on coverage and CPC. Cold is an acquisition cost you judge on contribution after product cost, which is the same standard as setting targets from contribution margin. They were never one number.

How Kopa keeps the split

In Kopa, Smart Campaigns isolates brand queries in their own bucket and excludes brand terms from non-brand and dynamic ad groups in the draft. The split is visible before you approve the launch, so brand efficiency cannot be blended into cold traffic after the fact.

What good looks like

  • Brand queries have their own campaign, budget, and target.
  • Brand terms and the misspellings you know are negatives on every non-brand Search campaign.
  • Other campaign types are not left free to claim the same brand query without an owner.
  • Reporting never presents one ROAS for brand and cold together as if it were a verdict.
  • The search term report is checked for your name outside the brand campaign.