Your Performance Max campaign shows a 12x ROAS. Looks phenomenal in the quarterly deck. But here's the uncomfortable truth: half those conversions came from people who were already searching for your brand name. PMax didn't create demand; it caught people who were already buying. That's not prospecting. That's credit theft from your own branded Search campaign.
I've seen this movie play out dozens of times. A marketing team celebrates their "high-performing" campaigns while the actual customer acquisition engine quietly bleeds money underneath the blended metrics. The brand conversions make the numbers look healthy. The non-branded spend is losing money. And Smart Bidding is using the blended signal to make bid decisions that are wrong for both traffic types.
Separating brand from non-brand isn't a campaign tidiness exercise. It's a prerequisite for knowing whether your advertising is actually working.
The 19x Gap Nobody Talks About
Dreamdata's analysis of B2B Google Search Ads revealed something that should make every CMO pause: branded campaigns with just 18% of budget are driving a 1299% return, while non-branded campaigns consuming 82% of budget deliver only 68% ROAS. That's a 19x performance gap.
The instinct is to celebrate the branded numbers. Don't. Those conversions were coming anyway. The person searching "YourCompany pricing" had already decided to evaluate you. Your ad didn't create that intent; it intercepted it. The real question is whether your non-branded spend is actually generating new customers or just burning cash while the brand traffic masks the damage.
When someone searches your brand name, they've already made a decision. When someone searches "best B2B marketing platform," they're evaluating options. These two behaviors require completely different bidding strategies, messaging, and success metrics. Mixing them in the same campaign is like measuring your marathon pace by averaging in your elevator rides.
What Mixed Campaigns Actually Hide
As one practitioner documented, the math gets ugly fast. If your campaign spends $5,000 total with $500 on brand (400 conversions) and $4,500 on non-brand (90 conversions), the blended numbers look like 490 conversions at $10 CPA and 800% ROAS. Looks great in the report.
The non-brand picture alone: 90 conversions at $50 CPA and 180% ROAS. Almost certainly unprofitable. But you can't see this because the brand performance is averaging over it.
Smart Bidding makes this worse. The algorithm doesn't know that brand traffic converts at 15% and non-brand at 1%. It sees a blended average of maybe 3-4% and sets bids accordingly. Too low for situations where you could bid more aggressively on high-value non-brand prospects. Unnecessarily high for brand traffic where you were going to get the conversion regardless.
You're making bidding and budget decisions based on numbers that obscure the performance that actually determines whether your growth spending is working.
The Performance Max Problem
Optmyzr's analysis of over 500 Google Ads accounts found that 91% had keyword overlap between Performance Max and Search campaigns. In 56% of Search campaigns, PMax still triggered for the same terms even when it shouldn't have. When there's a big performance gap, Search usually outperforms PMax.
The overlap is often caused by mismatched budgets, targeting, or schedule, not keyword setup. PMax is designed to find conversions wherever it can. If your branded terms convert at 15x the rate of your non-branded terms, guess where PMax is going to spend its time?
Google's January 2025 updates finally gave advertisers campaign-level negative keywords and better brand exclusion controls. For retail advertisers with product feeds, you can now apply brand exclusions to just Search text ads while keeping branded traffic for Shopping ads. This matters because it lets you separate the measurement while maintaining visibility where it counts.
An audit of 150+ Google Ads accounts found that one of the most common mistakes was branded traffic spread messily across multiple campaign types, inflating CPA while confusing the algorithm. In 2026, if you're not actively excluding brand terms from non-brand campaigns, you're buying what you already own.
The Governance Question
This isn't just about campaign structure. It's about organizational clarity.
Non-branded campaigns create demand. They help new people find you. Brand campaigns protect and convert demand. They make sure your earned attention doesn't leak at the last second. These are fundamentally different jobs with different success criteria.
When you mix them, you lose the ability to answer basic questions: Is our advertising actually acquiring new customers? What's our true cost to acquire someone who's never heard of us? Are we growing, or are we just capturing existing demand more expensively?
Seer Interactive's framework suggests five signals that it's time to test pulling back on brand search:
- You're ranking #1 organically
- Branded spend is climbing to 20-30%+ of your paid search budget
- There's low competitive pressure in Auction Insights
- You're optimizing for profit rather than just clicks
- You have full-funnel visibility through CRM or LTV data
If most of those ring true, you're probably paying for traffic you'd get anyway.
The Separation Playbook
Create a dedicated brand Search campaign with exact and phrase match on your brand name, common misspellings, branded product names, and your website domain. Use Maximize Clicks or manual CPC. Brand campaigns convert reliably enough that Smart Bidding's optimization adds little value.
For Performance Max, use brand exclusions to prevent your campaigns from serving for branded queries. Go to your campaign settings, click Additional settings, then Brand exclusions, and select the brand lists you want to exclude. You can still allow Shopping ads to appear for excluded brand searches if that visibility matters for your business.
Then run an incrementality test. Match your test duration to your goals:
- 4-8 weeks for short-term keyword overlap assessment
- 3-6 months for full-funnel CRM impact
- 6-12 months to uncover slow erosion in brand demand and direct traffic
Expect a short-term drop in ROAS and rise in CPA when you separate. That's not failure. That's clarity. You're finally seeing what your prospecting spend is actually doing.
The Uncomfortable Conversation
Here's what nobody wants to admit: separating brand and non-brand campaigns often reveals that the non-brand performance is worse than anyone thought. The blended metrics were flattering. The reality is harder.
But that's exactly why you need to do it. You can't fix what you can't see. And you can't scale customer acquisition profitably if you don't know what customer acquisition actually costs.
Brand exclusions don't make campaigns more profitable. They make them honest. And in a world where AI search advertising is projected to grow from $2.08 billion in 2026 to $25.93 billion by 2029, honest measurement is the only foundation for sustainable growth.
The dashboard might look less impressive after you separate. Your understanding of what's actually working will be dramatically better. And that's the trade every serious marketer should make.
