Branded vs Non-Branded Keywords: How to Allocate Budget
The “should we bid on our own brand?” debate has been running for over a decade and somehow still produces wrong answers in most accounts. The wrong answer in either direction — bidding too heavily or not at all — wastes meaningful budget. This guide walks through what we actually know, where the conventional wisdom is wrong, and how to set the right budget split for your stage and situation.
What “branded” and “non-branded” actually mean
Branded keywords: queries that contain your brand name. “Digitelia”, “Digitelia pricing”, “Digitelia vs competitor”, “Digitelia reviews.”
Non-branded keywords: queries that don’t contain your brand. “Marketing agency”, “Google Ads management”, “SEO consultant.”
The distinction matters because they represent different user intent:
- Branded: user already knows you (or someone told them about you).
- Non-branded: user is in market for a solution; you’re one of many options.
These are fundamentally different acquisition motions. Treating them identically in budgeting is a common mistake.
Why branded keyword ROAS lies
Branded keywords nearly always show dramatically better ROAS than non-branded in standard reporting. A typical account:
- Branded campaign: 12× ROAS
- Non-branded campaign: 3× ROAS
The naive conclusion: shift budget from non-branded to branded. The actual reality: most of those branded clicks were going to come for free anyway (the user typed your brand name; they were already going to find you).
This is the incrementality problem. Reported branded ROAS is inflated by users who would have arrived organically. The true incremental ROAS of branded paid is much lower — usually 30-60% of reported.
This doesn’t mean stop bidding on branded. It means understand what branded paid actually delivers (control + defense, mostly) and budget accordingly.
What branded paid actually does
Three real functions:
1. Brand defense. Competitors can (and do) bid on your brand name. If you don’t bid, they appear above your organic listing. Your branded paid prevents competitor capture.
2. Message control. Your organic listing shows your homepage meta description. Your branded ad can lead to a campaign-specific landing page with a specific offer. More conversion-friendly.
3. SERP real estate. With branded paid + organic + knowledge panel, you can own 70-90% of the visible above-the-fold space, leaving little for competitors or unrelated results.
These functions justify branded paid even at modest incrementality. The right question isn’t “should we do it” but “how much should we spend on it.”
Where non-branded paid delivers value
1. Demand capture from non-customers. Users searching for solutions in your category who don’t know your brand. Pure new-customer acquisition.
2. Awareness creation. Even users who don’t click your ad see your brand in their search results, contributing to brand recall.
3. Competitive intelligence. Search-term reports reveal what queries users actually use, informing SEO and product positioning.
4. Long-tail demand. Hyper-specific queries you’d never rank for organically.
Non-branded is structurally where new customers come from. It’s also where most accounts overspend in unprofitable ways without realizing.
The budget allocation framework
Brand stage drives the split:
| Brand stage | Branded share | Non-branded share | Notes |
|---|---|---|---|
| Unknown / brand-new | 5% | 95% | No branded search demand yet |
| Emerging / building | 10% | 90% | Some branded search starting |
| Recognized in category | 15% | 85% | Branded paid for defense + control |
| Established / mid-market | 20% | 80% | Branded important; non-branded primary growth |
| Category leader | 25-35% | 65-75% | Brand defense critical; competitors bid heavily on your brand |
| Mature, defending against challengers | 30-50% | 50-70% | Pure defense; growth limited by category |
These are guidelines. Specific situations move them up or down.
When to invest more in branded
Push branded budget higher when:
1. Competitors are bidding on your brand. Check by searching your brand from incognito. If competitor ads show, you need to defend.
2. You have known affinity-marketing campaigns running. TV, podcast, billboard — these drive branded search. Make sure paid + organic captures the lift.
3. You’ve launched a product or major news. Pulse spikes in branded search around announcements. Defend.
4. You’re entering a new market. Defend the brand globally as you launch.
5. Your brand name has commodity ambiguity. A brand named “Apex” or “Atlas” gets confused with other entities. Branded paid disambiguates.
When to invest less in branded
Pull branded budget down when:
1. Zero competitors bid on your brand. You probably can rely on organic + maybe a small defensive presence.
2. Branded search volume is tiny. If you’re below 100 branded searches/month, branded paid is a rounding error.
3. Your conversion rate from organic branded is already strong. Sometimes organic outperforms paid — let it.
4. Budget is severely constrained. When choosing between feeding non-branded acquisition or defending brand, non-branded wins for growing brands.
Measuring incrementality
The hard part of this analysis: how do you know your branded paid is incremental?
Method 1: Geographic holdout test.
- Pause branded paid in one region for 4-6 weeks.
- Measure: branded organic clicks in that region vs. control regions.
- The branded organic uplift = your branded paid was largely non-incremental.
- The total branded clicks drop = your true incremental loss.
Method 2: Time-period test.
- Pause branded paid nationally for 14-30 days.
- Compare branded organic + paid before/after.
- More volatile due to seasonality but cheaper to run.
Method 3: Match-type test.
- Move branded from exact match to broad match in one region (cheaper coverage).
- Measure efficiency.
Most accounts that run incrementality tests find branded paid is 30-60% incremental. The rest is air. Adjust budget accordingly.
The branded ad structure that works
Run branded as its own campaign with:
- Exact match keywords (your brand + brand variants like “Digitelia”, “digitelia.com”, “digitelia agency”)
- Tight ad copy that matches your value proposition exactly
- Landing page = your homepage or branded campaign page (not a generic landing page)
- High CTR (5-15% is typical; below 3% suggests creative or relevance issue)
- Smart Bidding: Maximize Conversion Value or Target Impression Share
- Bid strategy: ensure 90%+ impression share for brand defense
Avoid:
- Broad-match branded keywords (will pull in non-branded queries with your brand in them)
- Generic ad copy (this is your brand; the ad should feel premium)
- Sending branded traffic to a non-branded landing page
The non-branded structure that works
Non-branded should be your growth campaign:
- Mix of exact and phrase match keywords
- Multiple ad groups by intent or theme
- Smart Bidding: Maximize Conversions or tCPA
- Multiple ad copy variants for testing
- Negative keywords aggressively maintained
- Customer Match audiences applied as bid modifiers (e.g., +30% for past customers)
Non-branded should have higher CPA than branded (often 5-10×) — that’s normal. What matters is whether it’s profitable at the closed-won level.
How to think about competitor branded keywords
A related question: should you bid on competitors’ brand names?
Generally yes for:
- “Better than competitor X” or “competitor X alternatives” queries
- Specific competitor names where you have a competitive edge
Generally no for:
- Pure brand-name bidding where you have no differentiation (you’ll pay high CPC to be the third option)
- Competitors with strong trademark enforcement (some will issue cease-and-desist; legally allowed but creates friction)
This is its own playbook — see our separate article on competitor brand bidding.
Common mistakes
1. Defunding branded based on incrementality alone. Even if branded is only 40% incremental, the brand defense value is real. Don’t go to zero.
2. Overspending on branded because reported ROAS looks good. Same data, opposite mistake. Reported ROAS is inflated.
3. Running branded and non-branded in the same campaign. Loses control. Always separate.
4. No brand exclusion list on Performance Max. PMax will eat your branded auctions otherwise.
5. Identical ad copy for branded and non-branded. Different intent requires different messaging.
6. Not measuring closed-won, only form fills. Branded leads convert at higher rates than non-branded. Without closed-won data, you’re missing the picture.
A 60-day branded/non-branded optimization
Days 1-15: Audit current state.
- What’s your current split? What’s the reported ROAS of each?
- Run an incrementality test on branded (geographic holdout if budget allows).
- Confirm brand exclusion list on Performance Max.
Days 16-30: Restructure.
- If branded and non-branded are in one campaign, split them.
- Set bidding strategies appropriate to each (tROAS for branded, tCPA for non-branded typically).
- Apply Customer Match and audience signals.
Days 31-45: Optimize non-branded.
- Search-term audit. Add negatives aggressively.
- Identify under-performing ad groups; pause or restructure.
- Build new ad copy variants for testing.
Days 46-60: Re-evaluate split.
- Based on incrementality test results, adjust budget.
- Most accounts end up shifting 10-25% of budget from over-funded branded to non-branded for growth.
Frequently asked questions
What’s a healthy branded vs non-branded ratio? For most established accounts: 15-25% branded, 75-85% non-branded. Tightly category-leader brands sometimes 30-40% branded.
Should I bid on my own brand for SEO benefit? There’s no SEO benefit from branded paid. The argument is purely brand defense, message control, and SERP real estate.
What if competitors stop bidding on my brand? Then your branded paid is less defensive and more about message control. You can reduce share but rarely to zero — your branded paid landing page often has higher conversion than your homepage.
Is bidding on competitor brand names ethical? Legally yes in most jurisdictions (bidding on terms isn’t infringement; using trademarks in ad copy is). Ethically: gray area. Some industries respect a “no brand bidding” norm; most do not.
How do I tell if branded paid is worth it for my specific account? Run the incrementality test. 4-6 weeks of holdout in one region tells you. There’s no shortcut.
The branded vs non-branded debate is well-worn for a reason: it’s expensive to get wrong in either direction. The right answer for any specific account requires actual incrementality measurement, not just reported ROAS. Most accounts that haven’t done this measurement have a budget split that’s noticeably suboptimal — usually too much branded, too little non-branded.