The Campaign Type Most Brands Are Ignoring
Here's a question that changed how we think about Google Ads:
Which campaign type offers the most scale?
This framework considers how to expand eligible reach while protecting the actual account's economics and ability to learn. Demand Gen is one possible part of that mix.
We've spent the last 18 months tracking this across our client accounts. We looked at our fastest-growing brands. The ones that went from testing to seven figures on Google. (And yes, we were surprised by what we found.)
The pattern was consistent.
A brand we scaled from $2k/day to $16k/day within 3 months. A jewelry brand we took from $3.5k/day to $36k/day in 47 days. Another we grew from $10k/week to $26k/week in 2 months.
Different industries. Different products. Different price points.
But one thing in common: Demand Gen campaigns were the engine behind their growth. Every time.
This isn't about Demand Gen being "better" than Shopping or Search. It's about understanding which campaign type lets you scale fastest when you're ready to push beyond your current ceiling.
Many brands we talk to have tried Demand Gen. Many wrote it off in 2023 or early 2024 when the platform was still maturing. But Google shipped 60+ AI-powered improvements in 2025 alone, and the game has changed.
Confirm current Demand Gen availability and inventory in the actual account. Platform announcements and other advertisers' results do not establish this account's incremental return.
That last stat matters. It means Demand Gen isn't just reshuffling existing demand. It's creating new customers.
The question isn't whether Demand Gen works. It's whether you've built the system to make it scale.
Why Scale Matters More Than Efficiency
There's a trap that's easy to fall into. (And it feels like smart marketing until it isn't.)
They optimize for efficiency at the expense of growth. They find a 4x ROAS campaign and protect it like a fragile artifact. They're hesitant to add budget because it might "break" what's working. (We've been that team. The instinct to protect a winning campaign is strong.)
This is backwards thinking.
A campaign at $5K/day and 4× attributed ROAS and one at $50K/day and 2.5× have different revenue and spending levels. Compare contribution, cash limits, repeat behavior and incremental evidence before deciding which is preferable.
Campaign type changes the inventory, controls and creative available. Demand, budget, creative, offer, measurement and unit economics can each constrain useful scale.
Search responds to eligible query demand. Keyword estimates are planning inputs, not a fixed count of every auction or a permanent profitable-spend ceiling.
Shopping is capped by product demand. Similar constraint. The number of people actively shopping for what you sell limits how much you can spend profitably.
Demand Gen can reach beyond an active product search. Creative capacity matters, alongside eligible audiences, inventory, auction costs, product demand and contribution. Creative volume is not its only limit.
An observed spend plateau does not define a permanent channel ceiling. Check demand, inventory, economics, creative and measurement before expanding the campaign mix.
This is why scale matters as a strategic priority, not just a nice-to-have.
The brands making 7-8 figures on Google understand this. They don't ask "what's my most efficient campaign?" They ask "what gives me the most room to grow?"
Demand Gen answers that question differently than any other campaign type on Google.
The Data Behind the Fastest-Growing Accounts
Let's get specific.
We analyzed our accounts that scaled fastest in 2025. The ones that went from testing to serious volume. Here's what we found.
Case Study 1: Apparel Brand
The author reports a spend trajectory in this account example. Retain the actual dates, attributed outcome definitions and change log before inferring why it scaled; spend growth is not profit or causal evidence.
The interesting part wasn't just the scale. It was the sequence. They started with Performance Max for their Shopping placements. Hit a ceiling around $5k/day. ROAS was strong but volume wouldn't budge.
The author reports that additional campaign activity coincided with greater spend and changed PMax performance. That account sequence does not isolate Demand Gen's effect from other changes or establish incremental profit.
Case Study 2: Jewelry Brand
This one moved faster because we had strong creative assets from Meta. The brand was already doing well on Facebook and Instagram. We repurposed their best performers for YouTube.
The mechanism: their customer wasn't searching for jewelry on Google. They were browsing YouTube, watching style content, discovering new brands in their feed. Demand Gen put them in front of that audience. PMax captured them when they searched later.
Case Study 3: Supplements Brand
Supplements are notoriously hard on Meta. High CPMs, aggressive competition, restrictive policies. YouTube was the opposite. Lower CPMs, less saturation, and the ability to target people searching for topics like "moringa benefits" or "natural energy supplements."
We ran informational content as ads. Educational videos that happened to feature the product. This is where Demand Gen shines - you can reach people in learning mode, not just buying mode.
The Pattern Across All Three:
The account stories motivate a hypothesis about reach and later demand. They do not establish a universal new-customer percentage or a causal effect on other campaigns.
In one author-reported pause sequence, the account contracted within two weeks. That observation motivates further investigation; without a suitable comparison and concurrent-change record, it does not establish Demand Gen as the cause.
What Makes Demand Gen Different
Three things separate Demand Gen from other campaign types when it comes to scaling.
First: Reach is not constrained by existing demand.
This is the fundamental difference. Search campaigns require someone to type a query. Shopping campaigns require someone to be actively shopping. Demand Gen doesn't require any prior intent.
Available reach is one input. Actual audience behavior, inventory, auction costs and useful creative still constrain delivery and commercial results.
According to Google's data, Demand Gen campaigns can reach users across 3+ billion daily visits on YouTube and Discover combined. That's the scale we're talking about.
Second: Two channels in one campaign type.
This is something most operators don't fully appreciate.
YouTube Shorts and Discover placements are scroll environments. People are browsing, discovering, entertaining themselves. This is pure top-of-funnel awareness.
But long-form YouTube is different. People there are searching. Looking for information. Trying to solve problems. This is mid-funnel intent.
Demand Gen gives you both in a single campaign. You can adjust your strategy based on what you're selling:
- Apparel brand? Lean into Shorts for style inspiration.
- Supplements brand? Target long-form YouTube where people search for health information.
- Home goods? Use Discover for lifestyle content and in-stream for product education.
Choose the campaign type and inventory controls that support the actual buyer question. Format flexibility alone does not establish superiority over every other campaign type.
Third: Investigate effects beyond directly attributed conversions.
This is where it gets interesting for full-funnel operators.
The author reports other-channel changes alongside Demand Gen activity across selected accounts. Those observations suggest a cross-channel hypothesis; they do not establish a universal lift or isolate Demand Gen from concurrent work.
The mechanism: YouTube ads create awareness. People see your brand. Then when they search later, they're more likely to click and convert. The touchpoint attribution might go to Search, but the journey started on YouTube.
Clients have also reported changes in direct traffic and Amazon sales during YouTube activity. Those reports do not identify each customer's first exposure or isolate YouTube's incremental contribution.
This compound effect is why Demand Gen isn't just "another campaign type." It's infrastructure for scaling the entire account.
The Dual-Channel Advantage
Let's go deeper on the two environments within Demand Gen.
Scroll Environments: YouTube Shorts + Discover
These placements catch people in passive discovery mode. They're not looking for anything specific. They're scrolling, being entertained, letting the algorithm serve them content.
This is where pattern interrupts matter most. You have 0.3 seconds to stop the scroll. The hook has to hit immediately.
What works here:
- Unexpected visuals or sounds in the first frame
- Contrarian statements that challenge assumptions
- Satisfying results shown immediately (before/after content)
- Relatable pain points stated directly
We've tested hundreds of hooks in these placements. The winners aren't polished. They feel native to the environment. Almost like organic content that happens to be an ad.
Search-Adjacent Environments: Long-Form YouTube
This is fundamentally different. People watching 10-20 minute YouTube videos are in learning mode. They searched for something. They chose to watch. They're paying attention.
This is where VSL-style content works. You can take 30 seconds, 60 seconds, even 2+ minutes to make your case. The audience will stay if you're teaching them something useful.
What works here:
- Educational content that leads to your product
- Problem-solution narratives with clear mechanisms
- Social proof built into the content (results, testimonials)
- Strong CTAs with visual overlays reinforcing the action
The strategy should match the environment:
| Environment | Mindset | Creative Length | Hook Style | CTA Approach |
|---|---|---|---|---|
Shorts | Passive | 15-30 sec | Pattern interrupt | Soft, native |
Discover | Browsing | Static/carousel | Visual pop | Curiosity-driven |
In-stream | Active learning | 60-180 sec | Value promise | Direct, verbal |
A common approach is using the same creative across all placements. This is a mistake. The dual-channel nature of Demand Gen requires dual creative strategies.
The Compound Effect on Other Channels
Here's what happens when you run Demand Gen at scale:
Lift on Performance Max
The author's apparel sequence compared two periods with and without Demand Gen. A time-switch comparison can still be confounded by demand, promotions and other changes; the article does not establish randomized assignment or a stable counterfactual.
The retrospective reports PMax ROAS of 3.2× during one period and 2.4× during another. A time-period comparison can reflect demand, promotion or other changes; it is not a randomized lift estimate.
Similar products and visible settings do not establish that all relevant conditions were held constant. Retain the actual design and confounders before interpreting a change as causal.
The mechanism is straightforward. Demand Gen creates awareness. Some of those viewers convert directly. But many more convert later through other channels. PMax picks up a portion of that demand because it's running across Google's surfaces.
Lift on Brand Search
The author reports 20-30% changes in branded query volume during selected 4-6-week Demand Gen periods. Those retrospective observations need comparable query definitions, dates and other changes before supporting a causal interpretation.
This is valuable beyond the obvious. More brand searches means more cheap conversions. Brand terms on Search have the lowest CPAs in any account. When YouTube drives branded searches up, your overall blended metrics improve.
Lift Beyond Google
This is the part that surprised us.
Clients running heavy YouTube spend report increases on:
- Shopify direct traffic (no referral source)
- Amazon sales of the same products
- Retail partner sell-through
The customer sees the YouTube ad. Maybe they don't click. But the next time they're on Amazon searching for that product category, they recognize the brand. They click that listing instead of a competitor.
YouTube is creating memory. That memory converts across channels.
The Strategic Implication
Directly attributed ROAS may omit later effects, while adding later revenue without a valid comparison can overstate them. Assess the actual business outcome and use a suitable design to test the cross-channel hypothesis.
A difference between advertisers using different campaign mixes does not isolate the mix's effect. Compare the actual account decision with a suitable design and business outcome.
The Demand Gen Scaling Engine
Here's the framework we use to scale Demand Gen from zero to $100k/month+.
We call it the Demand Gen Scaling Engine. Four phases, each building on the previous.
Phase 1: Foundation (Week 1-2)
Goal: Establish baseline performance with proven elements.
Before scaling, you need something that works. This phase is about validating that your creative and targeting can drive conversions at an acceptable cost.
Setup checklist:
- Choose supported Demand Gen channel controls for the intended inventory
- Start with a manageable set of creatives matched to the buyer question
- Configure eligible audience segments and review expansion settings
- Verify measurement, consent and account-specific delivery
Success criteria: Review mature outcomes against the actual contribution boundary and test question. Positive attributed ROAS after 14 days does not establish profitability or sufficient evidence.
Phase 2: Validation (Week 3-4)
Goal: Identify winning creative angles and audiences.
Now we're testing. Not scaling yet. Testing to find what resonates.
What to test:
- 3-5 creative angles (different hooks, different value props)
- 3-4 audience segments (different intent signals, different lookalike seeds)
- Creative lengths (15s vs 30s vs 60s+)
Process:
- Add reviewed creative through a suitable test design
- Choose exposure and observation windows from the expected effect and conversion delay
- Use account-specific stop rules and uncertainty
- Allocate further budget only when the business evidence supports it
Success criteria: 2-3 clear winning creative + audience combinations identified.
Phase 3: Acceleration (Week 5-8)
Goal: Increase spend on validated winners.
This is where brands commonly stall. They found winners but are hesitant to push budget. (Totally understandable - scaling means spending more money before you know if it'll work. That's not fear, it's rational caution. The framework removes the guesswork.)
Scaling rules:
- Choose budget changes from mature outcomes, contribution and acceptable risk
- Preserve a clear question when testing new creative or audiences
- Use supported and permitted audience seeds
- Verify the selected format; Demand Gen in-stream video is skippable
What to monitor:
- Frequency (keep under 3 per 7 days initially)
- Audience overlap with other campaigns
- Creative fatigue (watch for CTR decay)
Example decision rule: agree an actual contribution boundary, acquisition outcome and useful evidence window before the test. Spend growth or a chosen channel-share target alone is not commercial success.
Phase 4: Compound (Week 9+)
Goal: Build the flywheel between Demand Gen and other channels.
Now we're optimizing the system, not just the campaigns.
Full-funnel integration:
- Review audience exclusions where supported and appropriate
- Treat PMax audience signals as suggestions, not exclusive targeting
- Investigate later brand demand without assuming its source
- Keep permitted customer and source records consistent across follow-up
Creative scaling:
- Produce 10-20 new creative concepts monthly
- AI tools for rapid iteration (we've tested brands doing $1m+ on AI-generated video ads)
- Test long-form VSL content for in-stream placements
Phase-by-Phase Implementation
Let me break down what each phase looks like operationally.
Phase 1 Implementation: Foundation
Campaign Structure:
The retained sketch below describes an earlier planning idea, not a valid Demand Gen import. Demand Gen uses ad groups; its in-stream video is skippable, so the non-skippable campaign shown is not a Demand Gen setup. Build from current Demand Gen format controls and the actual account.
Campaign 1: Demand Gen - Shorts
- Asset Group: [Product Line A]
- Creative: 1-2 vertical videos, 15-30 seconds
- Audiences: Custom intent (competitor keywords) + Affinity (relevant interest)
Campaign 2: Demand Gen - In-stream Skippable
- Asset Group: [Product Line A]
- Creative: 1-2 horizontal videos, 60-90 seconds
- Audiences: Custom intent (problem keywords) + Lookalike (purchasers)
Campaign 3: Demand Gen - Non-skippable 15s (Remarketing)
- Asset Group: [Product Line A]
- Creative: 1 high-impact 15-second ad
- Audiences: Website visitors (7-30 days)
Budget Allocation:
- Shorts: 20-25%
- In-stream Skippable: 50-55%
- Non-skippable Remarketing: 25-30%
Bidding:
- Start with Target CPA or Maximize Conversions
- Set realistic targets based on historical account data
- If new account, start with Maximize Conversions and let it learn
Phase 2 Implementation: Validation
Testing Framework:
| Week | Test Type | Method |
|---|---|---|
Week 3 | Creative angles | Add 3-5 new hooks to winning campaign structure |
Week 4 | Audiences | Duplicate winning campaigns with new audience combinations |
What "winning" looks like:
- Mature outcomes meeting the stated business decision rule
- A suitable design, exposure and uncertainty assessment; 50 conversions alone is not statistical significance
- Useful creative and audience evidence after reviewing delivery and confounders
Kill criteria:
- Below 50% of target ROAS after learning period
- CTR declining 20%+ week-over-week
- Frequency above 5 per 7 days
Phase 3 Implementation: Acceleration
Scaling Protocol:
Day 1-3: Increase budget 20% Day 4-6: Monitor ROAS (must be within 90% of baseline) Day 7-10: Increase budget another 20-30% Day 11-14: Evaluate and decide: continue scaling or stabilize
Warning Signs to Pause Scaling:
- ROAS drops below 70% of baseline
- CPM increases more than 30%
- Frequency exceeds 4 per 7 days
- Conversion rate declining
How to Fix:
- Refresh creative (new hooks, new assets)
- Expand audiences (add new lookalike seeds)
- Test new placements (add Discover if not running)
Phase 4 Implementation: Compound
Cross-Channel Optimization:
-
Audience sharing:
- Export Demand Gen video viewers as audience
- Use as signal in PMax audience suggestions
- Create dedicated remarketing campaigns in Search
-
Attribution tracking:
- Set up data-driven attribution to see YouTube's influence
- Monitor assisted conversions in Google Ads
- Track branded search volume weekly
-
Creative pipeline:
- Monthly creative production cycle
- Test 10+ new concepts per month minimum
- UGC and AI-generated content for rapid iteration
The Five Mistakes We See Repeatedly
Let me be direct about where brands go wrong.
Mistake 1: Judging too early
Plan the observation window from actual conversion delay, exposure, outcome rate and study design. A fixed number of weeks cannot establish that a campaign has optimized or that a short result is necessarily noise.
Mistake 2: Using Meta creative without adaptation
Your Facebook ads don't automatically work on YouTube. The environments are different. Scroll speed is different. Audio expectations are different. YouTube viewers expect sound on. Meta viewers often watch muted.
Adapt creative to the actual placement and buyer question, then compare like-for-like exposure. The earlier format-performance range was not supported by retained test evidence here.
Mistake 3: Wrong measurement window
Same-day reporting may omit delayed conversions, but a universal 60-70% missing-value share or 7-14-day delay is not established here. Inspect the actual conversion lag and attribution contract.
Choose conversion and follow-up windows from the actual buying cycle and study question. Keep attributed, modeled and incremental outcomes distinct.
Mistake 4: Treating it as a silo
Demand Gen works best as part of a system. Running it without PMax or Search to capture downstream demand wastes the awareness you're creating.
The brands that win run the full stack: Demand Gen for awareness, PMax for conversion, Search for capture, remarketing for close.
Mistake 5: Not producing enough creative
This is the bottleneck. Demand Gen scales with creative. If you're running 2-3 ads and wondering why you can't push past $20k/month, the answer is obvious.
The author reports high creative volume in selected larger accounts. Creative count alone does not explain their spend or establish a monthly requirement for another business.
The Window Is Open
Here's the reality of where we are right now.
YouTube and Meta costs depend on market, audience, format and period. The article does not retain a comparable current benchmark establishing universally lower YouTube CPMs or competition.
A third-party forecast does not establish the best budget allocation or likely outcome for this account. Base the expansion decision on current account evidence.
But there's a window. The brands figuring this out now will compound their advantage. They'll build creative libraries. They'll train their algorithms. They'll capture market share while the channel is still relatively uncrowded.
Competition and CPMs can change. The article does not establish a deadline or inevitable price increase.
The question isn't whether to start. It's whether you can afford to wait.
The author reports selected spend trajectories, including $2K/day to $16K/day and $3.5K to $36K. The examples do not establish typical outcomes, profitability or the effect of this framework.
The Demand Gen Scaling Engine is the system.
Foundation. Validation. Acceleration. Compound.
Four phases. Clear success criteria at each stage. A framework you can execute on starting tomorrow.
The framework organizes research, creative and measurement decisions. Whether it expands useful reach or commercial contribution needs actual account evidence; lower CPMs and compounding returns are not guaranteed.
That's the opportunity.
Put the framework to work
Plan a bounded Demand Gen test