How An Ecommerce Brand Scaled Beyond Search With Google Demand Gen Campaigns

Google Demand Gen campaigns are Google’s intent-rich, upper-funnel ad format designed to reach new audiences across YouTube, Gmail, and Discover feeds before they actively search. For ecommerce brands that have maxed out Search and Shopping growth, Demand Gen is the most viable path to filling the top of funnel without abandoning ROAS discipline. This article walks through how a mid-size ecommerce brand used a structured Google Demand Gen strategy to expand reach significantly, lift branded search volume, and actually improve bottom-funnel efficiency in the process. The key was not just turning on a new campaign type. It was the structural decisions around audience architecture, creative cadence, budget separation, and measurement that made the difference between wasted upper-funnel spend and a genuine growth lever.

The Situation: Strong ROAS, But A Clear Ceiling

The brand in question is a mid-size direct-to-consumer ecommerce business doing around $80K per month in Google Ads spend. Their account was well built. Search campaigns were segmented by intent tier. Shopping was running through a tightly structured Performance Max setup with proper brand exclusions. ROAS targets were hitting consistently.

The problem was not performance. The problem was growth.

Search Demand Was Finite

Their category had a fixed volume of monthly searches. They had already captured the majority of high-intent queries and were seeing diminishing returns from expanding into broader match types. CPCs were climbing quarter over quarter as competitors fought over the same finite demand pool.

The Blended Metric Trap

The team had tried standard Display campaigns before and abandoned them. The issue was not just creative quality or targeting. It was measurement. Display spend dragged down blended ROAS, and the CMO could not see any clear lift in bottom-funnel metrics to justify the cost. Every upper-funnel experiment died the same death: “it’s not performing” when judged by the same ROAS targets as Search.

This is the trap that keeps ecommerce brands stuck at whatever their search demand ceiling happens to be. If you only buy traffic from people already looking for your product, you can only grow as fast as that demand grows. And if your category is mature, that growth rate might be close to zero.

Why Google Demand Gen Was The Right Expansion Path

Google Demand Gen campaigns replaced the old Discovery ad format and represent something fundamentally different from both Performance Max and standard Display.

What Demand Gen Actually Is

Demand Gen is Google’s answer to Meta’s feed-based ad experience, but built on Google’s intent signals rather than social behavior data. The inventory spans YouTube in-feed, YouTube Shorts, Gmail promotions, and Google Discover. These are high-attention environments where users are actively consuming content, not passively scrolling past banner ads on random websites.

How It Differs From Performance Max And Standard Display

Performance Max is a conversion-optimized, all-inventory campaign type. It is designed to find people ready to convert right now, across every Google surface. The problem is that PMax often cannibalizes branded search and gives you credit for conversions that would have happened anyway.

Standard Display has a reach problem of a different kind: the inventory quality is low, the targeting is broad, and the environments (random news sites, apps, games) do not prime users for commercial consideration.

Demand Gen sits in between. It reaches users before they search, but in contexts where they are already engaging with content. The audience targeting is built on Google’s search and browsing behavior signals, which means you are reaching people who have demonstrated category interest through their actual behavior, not just their social profile.

Why The Audience Targeting Is Different From Meta

On Meta, you target based on interests and lookalike modeling derived from social activity. On Demand Gen, you can build custom segments based on what people have actually searched for on Google, what competitor sites they have visited, and what content they have consumed. For ecommerce brands, that distinction matters. You are not guessing at intent. You are using demonstrated search behavior to find people who have not yet searched for your brand specifically.

The Strategy They Used To Make Google Demand Gen Ads Work

This was not a case of turning on a campaign and letting Google figure it out. The team made several structural decisions before a single dollar was spent.

Audience Architecture Built On Competitor And Category Signals

The team built custom segments around three layers. First, people who had searched for competing brand names and product categories. Second, people who had visited competitor websites (using Google’s affinity and in-market signals). Third, lookalike segments based on their existing high-value customer list. Each layer ran as a separate ad group so they could see which signal quality drove the best downstream behavior.

Creative That Matched The Feed Environment

This is where most ecommerce brands fail with Demand Gen. They repurpose their standard product photography or Shopping feed images and wonder why engagement is low. The feed environments where Demand Gen serves (YouTube, Discover, Gmail) are content-first. Users are watching videos, reading articles, scanning email. A product-on-white image looks like an interruption.

The team produced short-form video (15-second spots shot specifically for vertical formats) and lifestyle imagery that showed the product in context. The creative answered a simple question: if you saw this in your YouTube feed between two videos you actually wanted to watch, would you stop?

Soft Conversion Instead Of Direct Purchase Push

This was the single most important tactical decision. Instead of sending Demand Gen traffic directly to a product page and measuring immediate purchase ROAS, the team used a soft conversion: a product quiz that led to a personalized recommendation and email capture. This accomplished two things. It gave the campaign a measurable conversion event that Google’s algorithm could optimize toward. And it built an owned audience for remarketing and email nurture, so the eventual purchase could happen through Search or direct traffic days or weeks later.

Budget Separation To Avoid Blended Metric Distortion

The team gave Demand Gen its own budget and its own KPIs, completely separate from Search and Shopping. Demand Gen was measured on cost per email capture, quiz completion rate, and 60-day view-through purchase contribution. Search and Shopping continued to be measured on direct ROAS. This prevented the CMO from killing the campaign after week one because blended ROAS dipped, which is exactly what happened the last time they tried Display.

What Happened Over 90 Days

Reach Expanded Without Cannibalizing Bottom-Funnel Budgets

Within the first month, Demand Gen was reaching roughly four times the audience volume of their Search and Shopping campaigns combined, at a fraction of the CPM they had seen on standard Display. Importantly, the Search and Shopping budgets did not change. This was net-new reach funded by net-new budget.

Branded Search Volume Increased

By week six, branded search volume had increased noticeably. The brand was seeing more people searching for their exact brand name and product names. This is the clearest signal that upper-funnel advertising is working: people who saw you in a non-search context later go to Google and search for you by name.

60-Day View-Through Attribution Showed Real Contribution

Using a 60-day view-through window, the team could see that users who were exposed to Demand Gen ads were converting through Search, Shopping, and direct traffic at a meaningfully higher rate than unexposed users. The contribution was not immediate, and it would not have shown up in a last-click ROAS report. But it was real and measurable.

Search Campaign ROAS Actually Improved

This is the counterintuitive result. As Demand Gen drove more informed, brand-aware traffic into the funnel, the people clicking on Search ads were more likely to convert. They had already seen the product in their feed. They were not cold clickers comparing five options. They were warm prospects coming to confirm a decision they had already started making. Search ROAS improved because the quality of the search audience improved.

This is why obsessing over ROAS targets without understanding what feeds them is one of the most common ways ecommerce brands limit their own growth.

The Key Structural Decisions That Made It Work

Three decisions separated this from every failed upper-funnel experiment the brand had tried before.

Separate Campaign, Separate Budget, Separate KPIs

If you judge Demand Gen by the same direct-purchase ROAS as Search, you will kill it within two weeks. Upper-funnel campaigns create demand. Bottom-funnel campaigns capture it. They need different scorecards.

Creative Refresh Every Three Weeks

The team refreshed creative assets on a roughly three-week cycle. Feed-based inventory is subject to frequency fatigue faster than Search because users see the same visual over and over. Stale creative was the number one reason performance degraded in the middle of month two, and the fix was simply having new assets ready to rotate in.

Audience Exclusions That Prevented Waste

Demand Gen was excluded from serving to existing customers and recent purchasers. The goal was net-new reach, not retargeting people who already bought. Without this exclusion, the campaign would have chased easy conversions from existing customers and looked great on paper while adding zero incremental value.

How Fully Managed Execution Changes The Math On Multi-Layer Google Ads

Here is the operational reality that most advice about Google Demand Gen strategy ignores: running Search, Shopping, and Demand Gen simultaneously with proper audience architecture, creative refresh cadence, budget separation, and cross-campaign attribution is a full-time job. For many ecommerce brands, it is more than a full-time job.

This is where the structural fixes that actually recover ROAS for ecommerce brands start to compound. Every new campaign layer you add increases the operational load on whoever is managing your account. Creative production, audience management, exclusion lists, attribution modeling, budget pacing across campaign types: these are not set-it-and-forget-it tasks.

For teams using groas, this operational burden is exactly what the proprietary engine handles around the clock. In a DFY engagement, your dedicated strategist owns the entire multi-layer architecture, from Search through Shopping through Demand Gen, and makes the structural decisions about budget allocation, audience segmentation, and creative rotation without you needing to manage any of it. In a DWY setup, your in-house team stays in the driver’s seat while the engine runs the heavy execution and a senior strategist advises on exactly these kinds of cross-campaign structural decisions.

The core issue with doing this through a traditional agency is capacity. An agency media buyer managing your account alongside a dozen others simply does not have the hours to monitor Demand Gen creative fatigue, adjust audience exclusions weekly, and cross-reference upper-funnel exposure with bottom-funnel conversion patterns. They are capped at whatever one person can physically get through in a week. groas puts a senior strategist on top of an engine trained on over $500 billion in profitable ad spend, so execution does not stop when a human runs out of hours.

For agencies themselves, the DIY product gives media buyers direct access to the groas engine to run this kind of multi-layer architecture across their entire client book without adding headcount. The engine handles the execution-heavy work while the agency focuses on strategy and client relationships.

What This Means For Any Ecommerce Brand Considering Demand Gen

Google Demand Gen campaigns for ecommerce are not a magic growth button. They are a structural expansion of your Google Ads architecture that requires deliberate decisions about audience targeting, creative production, budget separation, and measurement. Brands that approach Demand Gen with the same mindset they use for Search, judging everything by immediate direct-purchase ROAS, will kill the campaigns before they have a chance to work.

The brands that succeed treat Demand Gen as what it is: a demand creation layer that feeds the bottom of funnel, measured on its own terms, and managed with enough operational rigor to keep creative fresh, audiences clean, and attribution honest.

If your Search and Shopping campaigns are strong but growth has plateaued, and your category has a finite search demand ceiling, Demand Gen is likely your next move. The question is whether you have the operational capacity to run it properly alongside everything else. If the answer is no, that is exactly the problem groas exists to solve. No onboarding fee, no long-term contract, and a team that can have this running inside your account without you needing to manage it. Apply for DFY if you want groas to own it end to end. Get started with DWY if you want your team to stay in control with the engine and a strategist working alongside you.