Google Demand Gen campaigns for ecommerce are one of the most effective ways to break through a Search and Shopping plateau, and this case study shows exactly how. An established ecommerce brand, running roughly $80K/month in Google Ads across Search and Shopping, had hit a ceiling. Revenue was flat for two consecutive quarters despite strong ROAS. The team added Google Demand Gen campaigns as a dedicated upper-funnel channel, restructured audience targeting around Customer Match and YouTube segments, and within 90 days cut upper-funnel cost per lead by 44% compared to the Display remarketing baseline it replaced. Branded search volume climbed measurably. Search and Shopping ROAS held steady. This is the full breakdown of how they did it, what surprised them, and what it means for ecommerce brands stuck in the same position.
This was a mid-market ecommerce brand in the home goods category doing roughly $12M in annual revenue, with Google Ads as its primary paid acquisition channel for three years. The account was well-built. Search campaigns were segmented by product category and intent tier. Shopping campaigns ran through a standard Performance Max setup with properly segmented asset groups. ROAS was consistently in the 4-6x range depending on the month.
The problem was not performance. The problem was growth.
Search volume for the brand's core product categories was not expanding. The category itself was mature. Branded search had plateaued because the brand had not invested in any awareness-level media since a brief Facebook push two years earlier. Every dollar of incremental Search spend was producing diminishing returns because they were simply bidding more aggressively on the same finite pool of high-intent queries.
Meanwhile, the brand's target audience (homeowners aged 30-55 with above-average household income) was spending significant time on YouTube, Gmail, and Discover. The brand had zero presence in those environments.
The hypothesis was straightforward: Google Demand Gen campaigns could reach these audiences at scale, generate qualified upper-funnel engagement, and feed the existing Search and Shopping campaigns with new branded interest, all without cannibalizing what was already working.
Before the rebuild, the budget split was approximately 60% Search and 40% Performance Max (Shopping-heavy). A small Display remarketing campaign ran on the side, consuming about 8% of total spend, targeting cart abandoners and past purchasers.
When the team tried pushing an additional $15K/month into Search, CPC rose but conversion volume barely moved. The high-intent queries were already captured. Broad match expansion pulled in tangential traffic that converted at less than half the rate. Adding more budget to Performance Max mostly resulted in Google allocating spend to low-quality Display network placements rather than Shopping.
The Display remarketing campaign was the only "awareness-adjacent" activity in the account, and it was not actually generating awareness. It was retargeting existing visitors with banner ads on sites they were not paying attention to. CPL on Display remarketing had crept up steadily over two quarters as the retargeting pools shrank from tighter privacy controls and cookie deprecation.
Three signals pointed to a channel ceiling rather than a creative or offer problem:
This was not a campaign structure problem. It was a demand generation problem, in the literal sense. The brand needed to create demand, not just capture it.
The strategic decision was to treat Demand Gen as a dedicated channel alongside Search and Shopping, not as a replacement for anything. The existing campaigns stayed untouched. The budget for Demand Gen came partly from reallocating the underperforming Display remarketing spend and partly from a net-new investment the brand committed to for a 90-day test.
Two Demand Gen campaigns were launched:
Each campaign had distinct creative and distinct audience signals, keeping the learning data clean.
The audience targeting decisions were arguably the most important part of the entire effort.
For the prospecting campaign, the team used:
For the mid-funnel campaign:
The key insight: Demand Gen's audience-first approach meant the targeting quality determined everything. Unlike Search, where intent is expressed through the query, Demand Gen requires you to define the intent pool yourself. Getting this wrong means burning budget on impressions that will never convert.
This is where the results defied expectations. The team tested three creative formats: short-form video (15 seconds), carousel image ads, and single static image ads.
In the first 30 days, single static images outperformed video on both CTR and downstream CPL for this brand. The hypothesis is that the product category (home goods with strong visual appeal) communicated value faster in a high-quality lifestyle image than in a video that required time to process. Carousels performed well for the mid-funnel campaign where users already had some brand familiarity and were willing to browse multiple products.
Video eventually caught up in weeks 6-8 as the algorithm optimized delivery, but the early learning was clear: do not assume video wins in Demand Gen just because YouTube is a primary placement. Test everything.
The brand allocated 15% of total monthly Google Ads spend to Demand Gen for the test period. Roughly 60% of that came from eliminating the old Display remarketing campaigns, and 40% was incremental. Search and Shopping budgets were not touched.
This is critical. One of the most common mistakes with Google Ads upper-funnel strategy for ecommerce is robbing Search to fund awareness. That cannibalizes your best-performing campaigns and makes the test look like a failure even if Demand Gen is working, because total ROAS drops.
After 90 days, the numbers told a clear story.
The prospecting Demand Gen campaign delivered upper-funnel leads (defined as email signups and first-time site visitors who engaged with at least two pages) at a 44% lower cost per lead compared to the Display remarketing baseline it replaced. This was not an apples-to-oranges comparison. The brand used the same conversion actions, the same attribution window, and the same reporting methodology.
Over 70% of the conversions from the prospecting Demand Gen campaign came from users who had zero previous engagement with the brand: no site visits, no email interactions, no previous ad clicks. These were genuinely net-new prospects that Search and Shopping could never have reached because these people were not searching for the brand's products yet.
During the 90-day test, branded search query volume increased compared to the same period the prior year, even controlling for seasonality. While direct causal attribution between Demand Gen impressions and branded searches is imperfect, the timing correlation was strong, and no other awareness-level marketing was running during the test window.
This was the most important validation. Search and Shopping ROAS did not decline during the Demand Gen test. If Demand Gen were simply cannibalizing existing demand (showing ads to people who would have searched anyway), you would expect to see Search metrics soften. They did not. Conversion volume on Search actually ticked up slightly, consistent with the branded search lift.
This is the single biggest conceptual shift teams need to make. Google Demand Gen campaigns in 2026 are audience-first, creative-dependent, and algorithmically optimized across YouTube, Discover, and Gmail. They are not banner ads on random websites. The targeting philosophy, creative requirements, and measurement approach are fundamentally different from Display.
Teams that import their Display playbook into Demand Gen (retargeting-heavy audiences, generic banner creative, last-click CPA targets) will get mediocre results and conclude the channel does not work.
Last-click attribution systematically undercounts Demand Gen's contribution. Someone watches a YouTube ad, later searches for the brand on Google, and clicks a Search ad. Last-click gives 100% credit to Search. Demand Gen gets nothing.
The brand in this case study measured Demand Gen contribution using three signals together: upper-funnel CPL on the Demand Gen campaigns themselves, branded search volume lift during the test period, and new-to-brand customer rate from Search and Shopping during vs before the test. No single metric told the full story. Together, they made the contribution undeniable.
Google Ads Demand Gen campaigns for ecommerce are not universally the right next move. They make sense when:
If your ecommerce account is still leaving Search and Shopping performance on the table, fix that first. Demand Gen is a channel expansion play, not a foundation.
This is where most ecommerce brands hit a wall. Running Demand Gen campaigns well requires audience strategy, creative testing at volume, budget management across channels, and measurement sophistication that goes beyond what a single media buyer or generalist agency can physically execute in a week.
For brands that want Google Ads fully managed, groas runs your entire account end to end, including Demand Gen campaign builds, audience strategy, creative testing, and landing page optimization. A dedicated senior strategist owns every decision while the proprietary engine, trained on over $500 billion in profitable ad spend, handles execution around the clock. That means creative tests run and optimize continuously, not when someone gets around to checking the account on Tuesday.
For in-house teams that want to keep control, groas pairs the same engine with a strategist who works alongside your team. You stay in the driver's seat. The engine does the heavy lifting on bid management, audience signal processing, and performance analysis, while the strategist brings insights from across a massive dataset of what works in Demand Gen across categories.
For agencies managing ecommerce client books, running Demand Gen across 20+ client accounts creates a creative and strategic bottleneck that agencies struggle to scale without new hires. groas gives agencies direct access to the engine so they can scale operations across their entire client book, keeping their brand and margin while eliminating the execution ceiling.
In every case, onboarding is $0, there are no long-term contracts, and you can cancel anytime. groas earns the next month by performing.
The pattern this case study illustrates is not unique. It is one of the most common growth ceilings in ecommerce Google Ads: a well-optimized Search and Shopping account that has captured all available demand and has nowhere to grow because no one is generating new demand. Google Demand Gen campaigns are the native solution to that problem within the Google Ads ecosystem.
The execution requirements are real. Audience strategy, creative production and testing, cross-channel budget management, and multi-touch measurement are table stakes. Brands that treat Demand Gen as "set up a campaign and let it run" will waste budget.
Brands that get the execution right, whether through a strong in-house team, an agency with real Demand Gen expertise, or a fully managed service like groas, unlock a growth lever that Search and Shopping alone cannot provide.
If your ecommerce account has plateaued and your team is stretched, apply for groas and let a dedicated strategist assess whether Demand Gen is the right next channel for your business.
Google Demand Gen campaigns are audience-first ad campaigns that serve across YouTube, Gmail, and Discover. Unlike traditional Display campaigns, which rely heavily on contextual or retargeting placements across the Google Display Network, Demand Gen uses machine learning to match creative with high-intent audience segments in premium Google-owned environments. The creative requirements are different (lifestyle imagery and video rather than standard banners), the targeting philosophy is audience-signal driven rather than placement driven, and the optimization algorithms work across surfaces that have meaningfully higher engagement rates than typical Display inventory.
Demand Gen makes sense when your Search and Shopping campaigns have captured most available high-intent demand and incremental spend is producing diminishing returns. Key signals include impression share above 80% on core terms, flat or declining new customer acquisition rates, and stagnant branded search volume. If your target audience is active on YouTube, Gmail, or Discover, and you have strong visual creative or can produce it, Demand Gen is the logical next channel to open new audience segments without cannibalizing existing performance.
When structured correctly, no. The key is to fund Demand Gen from net-new budget or by reallocating underperforming Display spend, not by pulling budget away from Search and Shopping. In the case study covered here, Search and Shopping ROAS held steady throughout the 90-day Demand Gen test, and branded search volume actually increased. If you see Search metrics soften after launching Demand Gen, it is more likely a budget allocation or audience overlap problem than an inherent channel conflict.
Last-click attribution systematically undercounts Demand Gen's contribution because users often see a Demand Gen ad, then later convert through a branded Search click. Effective measurement combines three signals: the upper-funnel CPL within the Demand Gen campaigns themselves, branded search volume lift during vs before the test period, and new-to-brand customer rates from Search and Shopping. No single metric tells the full story, but together they reveal whether Demand Gen is actually generating incremental demand.
It depends on the product category and audience, which is exactly why testing matters. In many ecommerce verticals with visually strong products, high-quality static lifestyle images can outperform video in early testing because they communicate value instantly without requiring viewing time. Carousels tend to perform well for mid-funnel audiences with existing brand familiarity. Video often catches up as the algorithm optimizes delivery over several weeks. The lesson: do not assume video wins because YouTube is a primary placement. Run creative format tests from day one.
Performance Max is a full-funnel campaign type that distributes budget across all Google inventory, including Search, Shopping, Display, YouTube, and Discover, optimized toward conversions. Demand Gen is purpose-built for upper and mid-funnel audience engagement on YouTube, Gmail, and Discover only. They serve different strategic functions. Performance Max captures existing demand. Demand Gen creates new demand. Most ecommerce accounts benefit from running both, but they should be structured and measured as complementary channels with distinct goals.
Demand Gen creates a real execution bottleneck for agencies because each client needs custom audience strategy, dedicated creative testing, and ongoing performance analysis. Scaling this across 20 or more accounts without adding headcount is where most agencies hit a ceiling. groas solves this by giving agencies direct access to a proprietary engine trained on over $500 billion in profitable ad spend. Agencies keep their clients, brand, and margin while the engine handles execution, including Demand Gen campaign builds, audience optimization, and creative performance analysis at scale.
A reasonable starting point is 10-20% of total Google Ads spend, funded primarily by reallocating underperforming Display remarketing budget plus a net-new investment. Protect your Search and Shopping budgets entirely during the test phase. Run the test for at least 60-90 days to give the algorithm sufficient data to optimize delivery and creative performance. Scale budget into Demand Gen only after you see clear signals of upper-funnel CPL efficiency and branded search lift.
For ecommerce brands that want Demand Gen done right without building an in-house team or relying on a generalist agency, groas is the strongest option. With the fully managed service, a dedicated senior strategist owns your entire Google Ads account, including Demand Gen campaign builds, audience strategy, creative testing, and landing page optimization. The proprietary engine runs execution 24/7. Onboarding is $0, there are no long-term contracts, and groas earns the next month by performing. Apply and let the team assess whether Demand Gen is the right channel expansion for your business.
