Demand Gen is Google’s top-of-funnel campaign type for ecommerce. It puts your products in front of people who aren’t searching for them yet, across YouTube, Discover, Gmail, and other placements.
If you’re already running Search and Shopping campaigns profitably, Demand Gen is how you expand beyond people who already know what they want.
This guide covers how Demand Gen actually works, where your ads show up, how to set it up, and what to realistically expect. I’ll also be honest about where Google’s marketing doesn’t match reality.
What Are Demand Gen Campaigns?
Demand Gen (or Demand Generation) is Google’s latest push into the middle and top funnel of advertising. They reach shoppers across YouTube, Discover, Gmail, Google Network, Maps, and other placements.
Basically, anywhere Google wants to sell you ad space.
Here’s what matters. Google introduced Demand Gen around 2022 as a “discovery” focused product. Awareness-stage advertising. In 2025, they shifted the positioning and added tROAS bidding to make it look like a conversion-driving campaign.
Don’t be fooled by the repositioning.
Demand Gen is a top-of-funnel channel. The placements tell you everything: YouTube in-stream, Shorts, Discover feed, Gmail Promotions tab. These are places where people browse and watch. They’re not searching for your product.
No amount of bidding strategy changes the fundamental nature of the inventory. You’re interrupting people, not answering their questions.
That’s not a criticism. Top of funnel is valuable. But you need to treat it as top of funnel in your expectations, your budget allocation, and your performance benchmarks.
Expect Search-level ROAS? You’ll be disappointed. Expect it to build awareness and feed your lower-funnel campaigns with warmer audiences? It works.
Demand Gen campaigns use automated bidding, creative optimization, and audience signals to show products to potential customers. If you’ve used Performance Max campaigns before, Demand Gen will feel familiar. But there are some critical differences I’ll break down below.
Why Most Advertisers Actually Switched to Demand Gen
Here’s the part Google doesn’t emphasize in their marketing.
The biggest driver of Demand Gen adoption wasn’t that advertisers chose it. They were forced into it.
In 2024, Google sunset Video Action Campaigns (VAC). VAC was the go-to campaign type for running conversion-focused video ads on YouTube. If you wanted to run YouTube ads optimized for purchases or leads, not just views, VAC was your tool.
It was widely used, well understood, and it worked.
When Google killed VAC, Demand Gen became the only conversion-focused video campaign type in Google Ads. There was no alternative. If you were running VAC campaigns, your only option was to migrate to Demand Gen.
This is the real reason Demand Gen adoption exploded. Not because the product was so compelling on its own merits. Because Google eliminated the alternative.
Why does this matter? Because it changes how you should think about Demand Gen.
Google’s marketing will tell you that advertisers are “choosing” Demand Gen in record numbers. Technically true. But the context is that they had no choice if they wanted to keep running conversion-optimized video campaigns.
The good news: Demand Gen is actually a solid product. The audience controls are better than VAC, the product feed integration is a real improvement for ecommerce, and the multi-placement reach gives you more scale.
But don’t confuse forced adoption with organic enthusiasm.
What’s the Difference Between Demand Gen and Discovery Ads?
Good question. This gets confusing because Google basically sunsetted Discovery Ads as a distinct product type. They rolled it into Demand Gen instead.
If you were running Discovery campaigns a few years ago, you know they were limited. They only showed on Google’s own properties like YouTube, Discover feed, and Gmail. They had decent creative control but limited audience options.
Demand Gen is broader. It reaches across YouTube, Discover, Gmail, Google Display Network, Maps, and more. It’s designed to be a one-stop-shop for mid-funnel reach. Google’s pitch: “Why manage multiple campaigns when Demand Gen handles it all?”
My take: The broader reach is useful, but it also means less transparency. You’re relying more on Google’s algorithm to place your ads correctly. More on that below.
Where Demand Gen Ads Show Up
This is important because placement quality varies wildly. Here are the main placements.
YouTube (In-Feed, In-Stream, Shorts)
Your ads can appear in YouTube feeds, the list of recommended videos when someone’s browsing. They show in-stream (before or during videos) and in Shorts. YouTube traffic is typically high-quality for ecommerce, but CPMs are also higher. In-stream placements tend to get views without much intent, so keep an eye on your actual conversion rates.
Discover Feed
The Discover feed is Google’s recommendation engine. You see it as a carousel of content when you open Google News or swipe left on your phone. Placements here get lots of impressions but are lower intent. Users aren’t searching. They’re browsing. Conversion rates are typically lower here than YouTube, but CPM is also lower.
Gmail Promotions Tab
Your ads show up in the Promotions tab of Gmail inboxes. Gmail has a huge captive audience, but the Promotions tab is often ignored. Don’t expect stellar performance here, but if you have strong creative, it can work.
Google Display Network (GDN)
Your ads can show on third-party websites across the web that have enabled Google ads. Quality varies massively. Some GDN placements are great. Others are junk. The problem: Google doesn’t give you granular control over where your ads appear in Demand Gen campaigns. You can exclude categories, but that’s it.
Maps & Other Properties
Google occasionally places Demand Gen ads on Maps and other Google properties. Google is intentionally vague about where exactly your ads show. They call it “expanding placements.” I call it a lack of transparency.
How Demand Gen Works (And How It Changed in 2026)
Demand Gen uses machine learning to optimize placements, creative rotation, and audience targeting. You give Google your assets (videos, images, headlines), your audience signals (lookalike audiences, customer lists), and your budget. The algorithm does the rest.
Here’s the algorithm workflow.
1. Creative Rotation: Google tests different combinations of your images, videos, and headlines. It learns which combinations drive clicks and conversions, then allocates more budget to the winners.
2. Placement Optimization: The algorithm figures out which placements (YouTube, Discover, Gmail, etc.) drive the best results for your specific products and audience.
A big change in 2025 was that now you’re able to select which channels you want to appear in!
3. Audience Expansion: You set audience signals (like “reach people similar to my customers”), and Google expands beyond your specific targeting to find more people like your audience.
The 2025 Shift: In 2025, Google moved Demand Gen’s primary bidding strategy from tCPA to tROAS (target return on ad spend).
This sounds technical, but here’s what it means practically. Google wants you to think of Demand Gen as a conversion-driving campaign. They’re positioning it to compete with Meta and TikTok for performance budgets.
But changing the bidding strategy doesn’t change where your ads show.
You’re still running on YouTube, Discover, and Gmail. These are awareness placements. tROAS bidding just means the algorithm tries to find the most valuable users within those placements.
It’s smarter targeting within a top-of-funnel channel. It’s not turning Demand Gen into Search.
Bidding Strategies: What Google Recommends vs. What Actually Works
Google will recommend tROAS (target return on ad spend) for Demand Gen campaigns. This is their automated bid strategy where you tell Google “I need 3.1 ROAS” and the algorithm adjusts bids to hit that target.
For ecommerce, tROAS is solid. It’s designed to optimize for actual conversions and revenue, not just clicks. But there are caveats.
You need enough conversion volume. Google says you need at least 50 conversions in the past 30 days for tROAS to work properly. Smaller stores that don’t hit this threshold should use a different strategy or wait until they have enough data.
Google will underbid to meet your target. If your target ROAS is too aggressive, Google will reduce spend to hit it. You’ll get fewer conversions but better ROAS. Mathematically, this makes sense. But it’s limiting. Set a realistic target based on your actual business model, not wishful thinking.
Alternative: Maximize Conversions with target CPA. If you have enough conversion data, you can also use target CPA. Google will try to get you conversions at or below your target CPA. This works well if you know your profitable CPA threshold.
Demand Gen vs Performance Max: When to Use Each
Both Demand Gen and Performance Max are heavily automated campaign types. But they do completely different jobs.
Understanding this is critical.
Performance Max is a bottom-of-funnel campaign that pretends to be full-funnel. Google markets PMax as reaching “all of Google’s channels.” Technically true. But in practice, PMax goes as low in the funnel as it possibly can.
It loves branded search queries because those convert easily. It loves Shopping placements because intent is high. Left to its own devices, PMax will spend most of your budget on bottom-funnel placements and take credit for conversions that were already going to happen.
Demand Gen is a top-of-funnel campaign. It reaches people on YouTube, Discover, and Gmail. These people aren’t searching for your product. They’re browsing, watching videos, checking email.
You’re introducing your brand to new audiences and warming them up for later conversion.
They don’t compete with each other. They target different things entirely. PMax doesn’t do top of funnel in any meaningful way. Demand Gen doesn’t compete for the high-intent search and Shopping placements where PMax thrives.
Opposite ends of the funnel.
This means you should run both if your budget allows, and you should measure them differently. Judge PMax on direct ROAS and CPA. Judge Demand Gen on its contribution to the full funnel: branded search lift, new customer acquisition, and overall account performance.
Key Differences Between Demand Gen and Performance Max
Audience Control: Demand Gen gives you more control over audience selection. You can use lookalike audiences, customer lists, and audience signals more granularly. PMax uses “optimized targeting” which is basically Google’s way of saying “we’ll ignore your targeting and do what we want.”
Creative Control: Demand Gen is more transparent about how your creative is used. You can see which image/video/headline combos drive performance, kind of. Google still limits data. PMax generates AI variations of your creative without showing you exactly what’s running.
Placement Control: Demand Gen lets you select which channels to include. You can choose YouTube, Discover, Gmail, or GDN. PMax automatically includes everything and limits your control.
Bidding Transparency: Demand Gen’s bidding logic is slightly more transparent. You can see your tROAS target and roughly understand how bids are adjusted. PMax is more of a black box.
When to Use Demand Gen
Use Demand Gen when you want to reach new audiences who haven’t heard of your brand yet. Or warm up people who’ve shown some initial awareness.
It’s your top-of-funnel play.
Use Demand Gen if your products are visually appealing. Video and image quality matter more here because placements like YouTube Shorts and Discover feed are visual-first.
When to Use Performance Max
Use pMax to capture existing demand. It’s strongest on Search and Shopping placements where people are actively looking for products like yours.
Use pMax if you have limited creative assets. PMax will generate variations and test them for you. Though I’d rather have control over this.
The practical answer: Run both. pMax captures bottom-funnel demand. Demand Gen feeds the top of the funnel with new audiences. They’re working different stages of the buying journey.
For a deeper comparison, see our article on Demand Gen vs Performance Max.
Demand Gen vs Search Ads
Search ads like Standard Shopping target high-intent keywords. Someone searches “buy blue running shoes,” and your ad shows. That’s intent-driven, bottom-funnel advertising.
Demand Gen is different. You’re reaching people who aren’t actively searching for your product. You’re showing them a YouTube video, a Discover recommendation, or a Gmail ad. This is awareness and interest-stage advertising.
In ecommerce, you typically need both. Search ads capture ready-to-buy customers. Demand Gen builds awareness and interest, feeding the top funnel. But they serve different purposes, and you shouldn’t treat them the same.
Here’s a practical take. If your search campaigns are already profitable and scaled, adding Demand Gen makes sense. It reaches new people earlier in their journey. But if your search campaigns aren’t profitable yet, fix those first. Don’t add complexity with Demand Gen until your core paid search is working.
Budget Requirements: What Google Says vs. What Actually Works
Google recommends a minimum of $100 per day for Demand Gen campaigns. Or 20x your target CPA. So if your target CPA is $10, that’s $200 per day minimum.
Here’s my honest take. That’s expensive for a lot of ecommerce stores. If you’re a small store with a $500 total daily ad budget, dedicating $100+ to an experimental campaign is risky. You’re essentially gambling $30,000 a month on something unproven.
The real issue: Google’s recommended budgets are based on statistical significance. They want enough data to train the algorithm properly. $100 per day gives them 3,000+ impressions daily, which is enough for their machine learning to find patterns.
Can you run Demand Gen with $50 per day. Technically yes. But the algorithm will struggle to optimize effectively. You’ll get more variability in results, and the “learning period” will be longer.
My suggestion: Only run Demand Gen if you can sustainably spend $100+ per day. If you’re under that budget, focus on Search and Performance Max first. When you have sufficient budget and stable conversion volume, add Demand Gen.
Key Features for Ecommerce Stores
Product Feeds
Linking your Google Merchant Center product feed to Demand Gen is important. It tells Google about your inventory, pricing, and availability. Google can then show the right product to the right person.
Without a feed, Demand Gen still works but it’s flying blind. Google doesn’t know your product details. With a feed, the algorithm optimizes smarter. Learn more about product feed optimization.
Lookalike Audiences
Demand Gen lets you create lookalike audiences based on your website visitors and customers. This is powerful because you’re reaching people similar to your best customers, not random people on the internet.
Use this feature. Lookalike audiences typically outperform cold audiences in Demand Gen campaigns. For more on audience strategy, see our guide to Demand Gen audiences.
Customer Lists
You can upload customer emails and phone numbers. Google matches them to user accounts and treats them specially. You can exclude existing customers so you don’t waste budget re-acquiring them. Or bid more aggressively toward them.
Use Customer Match to build these lists, and leverage Audience Manager to organize them.
Channel-Level Controls
You can select which channels to include. You have YouTube, Discover, Gmail, GDN, Maps, and more. Start with YouTube and Discover only, then expand once you have performance data. GDN and Gmail often underperform for ecommerce, so don’t assume you need them.
Shoppable Ads (Connected TV)
Demand Gen supports “shoppable CTV” ads on connected televisions. Someone watches a video on their TV, and your product appears. They click through to your site using their remote.
This is technically cool but practically limited. Most people don’t click ads on their TV. CTV is great for brand awareness, not conversion. Don’t budget for this expecting a strong ROI. Learn more about Demand Gen for YouTube and CTV.
Common Mistakes I See Ecommerce Stores Make
Mistake 1: Running Demand Gen Like Performance Max
Some stores treat Demand Gen like pMax. They set it and forget it.
But because Demand Gen goes after specific audiences, you need to stay on top of what createive is working for which audience, which means constant monitoring and tweaking.
Mistake 2: Budget Too Low for the Learning Period
Demand Gen needs time to learn. Google recommends 2-4 weeks of data before making optimization decisions. If your budget is too low, you won’t get enough conversions per week for the algorithm to learn properly. Result: Poor performance and false conclusions.
Mistake 3: Ignoring Creative Quality
Demand Gen placements like YouTube Shorts and Discover are visual-first. If your images are blurry or your videos are boring, you’re wasting budget. Invest in actual creative production, not just product photos. See our tips on Demand Gen ad creative.
Mistake 4: Not Using Product Feeds
Connecting your Merchant Center feed is optional but should be required. Without it, Demand Gen can’t show specific products or price points. The algorithm has to work harder.
Mistake 5: Including All Channels From Day One
Google defaults to YouTube plus Discover plus Gmail plus GDN plus Maps. Don’t accept the defaults. Start with YouTube and Discover only. Once you see positive ROAS there, add other channels. This reduces wasted spend on low-performing placements.
Mistake 6: Expecting Bottom-Funnel Performance
This is the most common mistake I see.
Advertisers launch Demand Gen, look at the direct ROAS after two weeks, and conclude it doesn’t work.
But you’re measuring a top-of-funnel channel with bottom-of-funnel metrics.
Demand Gen’s job is to introduce your brand to new people and warm them up. The conversion often happens later through a branded search or a remarketing ad.
If you only look at last-click ROAS, Demand Gen will always look weak compared to Search or Shopping. That’s not a failure. That’s how top of funnel works.
Mistake 7: Unrealistic ROAS Targets
If you set a target ROAS of 5:1 but your business model only supports 3:1, Google will underbid and reduce spend. The campaign will look “profitable” because you hit your target, but you’ll miss revenue opportunities. Set ROAS targets based on your actual business metrics, not wishful thinking.
Setting Up Your First Demand Gen Campaign
I’ve written a full step-by-step guide here: How to Set Up a Demand Gen Campaign
But here’s the high-level process.
1. Prerequisites: Ensure you have conversion tracking set up, your product feed in Merchant Center, creative assets ready (images, videos, headlines), and a sufficient daily budget of $100 or more.
2. Create Campaign: In Google Ads, select “Demand Gen” as your campaign type. Set your objective usually “Sales” for ecommerce.
3. Campaign Settings: Choose your bidding strategy (recommend tROAS for ecommerce), set your daily budget, and target locations.
4. Audience Settings: Create audience signals (lookalike audiences, customer lists), and decide on your targeting strategy.
5. Channel Selection: Choose which channels to include. Start narrow: YouTube and Discover only.
6. Create Ads: Upload your image and video creative, write headlines, and link your product feed.
7. Review and Launch: Double-check everything and launch.
I’ll walk through each step in detail in the setup guide.
Is Demand Gen Worth It for Your Store?
Here’s my honest answer: It depends.
Demand Gen is worth it if: You have a daily ad budget of $100 or more. You have a catalog of visually appealing products. You’ve already optimized your search and Shopping campaigns. You want to reach lookalike audiences and build brand awareness. You have 6+ months of conversion data to feed the algorithm.
Demand Gen is NOT worth it if: Your daily budget is under $75. You’re new to Google Ads and still learning. Your products aren’t visually interesting like B2B software or dry goods. You haven’t stabilized your search campaigns yet. You don’t have creative resources like photography or videography.
My take: Demand Gen is a solid top of the funnel channel, but it’s not a replacement for Search or Shopping ads. It’s an addition. And it’s not for everyone. If you’re a bootstrapped ecommerce store with a $1,000 per month ad budget, you’re better off doubling down on Search and Shopping. If you’re a mid-size store with $10,000+ per month to spend, Demand Gen is worth testing.
The key question: Can you afford to spend $100-200 per day for 4-6 weeks while the algorithm learns. If yes, try it. If no, wait until you scale.
Check out Demand Gen benchmarks to see how your performance compares to industry standards.
Final Thoughts
Google’s Demand Gen is a powerful tool, but it’s not magic. It requires proper setup, adequate budget, good creative, and realistic expectations. And don’t believe Google’s benchmarks uncritically. When Google tells you Demand Gen drives “26% more conversions per dollar,” remember: That’s Google’s data from campaigns they selected. They have every incentive to make Demand Gen look good.
The real test is your own campaigns. Set up Demand Gen properly, give it time, and measure honestly. If it works for your products and audience, scale it. If it doesn’t, move on to other channels.
For a broader view of Google Ads strategy, remember that Demand Gen is one piece of your overall paid advertising plan.
