Every account that relies entirely on search advertising faces the same ceiling. You can only capture the demand that already exists. Once you have a reasonable share of the searches people are actively making for your category, growth requires either outbidding competitors for the same pool of queries or finding customers before they start searching.
Demand Gen is built for the second option.
It is the most underestimated campaign type in Google Ads. Most advertisers either skip it entirely, use it incorrectly as a bottom-funnel conversion driver, or try it briefly, find the last-click attribution numbers disappointing, and conclude it does not work. The ones who build it into a proper upper-funnel strategy find it does something no Search campaign can do: it creates demand rather than capturing it.
This guide covers what Demand Gen actually is, how it fits into a full-funnel account, what the research says about its real performance, and the specific setup decisions that determine whether it becomes a growth channel or a budget drain.
What Demand Gen Is and Where It Came From
Demand Gen was launched by Google in October 2023 as a successor to Discovery campaigns, absorbing Video Action Campaigns in early 2025 and becoming the single home for Google's visual, top-of-funnel inventory. It runs across YouTube (in-stream, in-feed, Shorts), Google Discover, and Gmail, with Google Display Network inventory added in 2025.
The campaign type reaches up to three billion monthly users across these surfaces. The core distinction from Search is audience-driven rather than keyword-driven targeting. Users are not searching for your product when they see a Demand Gen ad. They are watching a YouTube video, browsing their Discover feed, or checking Gmail. Your ad appears because Google's AI has identified them as likely to be interested based on their search history, content consumption, app behavior, and demographic characteristics.
This is also why Demand Gen is often misunderstood. Advertisers trained on Search think about advertising in terms of capturing intent. Someone types a query, you show an ad, they convert. Demand Gen does not work this way. It works more like good social advertising, placing your message in front of the right people before they have formed an active search intent, creating the awareness that eventually leads to a search.
As one practitioner put it in the 2026 Demand Gen playbook from Search Engine Land: "Demand Gen is here to create intent. Performance Max is here to capture it. When you let each one do its job, everything works better."
The Attribution Problem That Makes Demand Gen Look Worse Than It Is
The most common mistake when evaluating Demand Gen is using last-click attribution.
Last-click attribution gives 100 percent of conversion credit to the last ad interaction before conversion. If a user sees your Demand Gen YouTube ad on Tuesday, does some research on Thursday, and clicks your Search ad to convert on Friday, last-click gives all the credit to the Search campaign. The Demand Gen campaign that created the initial awareness appears to have zero contribution to that conversion.
This is how Demand Gen gets written off. The last-click numbers look bad. Leadership asks why you are spending money on campaigns that are not converting. You pause Demand Gen. Search costs gradually increase as more competitors bid for the same limited pool of high-intent queries that remain.
The reality is that Demand Gen's contribution is real but upstream. It operates on a timeline that last-click cannot see. Users who have been exposed to your brand on YouTube or Discover convert at higher rates when they later search on Google. The brand familiarity built by Demand Gen reduces the friction in Search campaigns. The pipeline that converts through Search was often seeded by Demand Gen.
Measuring this properly requires moving away from last-click toward data-driven attribution, which distributes credit across touchpoints based on actual conversion path analysis. It also helps to use Google's platform comparables metric, which uses view-through methodology similar to how social ads are evaluated. And it helps to run Demand Gen audience segments in observation mode on your Search campaigns, where you can track whether users who were previously exposed to Demand Gen ads convert at different rates on Search.
Marks and Spencer, running Demand Gen as part of a full-funnel strategy, exceeded their forecasted ROAS by 186 percent and achieved a 66 percent lower CPA compared to their paid social campaigns. That is the ceiling available when Demand Gen is measured and managed correctly.
Where Demand Gen Fits: The Power Pack
Google's recommended campaign architecture for 2026 is the Power Pack: Demand Gen to create awareness and intent, AI Max for Search to capture that intent when users search, and Performance Max to orchestrate full-funnel performance at scale.
Each campaign type does a distinct job. Demand Gen introduces your brand and offer to people who are not yet searching. When those people subsequently search, AI Max or standard Search campaigns are there to convert them. Performance Max handles the middle and lower funnel across multiple channels simultaneously.
The strategic implication is that Demand Gen should not be evaluated in isolation. Its value partially shows up in better performance from your Search campaigns as brand awareness builds. Accounts that run Demand Gen alongside Search typically see improved Search CTR for branded and category terms over time, because they are showing up to audiences who have already encountered the brand.
Demand Gen also makes sense as the first campaign type to launch when entering a new market or launching a new product. Search captures existing demand, but if your category is not established or your brand is unknown in a market, there is no existing demand to capture. Demand Gen builds the awareness that creates future search volume.
The Creative Requirement Is Real and Non-Negotiable
Demand Gen's performance ceiling is determined by creative quality more than any other single factor. This is different from Search, where ad copy matters but clever structure and keyword alignment can compensate for mediocre writing. In Demand Gen, if your video or image assets do not immediately capture attention and hold it, the campaign will not perform regardless of how well you have structured everything else.
Advertisers who uploaded both video and image assets to Demand Gen saw 20 percent more conversions at the same cost per action than those who uploaded only video assets, according to Google's internal data. Adding product feeds to Demand Gen campaigns produces a 33 percent increase in conversions without an increase in cost per action.
The creative hierarchy for Demand Gen in 2026 looks like this.
Video is the most powerful format but requires genuine production investment. A YouTube in-stream ad needs to hook the viewer in the first five seconds or they will skip it. The hook cannot be generic. It has to address something the target audience genuinely cares about, create curiosity, or deliver an immediate surprise. Vertical video for YouTube Shorts has a different grammar than landscape for standard in-stream. The mobile format requires faster pacing, text overlays for sound-off viewing, and a format that works in the full-screen vertical viewport.
A strong Demand Gen creative library includes at minimum one landscape video (16:9) for standard in-stream placements, one square video (1:1) for Discover and in-feed, one vertical video (9:16) for Shorts, and multiple static image ads in all three aspect ratios. Without the full format range, you are excluded from certain placements. YouTube Shorts, which drives significant reach particularly with younger audiences, requires vertical format.
On the image side, the quality bar is higher than Display advertising. These ads appear in the Discover feed alongside premium editorial content and on YouTube. An ad that looks like stock photography will be ignored. Ads that look like they belong in the environment, native in style and visual quality, attract significantly better engagement.
Audience Strategy: Prospecting and Remarketing Are Different Campaigns
One of the most consistent mistakes in Demand Gen setup is mixing prospecting and remarketing audiences into the same campaign. They have different objectives, different metrics, and different optimal bidding strategies. Running them together makes it impossible to optimize either effectively.
Remarketing audiences in Demand Gen are users who have already visited your website, viewed your YouTube channel, or are on your Customer Match lists. These users already know your brand. The job of remarketing Demand Gen is to re-engage them, reinforce your message, and nudge them back into consideration. Because they are already aware of you, conversion rates tend to be higher and attribution is more straightforward. Remarketing Demand Gen typically shows better last-click numbers, which is one reason it tends to get all the budget when the two are mixed.
Prospecting audiences are cold or warm audiences who have not yet encountered your brand. These are custom segments based on search behavior, in-market audiences, or lookalike audiences built from your Customer Match lists. Prospecting Demand Gen has longer time horizons to conversion and lower immediate conversion rates, but it is where actual new demand comes from. If your remarketing campaign runs out of new users to re-engage, it is because prospecting has stopped feeding them into the funnel.
Keep these as separate campaigns with separate budgets, separate audience segments, and separate evaluation criteria. Give prospecting at least 60 to 90 days before evaluating performance, because the conversion cycle from first exposure to eventual search-and-convert can take weeks.
Lookalike audiences, available in Demand Gen and not in most other Google campaign types, are particularly valuable for prospecting. They allow you to build targeting based on behavioral similarity to your best existing customers. The quality of your lookalike audience depends on the quality of the seed list, which is another reason to maintain clean, segmented Customer Match data.
Channel Controls and Placement Strategy
A 2025 update gave Demand Gen advertisers the ability to specify which surfaces their ads appear on: YouTube, YouTube Shorts, Discover, Gmail, and now Google Display Network. This is an important control that most advertisers leave at the default.
The default runs across all available surfaces, which is fine as a starting point but produces blended performance data that makes optimization difficult. A YouTube in-stream ad and a Gmail promotional ad perform very differently, cost differently, and serve different purposes. Running them in the same campaign and measuring aggregate conversion rates makes it impossible to know which surface is actually working.
The approach that produces the clearest data is starting with Google-owned properties, primarily YouTube, Discover, and Gmail, before expanding to Display. YouTube tends to have the highest-quality inventory and the strongest user engagement. Discover and Gmail add reach at typically lower CPMs. Display extends reach significantly but with more variable quality.
Google Maps was added as a Demand Gen placement in early 2026, giving location-based businesses promoted pin visibility at moments of local intent. For businesses with physical locations, this is worth testing as a dedicated placement-specific campaign.
Bidding for Demand Gen: Set Expectations Before You Set Targets
Demand Gen supports Maximize Conversions and Target CPA or Target ROAS bidding, the same Smart Bidding strategies available elsewhere. The right choice depends on where you are in the campaign lifecycle.
Starting with Maximize Conversions for the first 30 days allows the campaign to generate data without a restrictive efficiency constraint. The algorithm explores more broadly, learns which audiences and creative combinations drive results, and builds the foundation for subsequent optimization.
After 30 days and with sufficient conversion data, moving to Target CPA or Target ROAS makes sense if your primary goal is efficiency. If your primary goal is reach and brand awareness, Maximize Conversions with a budget cap continues to be the right choice, because applying a CPA constraint will reduce impression volume in exchange for conversion efficiency, which works against the campaign's upper-funnel purpose.
Be realistic about what conversion events to optimize toward. For a prospecting Demand Gen campaign, optimizing toward demo requests is probably setting an unrealistically tight conversion constraint. The user who has just seen your YouTube ad for the first time is not ready to request a demo. Optimize toward micro-conversions instead: video views, landing page visits, email signups, or content downloads. These are meaningful signals of engagement that the algorithm can find at the right scale.
The 30-Day Rule and Why You Cannot Shortcut It
Demand Gen campaigns need at least 30 days of data before you can draw any meaningful conclusions. This is not a soft guideline. The algorithm needs time to learn which audience segments, creative combinations, and placement contexts produce results. During the learning phase, which typically runs the first two weeks, performance can be erratic and the cost-per-action numbers will be higher than they will eventually stabilize to.
Making significant changes, pausing the campaign, or slashing the budget during the first 30 days resets whatever learning has occurred. This is the pattern that gets Demand Gen dismissed as ineffective. The advertiser launches the campaign, sees disappointing week-one numbers, makes changes, sees different disappointing numbers, makes more changes, and eventually concludes that Demand Gen just does not work for their business. What actually happened is that the campaign never got a chance to learn.
Budget through the learning period. Commit to a 60 to 90-day evaluation window before making strategic decisions. Build patience into stakeholder expectations before the campaign launches so that early-phase performance data is interpreted correctly rather than triggering premature intervention.
Measuring Demand Gen the Right Way
Given that last-click attribution undersells Demand Gen's contribution, here is the measurement framework that produces a more accurate picture.
Data-driven attribution is the starting point. In your Google Ads settings, ensure your account uses data-driven attribution as the default model. This distributes conversion credit across all touchpoints in the path rather than giving it entirely to the last click.
Platform comparables, available in Demand Gen reporting, use view-through methodology consistent with how social media platforms report ad performance. This metric captures conversions where a user saw your ad but did not click, then converted later. For upper-funnel campaigns, this is often where a significant portion of the real value lives.
Search term trend monitoring is a useful leading indicator. If your Demand Gen campaign is building genuine brand awareness, you should see increases in branded search volume over time. Adding your Demand Gen audience segments to Search campaigns in observation mode allows you to track whether users previously exposed to Demand Gen ads are converting on Search at higher rates.
Incrementality testing is the most rigorous measurement method. By running Demand Gen for one geographic region or audience segment while using a holdout group in a similar region or segment, you can measure the true incremental effect of the campaign, what conversions happened as a direct result of the Demand Gen exposure that would not have happened otherwise. Google's Conversion Lift studies support this kind of testing directly within the platform.
When Demand Gen Is and Is Not the Right Investment
Demand Gen makes sense when you have already captured a reasonable share of existing Search demand and need to grow the pool of potential customers. It makes sense when you are entering a new market where brand recognition needs to be built before people will search for you. It makes sense for product categories where purchase decisions are made before the search phase, where creative storytelling can influence preferences before intent has formed.
It does not make sense as your first campaign type if you are just starting with Google Ads and have not yet established Search campaigns. The right sequencing is to capture existing demand with Search first, build conversion data, and then add Demand Gen as an upper-funnel layer. Running Demand Gen without Search to capture the downstream intent it creates is like running television ads without any stores to buy the product in.
It also does not make sense if you cannot invest in quality creative. A Demand Gen campaign with poor video assets will underperform and produce data that suggests the campaign type itself does not work, when the actual problem is that the creative is not good enough to capture attention in a visual content environment. If you cannot fund proper video production and maintain a library of multiple formats, hold off until you can.
When the creative investment is there, the audience strategy is right, and the measurement framework reflects how upper-funnel marketing actually works, Demand Gen is one of the most powerful and underutilized tools available to Google Ads advertisers.
Running Demand Gen alongside Search and Performance Max requires keeping a close eye on how each campaign type contributes to the full funnel. ClickHub's cross-campaign view connects the dots between your upper-funnel activity and downstream conversions, so you can see what your Demand Gen spend is actually producing.