Demand Gen campaigns can now pull directly from business data feeds, auto-populating ads with product names, images, prices, promotions, and availability. I’ll explain why that got some eyeballs emojis when I shared the news with the JDM team.
Google is repositioning a channel most of us have treated as an awareness and consideration play into something that can assemble inventory-driven ads dynamically and optimize them toward conversions. The change is real, and the opportunity is real, but the thing you actually have to get right is the guardrails you build before you turn the automation loose.
Google now lets Demand Gen campaigns pull from business data feeds, auto-generating ads from your product, service, or use-case data.
This moves Demand Gen from a top-of-funnel tactic toward a lower-funnel commerce channel that could soon compete with Meta Advantage+ and catalog ads.
It is not "Display product ads with a feed." The difference is merchant-style feed automation combined with Demand Gen's YouTube- and social-style optimization and audiences.
B2B and SaaS can use it too, as long as you build a proper business data feed for products, features, or use cases.
Because you’re giving Google more creative and automation control, I would add more guardrails than a standard Demand Gen campaign: a capped test budget, strict feed hygiene, audience and exclusion controls, and conversion-quality monitoring.
For most of its life, Demand Gen has been a consideration channel. You built static creative by hand, pointed it at audiences across YouTube, Discover, and Gmail, and used it to warm people up. Commerce on Google lived mostly in Shopping and Performance Max.
Business data feeds break that pattern. Connect a feed and Google will assemble ads on the fly from your product names, images, prices, promotions, and availability, updating them automatically as your inventory or pricing changes. These business data feeds operate independently of Google Merchant Center, so a Merchant Center account is not required. For an advertiser with hundreds or thousands of products, that removes an operational bottleneck.
The bigger story is directional. Google keeps pushing every campaign type toward AI-driven creative, dynamic assets, and feed-based personalization, and Demand Gen is the latest to get pulled in. My read is that Demand Gen is being groomed into an AI-powered commerce engine, and over the next year it could stand on its own as a legitimate lower-funnel channel rather than a warm-up act.
Feed-based product ads are not new, so it would be easy to file this under "we've seen this before." That framing misses what makes it interesting.
Compare the three mechanisms directly:
In other words, you get feed automation on the input side and visual, social-style ad experiences with Demand Gen audience targeting on the output side. If your product or service catalog is deep and your best-performing placements are visual, this hits a gap the other two formats did not fully cover.
This is a test worth prioritizing under specific conditions:
If a client has a small, static product set or no existing Demand Gen traction, drop it down your priority list because it needs a catalog wide enough and a performance history real enough to make the automation worth the added complexity. Testing it on a thin catalog with no benchmark tells you almost nothing.
B2B and SaaS brands can run this feature, provided you build a proper business data feed. A SaaS product and feature library, or a B2B offering that spans multiple use cases and industries, maps well to feed-driven assembly. The requirement is the feed itself, so if you have never structured your offerings into feed format, that build is your real first step.
Here is where I part ways with the natural instinct. When a channel gets more automated, people tend to relax controls and let the system work. With Demand Gen business data feeds, I would do the opposite. You’re handing Google significantly more creative flexibility and automation than a standard Demand Gen campaign, and the automation won’t supply judgment about your brand, audiences, or conversion quality. That’s your job.
Don’t fold this into your existing Demand Gen campaigns. Carve out a separate test budget at 10 to 20 percent of Demand Gen spend, run it for four to six weeks, and evaluate against your existing benchmark. Contaminating a working campaign with an experiment is how you lose the ability to read either one.
Your feed is now your creative. Audit headlines, descriptions, images, URLs, and landing pages before launch. Remove duplicate offers, outdated products, low-converting services, and generic messaging. Segment feeds by theme so the system is working from coherent sets rather than a jumble. Establish approved brand language up front, because marketing teams routinely forget to define it when they build feeds, and the automation will surface whatever you leave in.
Start audiences narrow and intentional: Customer Match, website visitors, qualified lead lists, similar audiences, and high-intent custom segments. Then set exclusions with the same care. Exclude existing customers (if you’re prospecting), careers traffic, internal IPs, competitors where relevant, and poor-performing placements. Google's automation will not manage that judgment for you.
Conversion quality monitoring is the most important control on the list, and it matters most for B2B. A feed-driven system optimizing toward volume can pile up cheap conversions that never turn into pipeline.
Decide in advance what a good conversion looks like and hold the test to that standard, not to a raw CPA that looks flattering on the surface. For B2B campaigns, track metrics such as cost per qualified lead and lead-to-opportunity rate so you can see whether the additional conversion volume is producing genuine pipeline.
Demand Gen with business data feeds is a genuine step toward a lower-funnel commerce channel, and for advertisers with deep catalogs and existing Demand Gen traction it’s worth testing in time to know if you want to put it in play for Q4. The catch is that the value shows up only if you match Google's added automation with tighter controls on your side: a capped test, clean and themed feeds, deliberate audiences and exclusions, and honest conversion-quality tracking. Test it where the conditions fit, skip it where they don't, and build the guardrails before you flip the switch.
If you're weighing whether to connect a business data feed to Demand Gen or unsure how to structure the guardrails before you test it, JDM can help you build the feed strategy and campaign controls to do it right. Book a strategy call here.
B2B and SaaS brands can use it, as long as you build a proper business data feed. A SaaS feature library or a B2B offering spanning multiple industries and use cases maps well to feed-driven ad assembly. The prerequisite is structuring your products, features, or use cases into a feed, so that build is the real first step if you have never done it.
Display plus a feed delivers intent-lite placements with relatively static product cards, and Performance Max plus a feed is commerce-focused but offers limited creative control. Demand Gen with business data feeds combines merchant-style feed automation with Demand Gen's YouTube- and social-style optimization and audience targeting. The distinction is the combination of feed-driven inputs and visual, social-style ad experiences, which the other two formats did not fully cover.
Run it on a separate, capped budget rather than folding it into existing campaigns. Allocate 10 to 20 percent of your Demand Gen spend, run the test for four to six weeks, and evaluate against your current Demand Gen performance as a benchmark. Keeping it isolated is what lets you read the results of both the test and your existing campaigns cleanly.