Google Opens AI Video Ad Production to Every Advertiser

Google made its AI video production workflow generally available to every Demand Gen advertiser on Thursday.

For a high-ticket store owner, that one line moves a real cost off the books. Video has been the wall between independent operators and YouTube inventory, because nobody selling $3,000 saunas out of a one-person office is commissioning a videographer for six product lines. Multimodal Video Creation in Asset Studio now takes a storyboard and produces horizontal and vertical assets inside Google Ads, per Google’s August Demand Gen Drop announcement.

The video tool is one of three items in the drop, and it is the only one that is live for everyone today. The other two are a messaging test on YouTube and a set of travel features. I have been running Google Ads for high-ticket stores since before Demand Gen existed, and readers of Ecommerce Paradise know I am slow to get excited about Google shipping automation. This one lands differently, and the reason has more to do with what Google is taking away in January than what it gave away on Thursday.

Google has retired four campaign types in three years. Your registered agent should not be the next thing that changes on you. See why I still use Northwest after 25 years in business →

Google’s Demand Gen Drop Opens AI Video Creation to All Advertisers

Google published the August Demand Gen Drop on August 27, 2026, its twelfth in the series. Three changes were announced.

Multimodal Video Creation in Asset Studio moved to general availability. Google said advertisers “can now go from storyboarding to creating horizontal and vertical assets in one workflow, enabling them to show up across all Demand Gen inventory.” Before this, the workflow was limited, and producing the multiple aspect ratios Demand Gen inventory demands meant either an agency, a freelancer, or skipping the channel.

Second, Google is testing a way for viewers to start conversations with brands on messaging apps directly from Demand Gen ads on YouTube. Google framed it as “moving high-intent customers closer to purchase.” Search Engine Land’s paid media editor Anu Adegbola reported that for lead generation advertisers the test “introduces another potential action beyond sending users to a website or landing page.” It is a test, not a launch, and Google gave no timeline for wider availability.

Third, Google added travel-specific inventory features, surfacing local activities, events and real-time offers, and personalizing hotel ads by audience. That one does not touch a furniture or outdoor gear store.

The number Google put on the drop is a 30% average increase in conversions or conversion value. Read the footnote before you plan a budget around it. Google attributes that figure to “hundreds of improvements” made to Demand Gen during the second half of 2025, and sources it to “Google Internal Data, Global, H2 2025, DemandGen Experiment Results.” It is unaudited, it is Google measuring Google, and it describes last year’s changes rather than the three announced Thursday. MediaPost and Search Engine Land both flagged it as Google’s own footnoted number rather than an independent measurement.

Separately, Google said this week that better YouTube creative can more than double return on investment, another claim resting on Google’s own data. Barry Schwartz logged both items in the Search Engine Roundtable recap for August 28.

How the Display Network Shutdown Pushed Retargeting Into Demand Gen

The video tool matters because of a decision Google made in May. Standalone Display campaigns are being deprecated, joining Smart Shopping, Discovery and Local campaigns on the retired list.

Sarah Vlietstra laid out the timeline in Search Engine Land the same day the drop landed. Google released a migration tool that started appearing in accounts in June. In January 2027, the ability to create new standalone Display campaigns is permanently disabled. Auto-migration is anticipated later in 2027, and Google has not said when it begins.

The migration itself is reasonable on paper. Google preserves the last 42 days of performance history so the new Demand Gen campaign does not restart its learning period cold. The tool creates a Display-network-only Demand Gen campaign, and you add other channels afterward.

The trade-offs are where this gets expensive, and they are the counterpoint to every optimistic number in Google’s announcement. Demand Gen supports Target CPA, Target ROAS and Maximize Clicks. If you are running Manual CPC, Viewable Impressions or Pay for Conversions, those are gone. Bid adjustments, seasonality adjustments and portfolio bidding are not supported at all. Business Data feeds, uploaded HTML5 ads and third-party ads are unsupported, with the last two planned for late 2026 according to Google. Brand Lift and Search Lift studies do not work for Display inventory inside Demand Gen. Vlietstra also noted that budget already spent on migration day is not respected in the new campaign.

Two more changes worth knowing. In March 2026, Lookalike Audiences stopped being restrictive targeting and became “audience suggestions,” meaning your seed list is now a signal rather than a boundary, and a new Lookalike segment carries up to 96 hours of latency before it populates. In April, Google introduced view-through-conversion optimization in Demand Gen, then announced in July that Demand Gen campaigns running on Discover with that setting move from CPC to CPM billing.

That billing change is the same move I broke down when Google started charging for impressions in July. The pattern across all of it is consistent: more inventory, more automation, fewer levers.

The precedent is worth stating plainly, because it tells you what happens next. Smart Shopping went first, folded into Performance Max. Discovery campaigns became Demand Gen. Local campaigns disappeared into Performance Max as well. In every case the sequence was identical: a migration tool, a soft deadline, a hard deadline, then auto-migration for anyone who waited. Advertisers who moved early kept their history and picked their settings. Advertisers who waited got Google’s defaults during their busiest quarter. Display is the fourth run of the same play, and the calendar is already published.

What Demand Gen Video Means for High-Ticket Dropshipping Margins

My read is that the creative unlock is real and the performance claim is not something you should underwrite.

Run the math on the old way. A single product video for a $4,000 hot tub, shot properly with a freelancer, runs $800 to $2,500 and takes two to four weeks. You need horizontal for in-stream, vertical for Shorts, and usually a square variant. Call it $1,200 for one product done cheaply. If you carry 40 SKUs across three brands and you want video on the top ten, that is $12,000 and a quarter of calendar time before a single impression serves. That is exactly why almost every high-ticket store I look at runs Shopping and Search only, and treats YouTube as something other people do.

Asset Studio takes the production line item to roughly zero and the timeline to an afternoon. On my own stores I have been generating product imagery this way for a while, and I wrote up the whole process in my guide to using AI for product images and UGC videos. Video was the piece that stayed stubborn. It is not stubborn anymore.

Here is where I would push back on my own enthusiasm. AI-assembled video from a storyboard is going to be competent, not persuasive. High-ticket buyers are usually 45 to 70 years old, they are spending real money, and they convert on trust signals: a human demonstrating the product, a phone number, a warranty explained by someone who sounds like they have handled the thing. Generated b-roll set to stock music does not do that. What it does well is mid-funnel reminder creative, which is the actual job for most high-ticket retargeting anyway.

So the honest framing is this. If you are doing under 30 orders a month, your bottleneck is not creative, it is Shopping feed quality and search term hygiene, and you should fix that first using the three-tier Shopping structure I use on client accounts. Over 30 orders a month, you have enough remarketing volume that a Demand Gen campaign with generated video is worth a $30 to $50 per day test. Below that threshold, Demand Gen will spend your money teaching itself.

The messaging test is the item I am watching hardest. High-ticket sells on the phone. My rule for fifteen years has been that your number goes on every page, because a $4,000 purchase generates questions no product page answers. If Google ships a path from a YouTube ad straight into a text thread with your store, that maps onto how these sales actually close. A tool like Grasshopper gives you a business number that routes to your cell, which is what you want before that inventory opens up, not after.

Where this gets heavy is workload. You are now expected to manage Shopping, Search, a migrating Display budget, Demand Gen creative refreshes, and a feed, on top of supplier onboarding and order processing. That is a full-time job that has nothing to do with growing the business. If you would rather have a team that already runs this stack every day, that is exactly what my done-for-you turnkey service exists for.

New to high-ticket and wondering where ads even fit in the build order? Grab my free beginner’s guide →

How to Prep Your High-Ticket Store’s Demand Gen Campaigns Before Q4

Five things worth doing in the next seven days, in order.

Check whether the migration tool is in your account. Open Google Ads and look at any live Display campaign. If the tool is there, migrating now preserves your last 42 days of history and gives you runway before Q4 volume hits. Migrating in November does not.
Audit your bid strategies before you touch anything. If any Display campaign uses Manual CPC, Viewable Impressions or Pay for Conversions, pick a replacement from Target CPA, Target ROAS or Maximize Clicks and let it stabilize on the old campaign first. Migrating and changing bid strategy in the same week gives you no way to read the result.
Build your Lookalike seed lists this week. New segments take up to 96 hours to populate, and you want purchasers and quote-requesters seeded well before you need them. Pull the customer list out of Shopify and push it into both Google Ads and your email platform.
Generate three test videos in Asset Studio, not thirty. Pick your highest-margin SKU, produce a horizontal and a vertical, and run them against a warm remarketing audience only. Watch cost per quote, not view rate. If you want a second opinion on which SKU to lead with and what the target should be, that is the kind of thing I work through on one-on-one coaching calls.
Fix your measurement before you add a channel. Demand Gen is view-through heavy and it will claim credit generously. Tag your landing pages properly, keep Omnisend flows attributing separately, and pull organic and paid apart in SEMrush so you can tell what the new channel actually added.

If your creative production is the constraint rather than the strategy, hire it out instead of buying software. A part-time editor from OnlineJobs.ph running Asset Studio outputs through simple brand overlays will beat raw generated video, and costs less than one freelance shoot per month.

Frequently Asked Questions

Do I have to migrate my Display campaigns right now?
No. Creating new standalone Display campaigns is disabled in January 2027, and Google says auto-migration comes later in 2027 without giving a date. Doing it on your own schedule before Q4 is better than having it done to you mid-season.

Is the 30% conversion lift number reliable?
Treat it as marketing. It is Google’s own internal experiment data, it covers changes made in the second half of 2025, and it is not an independent measurement of the features announced this week.

Will AI-generated video actually convert high-ticket buyers?
For cold prospecting, probably not on its own. For retargeting someone who already viewed a $3,000 product page, it is good enough and it costs nothing to produce.

Can I still use my existing HTML5 banner ads?
Not in Demand Gen today. Uploaded HTML5 and third-party ads are unsupported, and Google says both are planned for late 2026.

Should a brand new store run Demand Gen at all?
No. Get Shopping profitable first, then add remarketing. My complete Shopping ads setup guide is the right starting point.

What tools help me edit the generated video?
Canva handles overlays and captions fine for most stores, and InVideo is worth a look if you want templated variants at volume.

How do I know if my niche can support this spend at all?
Check whether your average order value clears $800 and your gross margin clears 20%. My high ticket niches list breaks down which categories clear that bar.

Want my full step-by-step masterclass? Get the masterclass →

Google gave away the expensive part of Demand Gen and kept the part that decides whether it works. Test it small, measure it honestly, and do not let a self-reported 30% talk you into a budget you have not earned back. Subscribe to the YouTube channel for daily breakdowns. More breaking news tomorrow morning.

If this was useful, these go deeper:

Trevor Fenner

Trevor Fenner is an ecommerce entrepreneur and the founder of Ecommerce Paradise, a platform focused on helping entrepreneurs build and scale profitable high-ticket ecommerce and dropshipping businesses. With over a decade of hands-on experience, Trevor specializes in high-ticket dropshipping strategy, niche and product selection, supplier recruiting and onboarding, Google & Bing Shopping ads, ecommerce SEO, and systems-driven automation and scaling. Through Ecommerce Paradise, he provides free education via in-depth guides like How to Start High-Ticket Dropshipping, advanced training through the High-Ticket Dropshipping Masterclass, and fully done-for-you turnkey ecommerce services for entrepreneurs who want a faster, more hands-off path to growth. Trevor is known for emphasizing sustainable, real-world ecommerce models over hype-driven tactics, helping store owners build scalable, sellable, and location-independent brands.