ChatGPT Ads Adds Bulk Product Campaigns via Spreadsheet

OpenAI added bulk product campaigns to ChatGPT Ads on September 30, letting advertisers build many product-feed campaigns from one spreadsheet upload in Ads Manager.

For the high-ticket store owners at Ecommerce Paradise, this removes the chore that kept ChatGPT product ads a side project. If your catalog has a few hundred SKUs, you can now test a new traffic source this week without building every ad group by hand and without the $200,000 commitments OpenAI asked of early advertisers. It also means you can lose money faster, which is why the margin math below matters more than the feature.

Below: what OpenAI shipped, how the program went from invite-only to spreadsheet uploads in under eight months, what early advertisers have reported, and a test plan sized for a store that sells $1,000-plus products. If you are new to the model, start with my guide to what high-ticket dropshipping is.

Affiliate disclosure: some links in this article are affiliate links and I may earn a commission at no cost to you. It does not change which tools I recommend or what the research found.

Opening a new ad account? Open it under an LLC you formed the right way, with Northwest. Northwest Registered Agent is the LLC formation service I point ecommerce owners to, and because the address on an LLC filing is where plaintiffs and regulators look, its registered agent address keeps yours off the public record. Form your LLC with Northwest →

ChatGPT Ads Bulk Uploads Turn One Spreadsheet Into Many Campaigns

According to Search Engine Land’s September 30 report by Anu Adegbola, advertisers can now “create multiple product feed campaigns and ad groups through spreadsheet uploads in Ads Manager, with ad templates generated automatically.”

Search Engine Roundtable, also on September 30, put the path at Campaigns, then Bulk uploads.

The release carried four other changes, per Search Engine Roundtable. A new Products tab shows the review status of each item, so advertisers can see where a product sits in approval and why one was rejected. The public Insights API now returns attributed conversion events beyond a campaign’s selected goal, without touching the optimization setting.

Advertisers can also keep an internal account name, a public brand name and a legal billing name separately, under Settings, then General. And ChatGPT Ads launched in seven new markets: Indonesia, Malaysia, the Philippines, Singapore, Thailand, Vietnam and Taiwan. That brings the total to 67 countries.

Search Engine Land added one constraint. A brand name change triggers review and pauses ad serving until approval.

Why the product-feed format matters: per Search Engine Land, product-feed ads have been among the strongest-performing ads in OpenAI’s program to date. OpenAI has not published numbers behind that. In a September 8 analysis for Search Engine Journal, Brooke Osmundson, Director of Growth Marketing at Smith Micro Software, wrote that OpenAI states it “does not yet have performance benchmarks across advertisers, industries or campaign types.”

US store owners get the feed through a Shopify app. OpenAI’s help center page says the ChatGPT Ads for Shopify app connects a store’s product catalog and commerce-event data to Ads Manager and is currently available to all US-based Shopify merchants. It syncs inventory, catalog information, store updates and commerce events through the OpenAI Pixel.

The same page flags a limit of the app. Ads Manager offers more control, for example budgets of $28 instead of fixed $25 increments.

ChatGPT Ads Went From $200K Minimums to Self-Serve in 2026

On February 2, 2026, eMarketer reported that OpenAI was asking select advertisers for upfront commitments of roughly $200,000, with proposals ranging from $100,000 to $250,000. Ads were priced at $60 per thousand impressions rather than per click. Test campaigns ran from early February through March, with retail, streaming and internet services as the focus categories. eMarketer put ChatGPT at about 900 million weekly users.

On May 5, OpenAI announced a self-serve Ads Manager in beta, cost-per-click bidding, and a Conversions API plus pixel tracking. Agency partners named in the release included Dentsu, Omnicom, Publicis and WPP.

On September 4, Search Engine Land reported product-level reporting inside carousel ads, product-level fields in the Insights API, and a plugin that lets advertisers create campaigns in plain language from inside ChatGPT. That update also expanded the program to Europe, India, the Middle East and North Africa.

The numbers arrived on September 29. Graphite’s “The State of ChatGPT Ads” report, built on Similarweb panel data and covered by Search Engine Journal, put the US click-through rate at 0.73%. Brazil led at 1.91%, followed by Mexico at 1.55% and Japan at 1.53%. In the US between March 30 and August 4, retail and consumer brands held 10.4% of impressions across 1,510 advertisers, led by Home Depot. B2B software held 19.9% across 1,210 advertisers, led by Monday.com.

Search Engine Land’s September 29 write-up added that 43% of ad-eligible US users had seen an ad within six months and that the business had reached a $1 billion annualized revenue run rate. It described ChatGPT as an incremental channel, not a replacement for Google Search. I covered the Graphite data yesterday in my breakdown of the 0.73% US click-through rate.

Early advertisers disagree on whether the channel pays. Per the Search Engine Journal analysis, Common Thread Collective reported a $4.41 cost per click, a 0.94% click-through rate and an estimated return on ad spend of 3.3x to 6.8x. Synter reported a $9.89 average cost per click, ranging from $5.10 to $22.89 by region. Floyd Blaikie reported a $9.29 cost per click and a 0.7% click-through rate, then found that only five of its identifiable visitors matched its ideal customer profile.

This is one of three AI shopping fronts I have been tracking. Agents that buy are covered in my post on Shopify letting AI agents submit orders. Agents that recommend are covered in the study showing AI shopping picks big retailers 9 times in 10. Paid placement is the third, and it is the one you control with a budget.

What Product-Feed Ads in ChatGPT Mean for a High-Ticket Store

My read is that this release is about operations, not performance. Nothing in it proves that ChatGPT converts buyers of expensive products. What it does is push the cost of setting up a test close to zero, and that changes whether a small test is worth running. It is cheap to launch and expensive to run.

Here is hypothetical math, not reported data. Say your average order is $1,800 and your gross margin is 25%, which leaves $450 of gross profit per order. Your break-even conversion rate is your cost per click divided by $450.

At Common Thread Collective’s $4.41, break-even is 0.98%. At Synter’s $9.89 average, it is 2.2%. At the top of Synter’s range, $22.89, it is 5.1%. Net margin after shipping, chargebacks and your time is thinner than gross margin, so treat those thresholds as the floor.

Pull your own cold-traffic conversion rate from Google Ads before you spend a dollar here. If it sits below the threshold for the click cost you expect, the channel loses money however clean the feed is. My guide to turning Google Shopping clicks into sales covers what moves that number. So does my breakdown of why high-ticket ads need a complete conversion system.

Budget size is the second trap. At $25 a day for 30 days you spend $750. At $9.89 a click that buys about 76 clicks, and at a 1% conversion rate you should expect less than one order. You learn almost nothing from that.

I would commit to roughly 300 clicks before judging, which is about $2,970 at $9.89 or $1,323 at $4.41. Three hundred is a judgment call, not a statistical rule. If you cannot afford that test, wait and let other advertisers pay for the learning.

High-ticket buyers also call and request quotes before they order, so an order-only scorecard will undercount. Count phone calls and quote requests as results, and follow up fast. My breakdown of high-ticket sales follow-up shows how I handle that.

The feed is a spreadsheet, and the spreadsheet is where authorized-dealer stores win or lose. Titles need the brand and model number. Prices need to respect your supplier’s MAP policy. Availability needs to match what the supplier can ship. If your supplier data flows through a feed tool like Inventory Source, check that the sync is current before the first upload.

There is a real counterpoint to my caution. Common Thread Collective’s estimated 3.3x to 6.8x return suggests that disciplined advertisers can make this work. Blaikie’s finding cuts the other way: a decent click-through rate still delivered the wrong buyers. I could not find a high-ticket-specific benchmark from anyone, and OpenAI says it has none.

Google Shopping stays my first channel for high-ticket products, and I would not move budget out of it for this. Take a slice instead. Google is also changing its own results, as I covered in Google testing a sponsored grid that mimics free listings, so a second source of qualified traffic has value.

Paid placement also sidesteps one problem. If AI answers favor big retailers organically, an ad is the way a mid-sized authorized dealer gets in front of the same shopper. That is my read, not a finding from any of the reports above.

New traffic means new buyers, and new buyers mean new fraud and dispute exposure. Chargebacks on a $1,800 order cost far more than on a $40 one, so read my guide to chargeback prevention for high-ticket stores before the first order lands. Capture the email of every visitor you can with Klaviyo, because a buyer who does not purchase on the first visit is where high-ticket money gets made.

On the account itself, I would pick the public brand name before launch, since a brand name change pauses ads until review. I would also bill under a legal entity from day one so the legal name is never the thing you have to rework later.

If running another ad channel on top of supplier onboarding, feed cleanup and phone follow-up sounds like a second job, that is the kind of work my team takes off your plate in the turnkey done-for-you build.

Product-feed ads reward a clean catalog and tight margin math, and most owners do not have the hours to run one more channel. Want my team to scale the store you already have? See the scaling service →

Test ChatGPT Product Ads This Week Without Burning Budget

Run these five steps in order. The timing matters: Adobe forecasts $275.1 billion in US online holiday spending from November 1 to December 31, up 6.7% on last year, and you want your read on this channel before the season, not during it.

Confirm eligibility. Per OpenAI’s help center, you need a US-based Shopify store, admin access to Shopify and Ads Manager, and your site, industry, country, currency and timezone on hand. If you are outside the US, build the cleaned spreadsheet anyway so you are ready when the app reaches you.
Clean the feed before you upload it. Export your products, lead every title with brand and model number, match prices to your supplier’s MAP policy, and delete discontinued items. Claude is good at flagging titles that lack model numbers in a spreadsheet, but check its output against the supplier sheet. My guide to building a leaner high-ticket business with Claude shows how I use it.
Write the kill rule before you spend. Divide your expected cost per click by your gross profit per order to get break-even conversion, then decide the click count at which you stop. Track real profit in Finaloop. Ad revenue and cash in the bank are different numbers, as I explained in why your payout is not your profit.
Track calls and quotes, not only orders. Put a tracked phone number on landing pages and log every call. A free CRM such as HubSpot is enough. My walkthrough of setting up HubSpot Free for a high-ticket store shows the steps.
Add fraud screening to orders from the new channel. A tool like ClearSale reviews risky orders before they ship. My ClearSale setup guide covers the configuration.

If you would rather have a second set of eyes on your feed and margin math before you spend, book a discovery call with me this week.

Frequently Asked Questions

Who can run ChatGPT product-feed ads right now?
Per OpenAI’s help center, the Shopify app is currently available to all US-based Shopify merchants, with broader rollout planned as ads expand. Stores on other platforms can work through Ads Manager directly.

How much do ChatGPT ads cost?
Early advertisers reported costs per click from $3 to $22, according to Search Engine Journal, and OpenAI has published no official benchmarks. The $200,000 commitments reported in February applied to select advertisers in the early test, not to self-serve accounts.

Does bulk upload replace Google Shopping?
No. Search Engine Land describes ChatGPT as an incremental channel, and my own launch guide for high-ticket dropshipping still puts Google Shopping first.

Will ChatGPT recommend my store without ads?
Possibly, but a study I covered found AI shopping picks big retailers 9 times in 10. Treat organic recommendations as a bonus, and track whether your brand appears in AI answers with a tool like Semrush.

Do I need an LLC to open a ChatGPT Ads account?
The reporting I found does not describe an LLC requirement. Account settings do separate a legal billing name from a public brand name, which is why I would form the entity first and bill under it.

What if I do not have a store to test this on yet?
Start with niche selection. My free niches list at ecommerceparadise.com/niches has over 1,000 ideas. The high-ticket niches breakdown explains how I rank them.

Where do customer questions from a new channel go?
Into a helpdesk before they pile up in your inbox. Gorgias is a helpdesk built for Shopify stores and is worth a look before volume picks up.

Want 1-on-1 coaching to decide whether ChatGPT product ads belong in your high-ticket store’s budget? Get the coaching details →

I will keep tracking how OpenAI’s ad tools change for store owners and flag anything that touches your margin. Subscribe to the YouTube channel for daily breakdowns. More breaking news coming through the day.

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.