Your Shopify store is already inside ChatGPT's commerce layer. Shopify enrolled millions of merchants automatically in March, and OpenAI opened product feed ads to retail advertisers in June.
Below is what we are telling our clients to do right now. You do not need platform access to start, and you can apply the same principles, whether your store runs on Shopify, BigCommerce, Magento, or WooCommerce.
What ChatGPT Ads actually are
ChatGPT ads are contextual placements that appear below the chat response visible to users on the Free tier and the $8/month Go plan. They are clearly labeled as "Sponsored" or “Ad” and separated from the answer itself.
An example from a ChatGPT Add. The initial query was ‘Date night ideas in NY’
The format OpenAI is pushing hardest right now is product feed ads. You connect the same structured catalog you already use for Google Shopping, and the platform builds sponsored placements automatically from your product names, images, and attributes.
Unlike Google Ads, there is no keyword selection and no manual ad creation per SKU.
The platform reads the conversational intent of the chat and surfaces the most relevant products from your feed. It's not only about keywords and targeting. Your website also needs to be ready for AI search.
What we know so far (June 2026)
-
The rollout is limited to select markets, including the United States, the UK, Australia, New Zealand, and Canada, with recent expansion in Mexico, Brazil, Japan, and South Korea.
-
Advertisers do not get access to users’ chats, chat history, memories, or personal details.
-
Advertisers receive aggregated performance data, such as impressions, clicks, and conversions.
-
Users can manage ad settings, including personalization controls, ad history, and ad feedback options.
-
Free-plan users may have access to an Ads-Free option, but with reduced usage limits and fewer features.
-
ChatGPT Ads is currently in beta for advertisers.
-
OpenAI’s Ads Manager supports campaign setup, billing, reporting, CPC and CPM buying options, conversion measurement, and product-feed campaigns for retail advertisers.
-
For now, this appears to be a phased rollout rather than a universal change across all ChatGPT users.

Image taken from the official OpenAI guide on Ads
How much do ChatGPT Ads cost?
Hard CPC and ROAS data from verified ChatGPT advertisers are not widely available. We know that the early pilot was invite-only, with an initial $200,000 minimum commitment reserved for a select few, which dropped to $50,000 in April, and OpenAI removed it entirely in May.
ChatGPT Ads runs on both CPM and CPC. OpenAI recommends starting bids of $3 to $5 per click, with a default max bid of $60 per thousand impressions, which is significantly higher than Meta but comparable to premium placements on Google.

There is a hard floor of $3 in the ads manager system
How do ChatGPT ads compare to Meta and Google ads?
Numbers aside, keep in mind this: ChatGPT ads cost more per impression than Meta ads and more per click than Google Shopping ads, but the buyer arriving from a ChatGPT conversation has already described their need in detail.
If you are already running Google Shopping and Meta dynamic ads, here is where ChatGPT fits relative to what you know. We built the table with both confirmed and estimated information in mind, to leave you space for your own experimentation.
|
Google Shopping |
Meta dynamic ads |
ChatGPT product feed ads |
|
|
Match signal |
Keyword + product attributes vs search query |
Behavioural: past browsing, purchase history, lookalikes |
Conversational intent: what the user typed, not past behaviour |
|
Ad placement |
Above organic results, product carousel |
Feed, Stories, Reels. Interruption format |
Below the AI response, labelled "Sponsored" |
|
Buyer stage |
High intent. User knows what they want |
Discovery. User may not know the product exists |
Decision. User is describing a need, not browsing |
|
Avg CPM |
$2–5 Display [confirmed] |
$10–14 US eCommerce [confirmed] |
~$25 current rate [estimated] (was $60 at launch) |
|
Avg CPC (eCommerce) |
~$1.16 Shopping [confirmed] |
~$0.78 all-industry [confirmed] |
$3–5 recommended starting bid [confirmed] |
|
Conversion rate |
~1.5–3.5% Google Shopping [confirmed] |
~1.4% catalog ads [confirmed] |
1.5× other channels, Criteo retail cohort [estimated] |
|
Targeting |
Keywords, product groups, audiences, negatives |
Demographics, interests, lookalikes, retargeting |
Context hints (natural language). No cookies, no retargeting yet |
|
Measurement |
Mature. Full attribution, 20+ years of tooling |
Mature. Pixel + CAPI, robust attribution |
Early. Pixel + CAPI launched May 5, 2026 [estimated] |
|
Feed setup |
Google Merchant Center, standard product feed |
Meta Catalog, same feed structure |
Same Google Shopping feed. 100-product sample review required |
|
Min spend |
None |
None |
None as of May 5, 2026 [confirmed] |
|
Access |
Open self-serve |
Open self-serve |
Self-serve at ads.openai.com. US only, expanding [confirmed] |
How to create ChatGPT Ads for eCommerce?
1- The first thing is to sign up for ads on ChatGPT and wait for approval.

Set up OpenAI’s Ads Manager Beta to get started
2- Open the ads manager and finish setting up the account

It’s still a Beta version, easy to tell from the Basic layout compared to Meta ads or Google’s ads. There is not much to work with, leaving little space at the moment to optimize the ads themselves.
3- Create your first ad campaign
The only element that differs from the typical Meta ads campaign, besides it being much simpler, is the ‘Context Hints’ section. No one yet has a best practice for how to write it, but experience getting products featured in ChatGPT and Perplexity AI tells us to write as we would for an LLM-friendly product description; leave nothing to chance and go deep on who we are serving and what we are serving.

Finally, it’s the ad creative, which is still restricted to an image and text

The takeaway is clear: You can’t do much on the Ads manager to optimize your ads, but you can optimize your store for AI search.
Start preparing your store right now
The ad manager gives you limited levers, but your store does not have that excuse.
On March 24, 2026, Shopify activated Agentic Storefronts for all eligible merchants. In a single week, millions of Shopify stores became technically reachable through ChatGPT's commerce layer, alongside Microsoft Copilot, AI Mode in Google Search, and Gemini, all managed from a single setting in Shopify Admin.
Products are automatically syndicated across every major AI channel, with no apps to install, no custom integrations, and no transaction fees beyond standard processing rates.
There are two separate systems running inside ChatGPT with which you can work
-
The first is conversational brand recommendations.
This is where ChatGPT suggests a brand by name in response to "what is the best [category]" type queries. This is driven by your citation footprint across the web: press coverage, blog mentions, and content that gets referenced.

How AI Overview recommend products based on search intent. You are betting on your content's relevance, not your keyword strategy.
-
The second is the shopping surface.
This is where your products appear as sponsored or organic results when someone asks for a recommendation. This is where Agentic Storefront gets you. Feed quality, product data completeness, and how well your descriptions match the way buyers talk about their needs determine whether you show up here
Why your current product feed might not work as-is
Just a few months before the release of Agentic Commerce, we found that the average AI Search Readiness score for the top Shopify stores was 42 out of 100, far from what the hype would lead you to believe.

The confusion between Google Shopping feed optimization and ChatGPT feed optimization is often what keeps merchants behind the benchmark.
A Google Shopping feed is built around keywords, structured attributes, and category taxonomy. The algorithm is matching product data to queries.
A richer feed improves your Google Shopping performance, your Meta dynamic ads, and every other channel that uses product data to match intent. But while the SEO foundations are the same for both, they solve different problems and require different approaches.

Example from Google Shopping feed
ChatGPT is matching product data to conversation.
Someone types "I need a waterproof hiking backpack that fits three days of gear and costs under $150," and the platform has to decide which products in its index are the right answer to that specific situation. It’s not about exact matches as much as it’s about the semantic nature of search and how it’s changed since AI was introduced.
If your product description is not detailed enough to give the AI context on how it will match the user's request, that product will lose ranking compared to a competitor whose description more closely matches what the buyer typed. The same applies to ChatGPT ads.
Depth and context win over keywords
For every product in your catalog, ask yourself:
If a buyer described their problem to ChatGPT, is there enough language in your feed to address it?
Most catalogs fail this test. Descriptions are too short, too feature-focused, and written for a keyword grid rather than a conversation.
You can effectively turn your product page description into LLM-friendly ones. It's simple for a store with a few SKUs, as shown in the example below; complexity rises when thousands of products are involved, and custom automation is needed.
Below is an example of how to turn a basic product description into an LLM-friendly one.


Matching the description with structured data is key. The screenshot from the Ahrefs Chrome extension shows that AI Crawlers will have an easy time finding the answers.
What you should do next
The work that determines your ChatGPT ad's performance is happening in your product data and store configuration right now, not in the ad interface. You do not need platform access to start.
1. Check your Agentic Storefronts status
Go to Shopify Admin > Settings > Sales Channels > Agentic Storefronts and confirm your catalog is actively syncing. Check the feed status tab for any sync errors or rejected products. A store that enrolled successfully in March may have new errors if you have added products or changed variants since then.

Start by adding your first products if you haven’t already.
2. Audit your product descriptions for conversational intent
Pull your top 50 products by revenue. For each one, write out the most natural way a buyer would describe their need in a chat. Then compare that to your current description. The gap you find is the edit list.
3. Enrich your feed with use-case and context attributes
Add problem-solution framing. Add lifestyle and use-case context. Add compatibility details where relevant. Your Google Shopping title and description fields are the starting point, not the finish line. The products that surface in ChatGPT's feed will be the ones with enough context to match the situation the buyer described, not just the category they searched.
4. Set up conversion tracking before you spend a dollar
When access opens, you want the ChatGPT pixel and Conversions API firing from day one. Tracking infrastructure takes time to validate. Your audience configuration needs to be in place before the first campaign goes live, or you'll be flying blind from the start. Do not let setup be the reason you spend four weeks on the platform before you have anything to optimize against.
How will your paid ads strategy change with ChatGPT Ads?
ChatGPT ads do not replace your existing paid channels. They add a surface where high-intent buyers are already looking.
Treat ChatGPT as the third pillar of your paid media strategy, alongside the Google and Meta spend. The brands that figure out how to run all three in a coordinated way will have a structural advantage going into 2027.
If you want help auditing your feed, getting your Agentic Storefronts configuration right, have a full breakdown on AI readiness, or working out how ChatGPT ads fit into your paid media mix, our team works with Shopify Plus brands on exactly this.
Get in touch, and get a free consultation first.