On August 7, 2026, OpenAI extended conversion optimization to product feed campaigns in ChatGPT Ads, with conversion-optimized cost-per-click bidding (oCPC) available in beta. This is not an isolated feature: it arrives alongside new measurement integrations, more detailed pixel diagnostics, dynamic URL parameters, a multi-product carousel format in testing, and a behavior change in automatic advanced matching that already switched on by default for any account that did not opt out before August 17.
That last detail is the one most worth checking. If you did not log into your account before August 17 to turn it off, your pixel configuration already changed, without you having to do anything. This article explains what conversion optimization means for product feed campaigns, how we got here, exactly what that advanced matching change means, and what you should review in your account this week.
What is conversion optimization (oCPC) for product feed campaigns?
Conversion-optimized cost-per-click campaigns, known as oCPC, let a product feed campaign in ChatGPT Ads keep charging per click while adjusting ad delivery to favor users more likely to complete a conversion, not just click. It is the difference between paying for attention and paying for attention aimed at the outcome that actually matters to you.
To make the transition easier, OpenAI lets you clone an existing CPC campaign directly into a new oCPC campaign, or create several oCPC campaigns in bulk, instead of rebuilding each one from scratch. That cuts migration work down to minutes instead of hours, especially useful if you already manage several product feed campaigns with similar structure across different clients.
It is worth understanding that this feature is still in beta, and its current scope is specifically defined for product feed campaigns, meaning ecommerce. If you manage ChatGPT Ads campaigns that are not product feed campaigns, this specific version of oCPC does not apply to your account yet, though the platform's broader direction toward conversion optimization does affect you indirectly.
How did we get here?
Conversion optimization did not appear overnight, it was a deliberate three step progression over this same summer. In May 2026, OpenAI notified beta advertisers that conversion-optimized campaigns would start rolling out, with early access beginning June 5 for any account that already had conversions configured before June 1. That was the real structural shift: moving the platform from reach and click based objectives toward an objective based on real business outcomes.
On July 24, 2026, the option to create oCPC campaigns by selecting the Conversions objective became more broadly available, along with geo exclusions and bulk campaign management tools. And on August 7, the most recent update specifically extended that oCPC capability to product feed campaigns, closing the gap that existed between the general conversion objective and the ecommerce use case.
This phased rollout pace, general infrastructure first, then extension to specific use cases, is consistent with how ChatGPT Ads has matured since opening to all US businesses in May 2026. Each update builds on the previous one instead of replacing it, so if you manage active accounts on the platform, it is worth assuming more features will keep arriving in similar waves through the rest of the year.
What happened on August 17 with Automatic Advanced Matching, and did it already affect you?
Automatic Advanced Matching, or AAM, is a feature that helps match more conversions to the correct ads using hashed contact information a user enters into your website's forms, even when traditional cookie based tracking fails to capture that conversion. As of the August 7 announcement, AAM became the default behavior for any new web pixel immediately.
For existing pixels, the change had a specific date: on August 17, 2026, AAM would automatically switch on for all existing pixels, unless the advertiser had turned it off beforehand from Tools, then Conversions, then Data Source, then editing the pixel. That date has already passed. If you did not go in and check that setting before August 17, your existing pixel is already using automatic advanced matching right now, without you having made any active decision about it.
For most advertisers, this is a welcome improvement and requires no corrective action: a better conversion match rate generally means more accurate measurement and better informed bid optimization. But it is exactly the kind of change worth confirming actively rather than discovering by accident, especially if your business handles sensitive contact information in its forms, or if you manage accounts for several clients with different tolerance levels around how their users' data gets processed.
What other measurement tools arrived alongside this?
Alongside AAM, OpenAI expanded its measurement ecosystem with three new integrations. Triple Whale now lets you measure ChatGPT Ads performance alongside the rest of your marketing channels in a single dashboard, useful if you already use that tool to consolidate reporting across platforms. Sonar Optimize can send stronger conversion signals back to OpenAI, which in theory helps the bidding model itself learn faster what is working. Hightouch now supports sending conversion events directly into ChatGPT Ads for measurement and optimization purposes.
Ads Manager also gained more detailed pixel validation diagnostics, specifically showing why a conversion event was rejected before it was sent, which field was involved, what type of error occurred, and a concrete recommendation for fixing it. For any team that has lost hours guessing why a pixel did not fire correctly, this is the most directly useful improvement in the whole package, even if it is the least flashy.
What are dynamic URL parameters and the multi-product carousel?
Dynamic URL parameters let ChatGPT Ads automatically append campaign, ad group, and ad identifiers directly to your landing page's query parameters, without you having to build those parameters manually for every ad. This makes attribution and downstream analysis easier, because every click that lands on your site carries with it the exact information about which campaign and which specific ad generated it.
Additionally, OpenAI began testing a multi-product carousel format for product feed campaigns, allowing several products to appear within a single ad unit instead of being limited to just one. Alongside this, the platform announced plans to launch in Brazil and Mexico during the week of August 10, 2026, expanding advertiser access across Latin America. If you manage accounts with business in those markets, it is worth confirming directly in your Ads Manager whether you already have the option to target campaigns to those countries.
What should you check in your account right now?
The first thing, regardless of whether you manage product feed campaigns or not, is to go into Tools, then Conversions, and confirm the current state of AAM on every pixel you have configured, either to consciously keep it active or to turn it off if you have a specific reason to prefer the previous behavior. Do not assume it is still set the way you last left it.
If you manage product feed campaigns with conversion history already configured, the logical next step is cloning one of your best performing CPC campaigns into an oCPC version, instead of starting from scratch, and comparing performance between both over a reasonable period before deciding to migrate fully. We already covered the broader picture of what ChatGPT Ads is and how it compares to other platforms in our guide for digital marketing agencies, and this update confirms the same trend we documented there: the platform is maturing toward the same kind of performance tools you already expect from Google or Meta.
To size this with an illustrative example, not a real data point but a typical scenario, imagine an ecommerce account running a product feed campaign with solid click volume but mediocre conversion. Cloning that campaign into oCPC, letting both run in parallel for two weeks with the same budget, lets you directly compare whether the shift toward conversion improves the real outcome, instead of assuming a new feature is always better just because it is new.
This same principle, measure before migrating fully, connects with our article ChatGPT Ads 2026: What Every Digital Marketing Agency Needs to Know: a new optimization feature is only worth adopting once you confirm it improves your actual outcome, not just because the platform presents it as an improvement. At JP Director we are testing oCPC on low risk ecommerce accounts before recommending it as a standard for every client with an active product feed, and proactively reviewing the AAM status on every account we manage, instead of assuming the default behavior is always the right one for every client.
Frequently Asked Questions
What happens if I did not check my account before August 17, 2026?
If you did not turn off Automatic Advanced Matching before August 17, 2026, the feature is already active by default on your existing web pixels. There is no penalty or error associated with this, it simply means your account now uses hashed contact information from your forms to improve conversion matching. You can turn it off at any time from Tools, Conversions, Data Source, and editing the relevant pixel.
Does oCPC conversion optimization work for any type of campaign on ChatGPT Ads?
Not yet. The August 7, 2026 update specifically extended oCPC availability to product feed campaigns, meaning ecommerce. If you manage another type of campaign on ChatGPT Ads, this specific version of the feature does not apply to your account for now, though the platform's broader conversion optimization infrastructure keeps expanding progressively.
Do I need to rebuild my campaigns from scratch to use oCPC?
No. OpenAI lets you clone an existing CPC campaign directly into a new oCPC campaign, keeping its structure, or create several oCPC campaigns in bulk if you manage multiple at once. During the cloning process, Ads Manager automatically fills in the conversion event when only one is available, or prompts you to choose one if several are configured.
What is Automatic Advanced Matching and why should I care?
Automatic Advanced Matching is a feature that uses hashed contact information from your website's forms, like email or phone number, to more accurately match conversions with the ads that generated them, even when traditional cookie based tracking fails to capture them. It matters because it improves the accuracy of your measurement and, therefore, the quality of the optimization decisions you make on your campaigns, but it also matters because it involves processing your users' contact information, which is worth confirming consciously on every account you manage.
Last updated: August 2026. ChatGPT Ads' conversion and measurement features remain in active development; verify the current status directly in Ads Manager before making decisions about pixel configuration or campaign migration.
