ChatGPT Ads Adds Conversion Bidding, Product Carousels and Five New Markets

OpenAI is adding the infrastructure advertisers expect from a performance-marketing platform to ChatGPT Ads. The latest changes cover conversion optimization, product catalogs, campaign operations, attribution and pixel matching, while the advertising program has expanded beyond its original markets.

The important caveat is reach. ChatGPT Ads is limited to logged-in adults on the Free and Go plans, and OpenAI has not published a separate weekly user count for that eligible audience. Broad ChatGPT usage figures—or third-party estimates—should not be treated as guaranteed ad reach.

The biggest change: conversion-optimized CPC bidding

Product-feed campaigns can now use conversion-optimized cost-per-click, or oCPC, in beta. Advertisers continue to pay per click, but delivery is optimized toward clicks that are more likely to produce a selected conversion event. The initial release supports click-billed campaigns using fixed or manual bidding with a positive conversion bid cap.

This gives product advertisers a path beyond optimizing solely for traffic. It is still an early beta, so performance will depend on conversion-event quality, feed coverage and the amount of usable campaign data available to the system.

Campaign cloning and bulk creation

Advertisers do not have to rebuild every campaign from scratch. An existing CPC campaign can be cloned into an oCPC campaign, and oCPC campaigns can also be created in bulk. During cloning, Ads Manager can prefill the conversion event when only one eligible event is available; otherwise, the advertiser selects one.

For teams already testing product-feed campaigns, the workflow reduces the operational cost of comparing click-based and conversion-oriented delivery.

Product ads become more commerce-friendly

OpenAI is also testing a multi-product carousel format for product-feed ads. Rather than presenting a single product listing, the format can show several products from one advertiser in the same creative experience. Reporting on the test indicates that the carousel is confined to an individual advertiser’s product portfolio.

The format is a natural fit for shopping journeys in ChatGPT, where users may be comparing several products rather than looking for one exact item. Because the carousel remains a test, advertisers should treat availability and final behavior as subject to change.

Better attribution through URL parameters and measurement connections

ChatGPT Ads now supports dynamic URL macros that can pass identifiers into landing-page query parameters at delivery time. Reported parameters include {campaign_id}, {ad_group_id}, {ad_id} and {ad_account_id}. These values can help advertisers connect a landing-page visit with the campaign and ad that generated it without manually maintaining every identifier in the URL.

The platform’s measurement ecosystem is expanding as well:

• Triple Whale can connect ChatGPT Ads performance with broader ecommerce attribution and reporting workflows.
• Sonar Optimize can provide stronger conversion signals back to OpenAI, according to reporting on the integration.
• Hightouch’s Conversion API support gives brands another route for sending conversion data into advertising workflows.

These integrations do not eliminate attribution uncertainty. They make it easier to compare ChatGPT Ads with other channels, but the quality of the result still depends on consistent event definitions, clean landing-page tracking and a reliable measurement model.

Automatic Advanced Matching changes pixel defaults

Automatic Advanced Matching became the default for new web pixels and was scheduled to roll out to existing pixels on August 17, 2026, unless advertisers opted out. The feature uses hashed customer information collected through a pixel to improve conversion matching, including across sessions or devices.

Advertisers should therefore review pixel configuration, event quality and applicable consent requirements rather than treating the default change as a plug-and-play performance improvement. Better matching can improve attribution, but it also increases the importance of knowing what data the pixel collects and how that data is handled.

Where ChatGPT Ads is available

OpenAI’s advertising program began in the United States and was initially extended to Canada, Australia and New Zealand. On August 11, 2026, OpenAI said ChatGPT Ads had launched in five additional markets: the United Kingdom, Mexico, Brazil, Japan and South Korea. That brings the named rollout to nine markets.

Ads are intended for logged-in adult users on the Free and Go plans. Plus, Pro, Business, Enterprise and Education users do not receive ads, and accounts identified as belonging to people under 18 are excluded.

OpenAI has not broken out how many weekly users meet all of those conditions in the active markets. Some industry trackers cite more than 1 billion weekly ChatGPT users, but that is a platform-wide figure or estimate—not a disclosed count of ad-eligible Free and Go users.

Similarweb adds visibility into AI ad placements

Similarweb’s AI Ads dataset extends ad intelligence into ChatGPT, Google AI Mode and Google AI Overviews. It is designed to show which brands are buying placements, the creatives and sponsored content they use, the surrounding prompts or user intent, and their share of voice.

That matters because conversational AI surfaces have not offered the same public ad-library experience marketers are used to from major social and search platforms. Similarweb’s product is aimed at competitive analysis: identifying where competitors appear, what messages they use and which contexts they target.

Other monitoring products are pursuing a similar need. Adthena, for example, describes tracking brand presence, prompts where ads appear and competing advertisers across ChatGPT placements.

Omneky brings ChatGPT Ads into cross-channel workflows

Omneky has added self-service ChatGPT advertising to its campaign platform. Its customers can create, launch and optimize ChatGPT campaigns alongside channels including Meta, Google, LinkedIn, TikTok and Reddit.

The significance is operational rather than a new ChatGPT ad format: marketers can manage creative production, campaign execution, budget optimization, testing and reporting across multiple channels through a more unified workflow. Omneky also positions its platform as a way to connect advertising data and actions to AI assistants through its MCP offering.

What the update means for advertisers

Taken together, the changes point to a clear direction. ChatGPT Ads is evolving from a limited conversational-placement test toward a more conventional performance stack:

1. Optimize for outcomes: oCPC introduces conversion-oriented delivery for product-feed campaigns while retaining click-based billing.
2. Scale campaign operations: cloning and bulk creation make testing easier for advertisers with multiple feeds or markets.
3. Show more products: carousels support broader product consideration within one ad experience.
4. Improve measurement: dynamic identifiers and third-party integrations connect ChatGPT activity with downstream attribution systems.
5. Expand inventory: the named rollout now covers nine markets, although eligible user reach remains unknown.
6. Watch data governance: Automatic Advanced Matching makes pixel and consent reviews more important.

The strongest conclusion is not that ChatGPT Ads already matches the maturity or scale of established advertising networks. It is that OpenAI is building the components needed to compete for performance budgets: conversion signals, commerce formats, international inventory and an ecosystem of measurement and intelligence partners.