OpenAI Expands ChatGPT Ads with Richer Creatives, E-commerce Features, and CPA Models

I’ve been tracking OpenAI’s advertising rollout closely, and the latest expansion makes one thing clear: ChatGPT ads have moved well beyond a cautious test. They’re starting to look like a legitimate performance marketing channel.

What began in early 2026 as a limited trial of clearly labeled sponsored cards beneath ChatGPT responses has evolved quickly. OpenAI is now adding richer creative formats, stronger e-commerce functionality, and, most notably for performance marketers, CPA-based bidding. That shifts the conversation from visibility to measurable business results.

From ad test to performance engine

In January, OpenAI introduced a careful advertising framework: ads would appear only on free and Go tiers, stay separate from answers, and avoid sensitive categories like health and politics. That positioning mattered. OpenAI understood that trust would determine whether advertising inside a conversational interface could work.

By February, limited testing had begun for logged-in adult users in the US. The first format was simple: sponsored cards with product names, prices, and a small placement below responses. Functional, but clearly early-stage.

Since then, the pace has picked up. In May, OpenAI opened its self-serve Ads Manager more broadly in the US, lowered entry barriers for advertisers, introduced CPC bidding, added a conversion tracking pixel, and expanded product feed tools for e-commerce brands. Then came larger images, dynamic CTAs, dedicated shopping layouts, and carousel-style creative formats. Now CPA bidding is arriving for select advertisers.

That progression suggests OpenAI is doing more than monetizing attention. It’s building an ad system meant to compete for performance budgets.

Why richer creatives matter in ChatGPT

Creative behaves differently in a conversational environment than it does in social feeds or traditional search. In ChatGPT, ads show up alongside active intent. Users are already asking, comparing, researching, or shopping. That makes the ad less of an interruption and more of a relevant extension of the session.

The latest creative updates reflect that. Advertisers can now test larger visuals, more flexible calls to action such as “Shop now,” “Book now,” “Sign up,” or “Learn more,” and e-commerce-specific formats that include pricing, reviews, and multiple image orientations.

OpenAI appears to be adapting ads to the flow of conversation instead of forcing legacy formats into a new interface. That’s the right approach. In AI-driven environments, context matters more than almost anything else. A static creative that performs well in one prompt category may fall flat in another. OpenAI executives have already indicated that ad performance varies by conversational context, which makes perfect sense.

For brands, that means creative strategy needs to mature quickly. Repurposing existing paid social assets and hoping they work in chat won’t be enough. Marketers will need assets built for question-led discovery and intent-rich journeys.

The e-commerce layer is where things get interesting

The addition of product feeds and dedicated shopping units may be one of the most important developments so far. ChatGPT is becoming more transactional.

Earlier this year, OpenAI adjusted its commerce flow, moving away from a more native checkout concept and toward redirecting users to merchants’ own storefronts through an in-app browser. It sounds like a small change, but it matters. It reduces friction while leaving merchants in control of the final conversion experience.

Pair that with structured product cards, prices, reviews, and visual merchandising inside ChatGPT, and the platform starts to look less like a pure assistant and more like an intent-driven commerce gateway.

For e-commerce brands, that opens up a meaningful opportunity. If product catalogs can feed directly into ad creation and optimization, ChatGPT becomes a place where discovery, consideration, and click-out purchase intent can happen in the same session. That compresses the funnel in a way marketers should pay attention to.

It also reinforces a broader trend I’ve been talking about for a while: machine-readable commerce is becoming essential. Brands with better product data, cleaner metadata, stronger reviews, and more structured content will have a clear advantage in AI-mediated environments.

CPA bidding changes the advertiser conversation

The headline feature for many performance marketers is CPA bidding.

Once a platform moves beyond impression-based or click-based pricing toward cost-per-action, it starts speaking the language growth teams and direct-response buyers actually care about. Marketers don’t optimize for impressions. They optimize for purchases, sign-ups, leads, and profitable customer acquisition.

That’s why OpenAI’s move into CPA matters. It brings ChatGPT closer to established performance ecosystems like Meta and Google while preserving the distinct value of conversational intent. With the supporting infrastructure now taking shape, including pixels, conversion setup, and optimization pathways, OpenAI can start proving not only that users engage with ads, but that those ads drive outcomes.

From my perspective, this is the point where ChatGPT advertising becomes much harder to ignore. Brands that were merely curious now have a reason to test with a performance framework instead of treating the channel as a pure awareness play.

What this means for marketers right now

Marketers should see this as the beginning of an AI-native ad category.

Not exactly search. Not exactly social. Not exactly retail media. Something adjacent to all three, but driven by live conversational intent.

That makes a few priorities immediate:

• Creative needs to be context-aware. Ads in chat should match the question, the stage of intent, and the most likely next action.
• Measurement setup is no longer optional. If CPA and conversion optimization are in play, clean tracking becomes foundational.
• Product data quality matters more than ever. Structured feeds, accurate pricing, reviews, and availability signals will increasingly shape visibility and performance.
• Early testing creates an advantage. Advertisers that enter now will learn faster while competition is still taking shape.

There’s also a bigger strategic shift underway. OpenAI’s ad expansion is happening alongside major changes in AI search, agentic commerce, and platform monetization. As AI assistants play a larger role in discovery and decision-making, brands will need to think beyond traditional channel silos. The key question won’t just be “How do I rank?” or “How do I target?” It will be “How do I become the best machine-readable recommendation at the exact moment intent appears?”

The bigger picture

OpenAI’s advertising expansion is about more than revenue. It’s about defining how commercial experiences work inside AI interfaces without undermining user trust. That balance is critical.

If ads stay relevant, clearly labeled, and performance-oriented, ChatGPT could become one of the most important new channels in digital marketing. If they become intrusive or poorly matched to the conversation, the experience will deteriorate quickly. The platform’s future will depend heavily on execution.

Still, the direction is clear. Richer creatives improve relevance. E-commerce features reduce friction. CPA models add accountability. Together, they push ChatGPT from experimental monetization toward a serious ad ecosystem.

In my view, marketers who wait for the channel to fully mature may miss the most valuable window: when costs are still taking shape, competition is still relatively light, and early learning creates outsized gains. If you want to navigate that shift with a platform built around measurable ad efficiency and smarter revenue optimization, it’s worth taking a close look at ROAS Suite.

FAQ

What’s new in OpenAI’s ChatGPT ad expansion?
OpenAI has added richer ad creatives, shopping-focused formats, stronger e-commerce tools, conversion tracking, and CPA bidding for select advertisers.

Why does CPA bidding matter?
CPA bidding ties ad spend to actual outcomes like purchases or sign-ups, making ChatGPT more attractive to performance marketers focused on acquisition efficiency.

How are ChatGPT ads different from search or social ads?
ChatGPT ads appear within active conversations, where users are already expressing intent through questions, comparisons, and research. That creates a different context for targeting and creative.

Why is product data so important here?
Structured product feeds, accurate pricing, reviews, and availability help AI systems surface relevant offers more effectively and improve ad performance in conversational commerce environments.

Conclusion

ChatGPT advertising is no longer just an experiment. OpenAI is building the pieces of a real performance channel, with better creatives, stronger commerce capabilities, and bidding models tied to outcomes. For marketers, the opportunity now is to test early, build context-aware creative, clean up measurement, and treat machine-readable commerce as a competitive advantage rather than a technical detail.