Agentic commerce matters more than ChatGPT ads

Hardly a week goes by without marketing feeds buzzing about ChatGPT ads. Product feed integrations, the Ads Manager beta, and comparisons with Google's dominance in search advertising dominate the conversation. For agencies, it is a compelling narrative, but the view is often too narrow. The real shift is not advertising on large language models. It happens in checkout flows, voice assistants, and agentic commerce infrastructure, where the transaction itself becomes the ad unit. Brands quietly winning this cycle are refining product data so they factor into AI-powered purchase decisions.

Why ChatGPT Ads are structurally weak

Referral traffic from ChatGPT to the rest of the web grew 206 percent in 2025, according to Semrush analysis of 17 months of U.S. clickstream data. That is the headline most people stop at. The footnote says growth is driven by deeper use from existing users, not audience expansion. ChatGPT's U.S. user base has been essentially flat since September 2025. People who use it do so more intensely, but the addressable audience is not growing.

This matters when building an advertising business. Ad revenue scales with reach, and reach requires a growing user base. Without new users, you run the classic sequence of building an audience and then monetizing at scale in reverse. Leaked OpenAI financials show another challenge: $13 billion in revenue against $34 billion in total costs and expenses in 2025, translating to an operating loss of nearly $21 billion.

In 2024, OpenAI spent $2.37 to generate every $1 in revenue. By 2025, that ratio improved to $1.60 per dollar. Progress, but neither sufficient nor fast enough. It is no surprise OpenAI postponed its IPO until next year. Amazon lost $30 million the year it went public. Google and Meta were already profitable before their IPOs. OpenAI is not playing in the same league. How long investors stay patient remains open.

OpenAI's master plan is agentic

What the ChatGPT ads narrative misses: advertising is a defensive move, not a strategic vision. Sam Altman has always been against it. OpenAI needed revenue to fund its capital-intensive master plan, at least partially. The picture becomes clearer at Google, Amazon, and OpenAI itself.

At Google I/O 2026, Google introduced Universal Cart, built on the Universal Commerce Protocol (UCP). This infrastructure lets AI agents complete purchases on users' behalf. It is not a shopping tab redesign but a transaction layer between intent and purchase, with Gemini deciding what gets recommended and bought. Merchants can already onboard UCP. This is happening today, not in some distant future.

At Amazon, Rufus, the shopping assistant used by more than 300 million customers in 2025, and Alexa+, the personalized AI assistant across hundreds of millions of devices, merged into Alexa for Shopping. Customers automate deal-finding and routine purchases based on personalized insights. Like Gemini, Alexa can complete the transaction and dramatically shorten the path from ad impression to purchase.

OpenAI also integrated product feeds in Ads Manager. At Google, Meta, and Amazon that feels basic; in a brand-new ad platform it is notable. Product feed ads in 2026 are not an advertising innovation but a platform reaching for familiar revenue streams while a more interesting architecture takes shape beneath the surface.

What this means for paid media today

The industry often asks the wrong question. Not whether you should test ChatGPT ads, obviously yes, or whether you should diversify beyond Google, also yes. The right question is: is your product data ready for agentic commerce?

When an AI agent, whether Alexa, Google's shopping agent, or whatever OpenAI builds next, makes a purchase recommendation, it will not pull from your campaign creatives. It will pull from your product feed. The cleanliness and completeness of that data determines whether you exist in the recommendation. We have seen this pattern before. When Google moved from keywords to audiences to intent signals, advertisers with cleaner conversion tracking and stronger first-party data won. When Meta moved to Advantage+ and black-box optimization, winners built better creative systems. Agentic commerce applies the same dynamic to the transaction layer: data is king.

• Maintain product feeds: Keep them complete, accurate, and updated in near real time.
• Implement structured data: Mark product attributes, availability, pricing, and other signals consistently across your catalog.
• Build API integrations: Connect to platforms developing agentic infrastructure.

Treat product data the way you should have been treating conversion tracking for the past decade: as a competitive advantage, not a maintenance task.

The real battle is not in the ad console

ChatGPT ads will generate revenue, attract advertisers, and appear in media plans. But it will not become the next Google Ads. The audience ceiling is real, the cost structure is brutal, and the competitive moat that made Google search ads irreplaceable for 20 years is not there.

The wave is not in the ad console. It is in infrastructure for task completion, automated purchasing, and agent-to-agent commerce. Google and Amazon are already building it. Brands that show up there will not do it through better bidding strategies. They will do it through better data. That is where SEO, GEO, and ecommerce teams should start.