OpenAI Told Investors ChatGPT Ads Would Be a $100 Billion Business. Analysts Cap the Whole Market Near $5 Billion.

eMarketer published its first full US AI advertising forecast on June 4, and the number that should worry OpenAI is the one it left out. The firm put the entire chatbot ad market, ChatGPT and Microsoft Copilot and Google’s AI Mode and Amazon’s Alexa shopping assistant combined, at less than $1 billion this year and $5.41 billion by 2030. OpenAI has told investors that ChatGPT ads alone will be a $100 billion business by then.

That gap is the story. On its current trajectory, OpenAI’s ad business is on pace to miss its own five-year forecast by 90%, Adweek reported this month, citing eMarketer’s data. Even if ChatGPT captured every dollar spent inside every chatbot on the market, it would still land well short of the target its own investor deck promised.

Nate Elliott, Principal Analyst for AI in Marketing and Commerce at eMarketer, draws the distinction marketers should copy into their own planning:

“Search monetized because users were already in commercial retrieval mode. Ads appeared next to expressed demand. They helped.”

A search box is a person telling you what they want to buy. A chatbot is a person asking for help thinking.

Selling ads against the second is selling attention where the intent to buy hasn’t formed yet.

OpenAI’s pitch runs the other way. David Dugan, the former Meta executive it hired as Head of Global Ads Solutions, told the industry at Cannes that ChatGPT reads user context better than a keyword does, and that the early reception backs the bet. “The feedback that I’ve gotten from the agencies is they’re excited to have a new platform to experiment with,” he said. The adoption is real: through its partner Criteo, more than 2,000 brands were advertising on ChatGPT as of June, after a February test that has since added cost-per-click buying and a self-serve ads manager.

Elliott isn’t buying the line from a fast start to a $100 billion business. Asked why OpenAI’s projection sits so far above his own, he didn’t hedge:

“my guess is lazy forecasting, and obvious ulterior motives.”

The mechanics behind his number are the part worth sitting with. eMarketer expects US AI ad spending to more than double, from about $32 billion in 2026 to $68 billion in 2030. But more than 80% of that money lands beside AI-generated content, not inside a chatbot conversation. Picture the ads stacked around a Google AI Overview, the AI summary that now sits atop search results. The dollars are real and growing. They’re just pooling where a person is still in shopping mode, not where they’re deep in a chat. OpenAI’s forecast assumes it can pull commercial intent into the conversation itself, and nobody has shown that it moves.

That’s why the operators closest to the buy are hedging. Sjeel, a fractional CMO who writes about AI marketing, walked through the reasons a B2B team shouldn’t rush in. The senior decision-makers who matter sit on paid ChatGPT tiers that don’t show ads. A paid placement mostly amplifies a company the model already names on its own. Her verdict was plain:

“The fact that you can run ChatGPT ads now doesn’t mean you should.”

She’d rather teams earn the mention before they pay for it:

“Building for them is the move. Buying them isn’t… yet.”

“Building for them” means answer-engine optimization, the work of getting a brand cited inside AI answers the way SEO once earned a blue link. It’s slower than buying a placement, and it’s the thing that actually compounds.

For a marketing budget, the pattern rhymes with the rest of the year. Global ad spend cleared $1 trillion in 2026 while marketing budgets stayed flat, so every new channel gets funded by cannibalizing an old one. Moving money into ChatGPT ads to chase a $100 billion forecast means pulling it out of channels that convert today. And the same shift that’s turning search into AI answers is the one making those answers hard to buy your way into.

The forecast will keep making headlines, because a $100 billion number is built to. The market underneath it is a fraction of that, and the ads are landing next to the answer, not in it. Treat ChatGPT ads as an experiment with a learning budget you can afford to lose, not a line item you move real money into on the strength of someone else’s projection. Earn the citation first. The ad, if it ever pays, comes after.