OpenAI just turned ChatGPT into an ad platform. Here's what brands need to do in the next 90 days.

On May 5, 2026, OpenAI quietly did something that will reshape the next decade of digital advertising. They dropped the $50,000 minimum spend requirement, opened ChatGPT Ads to every U.S. advertiser through self-serve, and added cost-per-click bidding. The pilot that started with a handful of enterprise brands is now open to everyone.

The numbers already in market are not small. The ChatGPT ad business is pulling $100M+ in annualized revenue with an internal target of $102 billion by 2030. That is not a test. That is a land grab — and it is happening faster than any platform-to-ad-network transition in digital history, including Google's.

Why this matters more than the headlines suggest

Every major ad story gets covered the same way: platform launches ads, brands rejoice or panic, the cycle continues. But this one is structurally different, and most of the coverage is missing the point.

The core difference is where in the purchase journey ChatGPT lives. Google Ads intercept a user who is already searching — intent already formed, query already typed. ChatGPT sits earlier and later than that. It's where people go to think through decisions, compare options, understand categories, and increasingly, to get a direct recommendation instead of a list of links.

When a user asks ChatGPT “what's the best project management tool for a 50-person team?” — that is not a search. That is a buying conversation. The answer that comes back carries the weight of a trusted advisor, not a search result. Paid placement in that context is categorically different from a banner ad or even a sponsored link.

The 90-day playbook

Here is what I would actually do right now if I were running marketing or media strategy for an enterprise brand:

First: Audit your organic presence before touching paid.Run your category queries in ChatGPT, Claude, Perplexity, and Gemini. See what comes back. If your brand isn't appearing organically in the answer layer, paying for placement will help you show up — but without the surrounding context, the model has nothing to reinforce. You'll get visibility without credibility, which converts poorly.

Second: Treat GEO like the prerequisite it is. Generative Engine Optimization — structured data, authoritative citations, schema markup, content written for how LLMs retrieve and synthesize — is now table stakes before you run LLM advertising. The brands that will win paid placement in ChatGPT are the ones the model already trusts organically. This is not different from how Google Ads works at the quality score level. Organic authority amplifies paid.

Third: Start small, measure differently. Standard attribution models do not work here. ChatGPT ad interactions are conversational — a user might click a sponsored result mid-conversation, bounce, and return three days later through direct. Last-click attribution will undercount impact significantly. Build for incrementality testing from day one or you will pull budget from something that is actually working.

Fourth: Think about format. Early reports from the enterprise pilot suggest that sponsor callouts appearing within conversational responses outperform traditional display-style placements. This should surprise no one — the format that fits the container works better. Write ad copy that reads like a recommendation, not an ad.

The harder question

The thing nobody wants to say out loud: what happens when the answer layer becomes pay-to-play at scale?

Right now, the brands appearing in ChatGPT responses are there because of content quality, schema, citations, and genuine authority. The moment heavy paid penetration hits, smaller brands and independent voices that built organic presence get crowded out — not because their content got worse, but because their budgets can't compete. We have seen this movie before, on every platform that started as a meritocracy and became a media buy.

The window to build durable organic presence in the answer layer is closing. Not tomorrow, but within 18-24 months. The brands treating GEO as a “we'll get to it” initiative will be in the same position in 2028 that brands who ignored SEO in 2004 found themselves in 2008.

Build the organic foundation now. Run the paid experiments. Measure differently than you ever have before. The next 90 days are genuinely important.