ChatGPT Ads Manager is Open to Everyone: How Should You React?

There is a type of opportunity that people only look back on with regret months later. It usually doesn't make a lot of noise when it first appears. Google Ads in 2002, Facebook Ads in 2007—at the time, they didn't look like “billion-dollar channels” at all. They were just a crack in the door: cheap, with very few people crowding in. And precisely when there were the fewest people crowding in, the costs were the lowest.

A similar story is unfolding with ChatGPT Ads Manager. From a playground that was almost exclusively reserved for major brands—where entry required a minimum commitment of $50,000 and usually meant going through major agencies—OpenAI has rolled out a self-serve mode for every business in the US. No big contracts required, no middlemen needed. You sign up yourself, set your own budget, and run your campaigns. It sounds simple, but behind it lies a much deeper shift in how people make purchasing decisions.

What the $100 Million Figure in Six Weeks Tells Us

ChatGPT crossed the milestone of $100 million in annualized advertising revenue within its first six weeks—and more notably: that figure came from less than 20% of the user base eligible to see ads every day, while roughly 85% of users on the Free and Go plans are already eligible. In other words, the platform is currently operating at a tiny fraction of its actual capacity.

What caught my attention wasn't the revenue figure, but the supporting infrastructure. Alongside the self-serve Ads Manager, the platform added CPC and CPM bidding, conversion tracking, pixel measurement, and attribution capabilities. This is the pivotal turning point: it transforms ChatGPT Ads from an experimental brand-awareness product into a measurable performance channel. An advertising ecosystem only truly expands when people spend money and can actually see the results.

Here are a few key data points to visualize the scale:

• 800 million active weekly users, processing roughly 2.5 billion prompts every day.
• Ads currently only display to logged-in adult users on the Free and Go plans. Users on Plus, Pro, Business, Enterprise, and Education tiers do not see ads.
• Geographic expansion has kicked off in Canada, Australia, New Zealand, the UK, Japan, South Korea, Brazil, and Mexico.
• OpenAI's 2026 ad revenue target is $2.5 billion, with much larger long-term expectations.

Why Advertising in ChatGPT is Fundamentally Different

The easiest mistake to make is to lift creative assets straight from Google or Meta and expect them to perform well. The environment here is fundamentally different. ChatGPT is a place where multi-turn conversations take place: people ask, follow up, compare, and only then arrive at a decision. Before they ever see an ad, they have already spent time narrowing down their problem. The AI has already helped them with the research and comparison phase.

This means people coming from a ChatGPT ad are usually much further down the funnel than people coming from display ads or social media. They don't need another blog post titled “5 Things to Know Before You Buy.” They need a specific answer or a clear solution. If your landing page is still a generic homepage, your conversion rate will suffer—not because the channel is weak, but because your messaging is out of sync.

When someone clicks a Google ad, they are usually at the beginning or middle of their research journey. When someone encounters an ad within ChatGPT, they have already gone through a specific conversation and successfully narrowed down their problem.

This is why I advise against measuring this channel with the same set of KPIs used for search advertising. If you evaluate it by ROAS in the very first month, you will draw the wrong conclusions. The initial phase should be viewed as learning how advertising works within a conversation.

The Two Active Formats and How They Work

Both current formats appear beneath the AI's response, clearly labeled as sponsored, and visually distinct from organic content. This point is crucial: OpenAI has made it clear that ads do not influence the response content, and advertisers only receive aggregated performance data—users' private conversations remain confidential.

• Shopping product carousels, integrated with checkout capabilities. Ideal for e-commerce brands selling visually intuitive products with a straightforward purchasing path.
• Conversational banners featuring a call to action and a “Ask ChatGPT about this ad” button. When users click, they enter a conversation fueled by data uploaded directly by the brand: product details, FAQs, and service specifics.

The second format is what truly deserves a pause for thought. Users ask about prices, sizes, and features—and receive direct answers using your data, right within the platform. For products requiring careful consideration or B2B offerings with long sales cycles, this is practically an automated sales consulting layer. However, it also introduces a new requirement: the data you upload must be accurate, complete, and up to date. If your documentation is flawed, the AI will give incorrect answers on your behalf, very smoothly and with great confidence.

Contextual Targeting, Not Keywords

ChatGPT doesn't target by familiar keywords or demographics. It uses contextual matching based on the topic of the current conversation, previous chat history, and the user's past ad interactions. Theoretically, this is a much deeper intent signal than interest-based targeting.

However, “deeper” also means less predictable. You no longer have sliders to adjust age, gender, or location. You have to think in the language of the buyer. The most practical approach I find reasonable is to use ChatGPT itself to study how it talks about your industry. The way the AI expresses itself is a preview of the context your ad will appear within. If your messaging doesn't match that context, you are talking to a completely different room.

Which Industries Should Jump In Early, and Which Should Wait?

The groups with the clearest opportunities are those where users already turn to ChatGPT for research and decision-making. Based on insights from initial testing data:

• B2B software and professional services—complex questions, long sales cycles, multiple conversational sessions.
• Financial products, health, and wellness—users ask highly detailed questions before making decisions.
• Travel and hospitality—comparing, weighing options, and asking questions iteratively.
• Consideration-heavy e-commerce—where people compare specifications or ask AI to evaluate options.

Conversely, low-priced commodity goods driven by impulse buying will see a lower signal-to-noise ratio—at least initially, before ad formats expand further. This doesn't mean you shouldn't try, but rather that you shouldn't set the wrong expectations.

If You Are Starting This Week, Here Is What You Should Do

1. List the real questions potential customers are asking ChatGPT about your product. This is the new version of keyword research.
2. Align your language: rewrite your messaging the way people actually express themselves, not the way your marketing department prefers.
3. Set aside a learning budget for the first few months, treating it as team training expenses rather than an instant ROAS machine.
4. Redesign your landing pages for context: directly acknowledge the problem the user just discussed, provide a specific answer, and make the next step crystal clear.
5. Prepare data for the conversational format: product info, FAQs, policies—write them as if you are training a brand-new customer service representative.
6. Establish measurement standards from day one, because when competitors flood in, your foundational data will become a hard-to-copy advantage.

A key point to note: currently, less than 7% of ads are rated by users as low-relevance, and OpenAI states they are actively improving this metric alongside user trust. For advertisers, this is a positive signal—a high-quality baseline gives decent ads a fair chance of being seen.

The Worries, and Why I Still Recommend Trying It

To be fair, this is still an immature platform. Measurement tools will evolve, policies will shift, and relying solely on Free and Go tier users quite clearly limits your target audience. Anyone who goes all-in with their entire budget right away is gambling on an unstable foundation.

Yet an undeniable truth remains: the cost of waiting is usually higher than the cost of experimenting. When minimum budget barriers are removed and eligible user groups expand, bid prices will climb rapidly. Brands that build expertise in conversational advertising during this low-competition phase will enter the later game with a structural advantage—much like early adopters of Google Ads or Facebook Ads did.

I don't think the right question is “should we run ChatGPT Ads or not.” The better question is: what can you learn from a small budget over a couple of months, when almost no one around you is doing it yet?

And this is where I want to hear your thoughts. Some say sneaking ads into an AI conversation is natural because users need answers and brands need to be found. Others argue it's a boundary that should be preserved—that spaces for reflection and inquiry should remain ad-free. Which side are you on? Will you test this channel, or wait until it stabilizes before stepping in?

Leave a comment below—even just a single line—I'd love to know what you're thinking. If you're considering trying it out, start with a budget small enough so you won't regret it, but large enough to learn something. Then come back and tell me the results.