What ChatGPT Ads Reveal about the Future of AI Monetization

Having launched in February this year, ChatGPT's ad rollout is still in its early stages, but the signals are already attracting attention. According to Reuters, OpenAI's U.S. advertising business exceeded $100 million in annualized revenue within weeks of launch, while remaining deliberately constrained. Exposure is limited, formats are simple, and measurement tooling is still evolving, yet the early results are already meaningful.

But the ads themselves are not the most interesting part of the story. What matters is what they reveal about where AI monetization is heading, and what it will take to get there.

This isn't a new placement. It's a new signal.

It is easy to look at ChatGPT ads and see a familiar format: a sponsored unit appearing alongside a response. But that interpretation misses the bigger shift.

The real change is not where the ad appears, it's the signal powering it.

For decades, digital advertising has relied on increasingly sophisticated ways of approximating intent. Display advertising used audiences and context. Social platforms used behavioral signals. Search represented a major breakthrough by allowing users to explicitly express intent through a query.

Conversational AI takes that one step further. Users just don't state a need, they explain it. A conversation about running shoes might include training goals, a budget, a preference for certain brands, a previous injury, and a question about timing. These are signals that would less likely all surface in a traditional search query.

And unlike a query, the conversation doesn't end with a single expression of intent. It evolves with each engagement adding context. Each response reveals more about what the user is actually trying to accomplish.

Intent is no longer captured once. It is continuously revealed, creating a fundamentally different signal from anything the industry has previously had access to.

Intent driving early performance

One of the most telling aspects of ChatGPT's rollout is that strong performance is emerging despite infrastructure that is, by the standards of established advertising channels, still quite limited.

Formats are simple, tooling is still developing and measurement is immature, and optimization capabilities are constrained. And yet advertisers keep showing up. The reason is straightforward: the underlying signal is unusually strong.

Advertisers have spent years investing in systems designed to infer what users might want. Conversational AI reduces the amount of guesswork dramatically.

In the running shoes conversation above, the signals of training goals, budget, preferred brands, previous injuries, and purchase timing would rarely be captured in a traditional search query.

Strong signal quality changes the economics of the opportunity.

When intent is clearer, relevance improves. When relevance improves, performance follows. The early success of conversational advertising is not primarily a format story. It is a signal story.

The infrastructure moment

Most discussions about ChatGPT ads focus on the ads themselves. The more consequential story is what needs to be built behind them.

Every major platform shift in digital advertising has required new infrastructure. Search was not simply a new ad format. It introduced keyword auctions, quality scoring, and entirely new approaches to targeting and measurement. Mobile was not simply a smaller screen. It required new networks, mediation platforms, attribution systems, and optimization tools. Both were new ecosystems built from scratch.

Conversational AI is following the same pattern, but moving faster.

The challenge is no longer simply showing an ad inside of a chat interface. It is understanding intent in real-time, creating competition for demand, and delivering relevant experiences that fit the conversational format, and measuring performance across decision journeys that don't fit neatly into impression and click models.

That requires systems purpose-built for this environment, not extensions of what already exists.

What comes next

The first phase of conversational advertising has proven something important: users will engage with sponsored experiences when they are relevant to the conversation.

The second phase is harder. It is about building the infrastructure to make that relevance consistent, measurable and scalable across the broader ecosystem of AI-native products, not just the largest platforms.

Right now, the value being generated in conversational environments is heavily concentrated. ChatGPT has the scale and resources to build its own ad infrastructure. Most AI-native products do not. They are sitting on the same quality of intent signals of users expressing goals, preferences, and decisions in natural language, but they have no mechanism to connect those signals with advertiser demand.

Closing that gap requires the same kind of infrastructure build that defined earlier platform shifts. Keyword auctions didn't exist before search needed them. Mobile mediation was built because the ecosystem required it. The infrastructure for conversational advertising, systems that process deep intent, create competition for demand, and measure performance across non-linear decision journeys, will be built the same way.

What makes ChatGPT's ad rollout significant isn't because it proves ads can exist inside AI. It's that it proves the market, at scale, with real advertiser spend, before most of the ecosystem has the infrastructure to participate in it. The gap between demonstrated value and available infrastructure is where the next consequential build happens.

Velocity is building that infrastructure layer for AI products that don't have the scale to build it themselves. velocity.io