OpenAI Ads: Where Search Meets Social (And What the Data Actually Shows)

When OpenAI first introduced ads, Booyah did not jump in. The rates were unreasonably high, and we were not going to recommend our clients spend against an unproven channel at a premium that could not be justified. So we watched, evaluated, and waited.

What changed is not that the rates have normalized; they have not. What changed is the pace of innovation on the platform. Features, formats, and targeting capabilities are evolving fast enough that sitting on the sidelines is starting to feel like the riskier move. We are at a point where the brands that have been testing, building playbooks, and learning the platform’s nuances will have a real advantage when rates do come down and conversion optimization matures. We would rather be shaping best practices than catching up to them.

With 900 million weekly users on ChatGPT, the audience is there. These are not passive scrollers. They are people typing specific questions, actively seeking answers, and making decisions in real time. Getting in front of that intent, even at a premium, has strategic logic.

What the Early Data Tells Us

The most interesting thing about OpenAI’s early performance is that it does not fit neatly into any existing category.

CTR Index by Platform

All platforms indexed to a common baseline (cross-platform average = 100). Client-specific figures are not disclosed.

OpenAI’s CTR index of 95 puts it in its own zone. It is not behaving like traditional search, which sits well below the average at 56. But calling it “social-like” overstates it too. Meta is at 142, a full 50% higher. What you can say is that OpenAI’s click engagement is much closer to the social end of the spectrum than the search end. Users are clicking at rates that reflect genuine interest, not just navigational intent.

CPC Index by Platform

All platforms indexed to a common baseline (cross-platform average = 100). Client-specific figures are not disclosed.

All platforms indexed to a common baseline (cross-platform average = 100). Client-specific figures are not disclosed.

The CPC picture is where it gets uncomfortable and honest. At an index of 238, OpenAI is the most expensive platform in this set by a significant margin, sitting well above even traditional search at 134. That is the reality of where rates are right now.

The case for being there anyway is this: CPCs on new ad platforms almost always start high. Early inventory is limited, demand from early adopters compresses supply, and measurement tools are not mature enough yet to optimize toward conversion efficiently. These are solvable problems, and they tend to get solved as platforms scale. The brands with a head start on creative learning, audience testing, and performance benchmarks will be better positioned to capture the efficiency gains when they arrive.

What This Signals for the Broader Landscape

LLMs are not replacing Google. They are complementing it and carving out their own space in the consumer decision-making journey. Some users move fluidly across channels: a ChatGPT query to frame the problem, search to compare options, a retargeting ad to convert. But a growing segment never leaves ChatGPT at all. Research, evaluation, and decision happen in a single session, compressing the awareness-to-conversion cycle into one environment. For those users, if your brand is not present inside that conversation, it does not exist. Both behaviors make the case for being there.

The early CTR data supports that case. Users encountering ads in this environment are engaging at rates that reflect genuine intent, not passive exposure. Whether that intent translates to downstream conversion at a rate that justifies today’s CPC is the question we are actively working to answer.

As the landscape continues to evolve, we are committed to being more than early adopters. We aim to help shape what best practices look like in this new era of search through testing, learning, and bringing those insights back to the brands we partner with.

Indexed performance data reflects early results across Booyah client campaigns. Individual results will vary.