By Elena Vasquez

More than 100,000 advertisers have tested ChatGPT Ads, but a $1 billion annualized revenue run rate has not resolved the central issue: whether brands can treat the platform as a durable media line rather than an experiment.

OpenAI launched ChatGPT Ads about nine months ago and reached the $1 billion run rate this summer while adding new formats, partnerships and tools at a rapid pace.

OpenAI must now show that the ads move hard business metrics, a test that chief marketing officers and data specialists have judged mixed. Twilio CMO Chris Koehler said: "We've dabbled in it." Intuit CMO Thomas Ranese described the company as "experimenting."

What do early performance studies show?

Maximilian Kaiser and Christian Schulze's working paper, which remained under review, tested 102 advertiser sites. ChatGPT Ads generated $0.89 in revenue per U.S. session, compared with $2.45 for Google search ads, a difference of $1.56 per session.

The researchers told CMO Insider by email: "What makes us cautious about 'it will come with time' is the size of the gap."

Adthena supplied a separate benchmark based on 30 enterprise brands from July through September: 52% struggled to recoup ChatGPT ad spending, while 26% matched or beat Google ads. The remainder reported mixed results.

OpenAI's October case study of Evernote provided a counterpoint. The note-taking app recorded a trial-to-paid conversion rate about 10% to 20% above its historical average across paid-acquisition channels earlier that year.

Can ChatGPT build enough ad inventory?

OpenAI's ad growth faces three constraints: the size of ChatGPT's free user base, a limited supply of conversations relevant to advertisers, and a ceiling on how many ads users will tolerate before leaving the product.

Dave Dugan, OpenAI's head of global ad solutions, said in a June interview that ad delivery and volume had improved as the company grew more confident in the ChatGPT Ads user experience. Brands still struggled to secure consistent visibility at scale.

Qualcomm CMO Don McGuire said: "It's really hard for these LLMs to find audience. Unless it's super, super broad, targeting is not really happening right now, so that really begs the ROI question." The return-on-investment question therefore remains central to scaling the channel.

Will ads inside answers improve returns?

Kaiser and Schulze proposed one route to better performance: shift ads from cards beside ChatGPT conversations to placements woven more naturally into answers. The trade-off is substantial. OpenAI's ad principles say ads will not influence ChatGPT's answers and that the company prioritizes user experience over revenue, making a more intrusive format unlikely.

Will better measurement settle the ROI debate?

Measurement remains a "sore point" for marketers, Adthena CMO Ashley Fletcher said. OpenAI has introduced a measurement pixel and conversion bidding, allowing advertisers to set targets for cost per acquisition or cost per click.

Fletcher said the tools should give CMOs better data for modeling future spending, but warned that the findings may disappoint. Cleaner measurement may show ChatGPT ads mostly "nudge people" during early brand consideration, not deliver the rapid conversions advertisers expect from established platforms; Black Friday will test that proposition.

Can advertisers turn tests into permanent budgets?

Hans Fischmann, AdRoll's vice president of product, tags every ChatGPT ad the company runs and uses the data to inform broader media buying. "Someone asks ChatGPT about a hotel on Tuesday, and sees that brand on their TV Thursday," Fischmann said. "That's the full-funnel play, and it's what earns ChatGPT a permanent spot on the plan."

Natural Intelligence, which operates Top10.com, BestMoney.com and a performance-marketing business, offers a more direct early win. Jonathan Vardi, its head of strategy, moved ChatGPT Ads from the test budget to the growth budget, where spending scales as performance improves.

Vardi said the team turns each product's pain points into specific ads designed to appear when users ask difficult questions. He also treats accurate conversion reporting as essential to the algorithm's learning.

"If you don't tell ChatGPT what users are doing on your site after they click your ad, you're not allowing the ads algorithm to learn which users actually engage with what you have to offer and become paying customers," Vardi said.

"Stopping at views or clicks really constrains your ROI now that there's more competition on the channel."

OpenAI has advertiser reach and a $1 billion revenue run rate. It still lacks the consistent performance, targeting scale and conversion evidence required to turn broad testing into recurring media commitments.