Here’s what OpenAI needs to do to reach its billion-dollar ad ambitions

Just seven months after its initial launch, OpenAI’s ad business is running at a $1 billion annualized revenue run rate. By the end of the year, the tech company projects its ads business will rake in $25 billion in ad revenue. That projection surges to $100 billion by 2030.

Notably, there’s a difference between annualized revenue run rate and actual ad revenue. OpenAI is effectively saying that the goal is doable if its current revenue holds steady for a year.

But change is exactly what advertisers need to spend beyond test and learn budgets. To meet the goals it has set for itself, OpenAI needs changes around measurement, inventory, data, adoption and infrastructure, according to Krystal Scanlon, senior platforms reporter at Digiday. Scanlon recently joined the Digiday Podcast to check in on how OpenAI’s ad business is shaping up.

Here are five things OpenAI needs to make good on its billion-dollar ad revenue bets:

This interview has been lightly edited for clarity. 

Measurement and attribution

Scanlon: Measurement seems to be a big question mark for a lot of advertisers that I’ve spoken to. Because obviously, when it comes to having massive budgets, no marketer in this world, from what I know of, can literally hand over thousands or hundreds of thousands of dollars in any capacity — or even millions — without being able to justify it, and have a way to measure that this is exactly why this ad should be placed here on this platform. 

Data and privacy

Scanlon: I’ve had other advertisers saying that they are uncomfortable with signing contracts around the conversion side of things due to the Ts and Cs [terms and conditions] that OpenAI have provided, which are just considered really rigid and there’s too much liability on the advertiser — probably around things like the data and how it’s going to be used and shared. There’s still a lot of ironing out that needs to be done before this can be anywhere near the scale of a Google or a Meta. There’s a lot more privacy issues right now at the moment. And then obviously, we’ve got the whole AI regulations, the issues around like what AI is doing in terms of when it’s going a little bit rogue here, there, and everywhere. (OpenAI agents escaped a test environment and hacked into other companies, including software company Hugging Face and multiple government websites, heightening global concerns around AI safety and privacy.)

Inventory

Peterson: When OpenAI first launched ads in ChatGPT, it was asking for these big six, seven-figure commitments from advertisers to spend. But then, you know Kimeko, [Kimeko McCoy, senior marketing reporter at Digiday] at the Digiday Programmatic Marketing Summit in May in the behind closed doors town halls we had with the ad buyers who were in attendance, a number of them were saying, “Look, we’ve committed to that spend, but we’re not actually spending that money. Like there’s not enough inventory here where we just have money that just is sitting there unused. We would like to spend that money, but again, there’s not enough inventory.” I’m still hearing that from ad buyers.

Adoption

Peterson: OpenAI needs to build out this long tail of small and medium-sized advertisers, just like Google and Meta and Amazon have done, the Shopify integration seems like one way for it to do that. Especially because, again the integration of the product feeds that seems to be the thing that OpenAI has done this year that’s had the most positive impact on its ad business.

Infrastructure

Scanlon: They’ve got Criteo, they’ve got Kargo, they’ve got numerous different partners as well. That automatically opens them up to hundreds of thousands of different advertisers that they could have on board already. They are seemingly getting these advertisers on board — and the more partners they have, they’re just opening themselves up [to greater ad spend].