OpenAI's advertising business reached a $1 billion annualised revenue run rate in under 200 days. Google's AdWords took roughly four years to reach the same number. However you feel about ads inside a chat window, the growth rate is not in question, and it is now backed by an infrastructure expansion that changes who can buy in.
On 31 August 2026, OpenAI opened beta self-serve access to Ads Manager for advertisers across 31 European markets, following an August rollout that had run through agency partners only. Self-serve means a startup or small business can build and manage a campaign directly, without an agency in between. OpenAI's own monthly revenue from ads is currently around $83 million, and the company is targeting $2.5 billion in recognised advertising revenue for 2026, with a long-term ambition of $100 billion by 2030.
Where the ads actually sit
Ads appear as sponsored links beneath AI-generated answers, not as a separate results page. Crucially, they are shown only to Free and ChatGPT Go subscribers. Plus, Pro, Business and Enterprise tiers remain ad-free. That single detail matters more than it looks.
If your customer base skews toward people who pay for a subscription to the tools they use daily — which describes most B2B SaaS buyers with any budget authority — the audience seeing these ads is disproportionately the free tier: students, casual users, people not yet paying for anything. That is not a reason to ignore the channel, but it is a reason to size your expectations correctly before you commit budget to it.
Why this is a new kind of channel, not just a new ad network
Search advertising and organic ranking have always been visually and structurally separate — a labelled ad block above a set of unlabelled results. A sponsored link sitting beneath a synthesised answer, in the same conversational flow the user just asked a question in, is a different object. It has not existed at this scale before, and nobody outside OpenAI has a mature playbook for how it performs against a direct-response benchmark.
That cuts both ways for a small SaaS. The upside is a channel your larger competitors have not yet optimised either — everyone is early. The downside is that "advertise here" and "get cited here" are now two separate things happening in the same box, and a founder chasing AI visibility needs to be clear about which one they are pursuing. Earning a mention in an answer through genuine product fit is not the same purchase as buying a sponsored link beneath one, and the second does not substitute for the first if your buyers are mostly on paid tiers where no ad ever appears.
What to actually do about it this week
Do not rush to buy media here based on the growth headline alone. A billion-dollar run rate tells you the channel scaled; it tells you nothing about your specific cost per acquisition. Ask for a small, capped test once self-serve access is available in your market, and measure it against your existing paid channels rather than against the hype.
Check whether your buyer is even in the audience. If your ideal customer is a founder or ops lead who pays for ChatGPT Plus or above, this specific ad placement will not reach them at all right now. Your organic-citation work — being the product an assistant names when asked, not the one it advertises beneath its own answer — stays the more relevant lever for that audience.
Watch the pricing model as it matures. Self-serve access at this stage typically means simple campaign setup ahead of a fully mature auction. Early movers into a new ad platform have historically found costs rise sharply as competition catches up, so testing now, before the market prices it correctly, has a real advantage if the audience fit is right.
The bottom line
A billion-dollar ad business built in under 200 days is a genuine milestone, and self-serve access across Europe makes it reachable for small advertisers for the first time. But the audience is explicitly the non-paying tier, and that alone should decide whether this belongs in your marketing budget this quarter or on your watchlist for next year.