OpenAI’s ambitious (and unrealistic?) advertising targets
Within just five years, the company believes it can reach $100 billion in ad revenue
Barely launched in the advertising market, OpenAI is already setting its sights high. Buoyed by what it calls “encouraging early results,” the creator of ChatGPT is forecasting explosive, unprecedented growth, according to documents shown to investors and reviewed by U.S. outlets The Information and Axios. Within just five years, the company believes it can reach $100 billion in ad revenue — a level that would place it among the global top four, behind only Google, Meta and Amazon, but ahead of YouTube and TikTok.
OpenAI’s optimism rests on two pillars. First, the ever-expanding adoption of ChatGPT. Three and a half years after its launch, the chatbot already attracts more than 900 million weekly active users. The company projects that figure will reach 2.75 billion by 2030 — on par with Facebook or Instagram, and ahead of TikTok. Second, it is banking on a strategic position closer to users’ purchase intent, which could translate into higher conversion rates. Together, these advantages are meant to make the platform a must-have for advertisers.
$8 million in a month
For now, however, OpenAI is still in the early stages of its advertising push. A first pilot phase was launched in early February, initially limited to the United States. It has since been extended through at least the end of April and expanded to three additional markets: Canada, Australia and New Zealand. “We hope to continue expanding into many more markets this year,” the company says. Ultimately, ads are expected to be rolled out to all free users, as well as subscribers to its new low-cost Go plan.
In the first six weeks of testing, more than 600 advertisers ran campaigns on the chatbot. Revenue exceeded a $100 million annualized run rate—equivalent to just over $8 million in a single month. OpenAI, led by Sam Altman, is anticipating a dramatic ramp-up. It is targeting $2.6 billion in ad revenue this year, rising to $11 billion in 2027, $25 billion in 2028, $53 billion in 2029, and surpassing $100 billion the following year.
An advertising infrastructure to build
To scale, OpenAI still needs to build a full-fledged advertising infrastructure, which remains extremely limited today. The company has only just begun testing a self-service platform that would allow advertisers to create and manage their own campaigns, an essential tool to reach the millions of SMEs that underpin the ad models of Google and Meta. It has also lowered its minimum spend requirements and will need to strengthen its measurement tools to help advertisers assess campaign effectiveness.
That step will be critical to justify its pricing model. Unlike Google or Amazon, OpenAI has not opted for a cost-per-click model, shifting more risk onto brands. Its prices are also relatively high, exceeding those of streaming platforms, for example. Another challenge will be improving ad relevance. According to Adweek, some advertisers have reported click-through rates below 1%. A particularly complex task as the startup simultaneously seeks to expand ad coverage across a growing number of queries.
Ambitious or unrealistic?
Such a growth trajectory would be unprecedented. Google took 15 years to reach 12-digit ad revenues; Meta took 14. Amazon is not expected to get there until 2028, according to eMarketer, 12 years after introducing ads. But the context has changed. The online advertising market is far more mature, and the rise of AI is happening at a much faster pace. For now, OpenAI is also the only major player fully committing to this path.
Whether merely ambitious or outright unrealistic, these targets should above all be seen in the context of a potential IPO, likely by the end of the year. Beyond the headline figures, the goal is to convince investors that the company has a credible roadmap to monetize its vast base of free users. The stakes are high: achieving profitability will depend on it, especially as this audience generates significant costs due to the compute required for text generation.
Photo: OpenAI