OpenAI missed internal user and revenue targets.
OpenAI Missed Its Own Targets. The IPO Cant Come Fast Enough.
The Wall Street Journal broke the story on Monday: OpenAI recently missed its own internal targets for user growth and revenue, less than a year before the company expects to list on public markets at a valuation that could exceed one trillion dollars. Two people familiar with the matter told the Journal that OpenAI fell short of an internal goal to reach one billion weekly active users for ChatGPT by the end of 2025. The company also missed several monthly sales targets in 2026, according to Bloomberg, which confirmed the broad strokes independently.
The timing is awkward. For months, OpenAI has been building a public narrative around scale. The company told investors and markets it had eight hundred million weekly active users as of December. It reported twenty billion dollars in annualized revenue in January, up from six billion the prior year. OpenAI CFO Sarah Friar described the growth as closely tracking an expansion in computing capacity. The numbers were supposed to demonstrate momentum. Internal targets, it turns out, were set higher.
What changed is not a quarter of bad luck. It is a competitive inflection. Rival Anthropic gained ground in coding and enterprise markets during precisely the period when OpenAI was missing its marks, according to people familiar with the matter. Bloomberg reported that Anthropic, structured as a public benefit corporation, has been more aggressive in courting business customers who want contractual commitments and dedicated infrastructure. OpenAI, whose growth has been driven heavily by consumer and developer API usage, found the conversion funnel slower than its internal models projected.
The stakes extend beyond one company missing a quarter. OpenAI expects to spend one hundred twenty-one billion dollars in its current spending period, according to Seeking Alpha. Against twenty billion dollars in 2025 revenue, that is a six-to-one spend-to-revenue ratio. No major technology company has come to public markets burning at that multiple. The comparison that people in Silicon Valley are quietly making is to the early infrastructure buildouts of the internet era, when companies spent aggressively to capture market share and trusted that revenue would follow. Some of those bets paid off. Many did not.
OpenAI is not a typical pre-IPO company. It has raised more than one hundred twenty-two billion dollars to date, including a forty billion dollar investment from SoftBank and a thirty billion dollar commitment from Nvidia, both finalized earlier this year. It has a revenue-sharing arrangement with Microsoft, whose investment and cloud partnership has been the financial backbone of the company since 2019. It is restructuring from a nonprofit governed by a board of directors to a public benefit corporation, a transition that will strip away some of the oversight mechanisms that were designed to constrain commercial incentives. The IPO itself is the plan, not the backup.
The question that matters now is not whether OpenAI is a significant company. It is whether the valuation math works at the price being assembled. At an eight hundred thirty billion dollar valuation in the most recent funding round, or the one trillion plus that some bankers have discussed for the listing, OpenAI would be valued at roughly forty to fifty times trailing revenue. For a company burning six dollars for every dollar it takes in, that multiple assumes a transformation in unit economics that has not yet arrived. If the best-funded, most-celebrated artificial intelligence company in history cannot reliably hit its own commercial milestones, the logical follow-up question is what that means for the entire sector that has priced itself on the assumption that it will.
Anthropic is not waiting for an answer. The company filed paperwork for a public offering of its own in late 2026, targeting many of the same institutional investors who are being asked to buy OpenAI at the top of the market. The two companies will present different stories to those investors. OpenAI will argue scale and momentum. Anthropic will argue that its enterprise-focused model, with revenue derived from contractual arrangements rather than usage-based API calls, is more predictable. Both will be asking for money at a moment when the fundamental question of whether foundation model companies can generate durable profits remains genuinely open.
OpenAI declined to comment on the specific missed targets. The company has not publicly addressed the gap between its internal goals and its public statements about user growth, nor has it provided updated figures that would allow outside observers to assess current performance against the targets it set for itself. The IPO process, when it arrives, will require certified financials and a prospectus reviewed by regulators. Until then, the only numbers available are the ones the company chooses to share and the ones that find their way into reporting by journalists with sources inside the building.
What is clear is that the pre-IPO sprint has intensified the pressure on every commercial metric the company tracks. The gap between what OpenAI told public markets and what it told itself is now a matter of public record. Whether investors care depends on whether they believe the next set of numbers will be the ones that close it.