C3.ai's debut shows public investors pay up for enterprise AI

C3.ai more than doubled on its first day while growth slowed to 11%. Public markets price AI software on expectations, and private rounds will follow.
A 120% gain on the first day of trading greeted C3.ai, the enterprise AI software company founded by Tom Siebel, when it listed on the New York Stock Exchange under the ticker AI on 9 December. The shares priced at $42, above a range the company had already raised once, for gross proceeds of $651 million.
The growth behind that price is slowing. Revenue rose 71% in the fiscal year to April 2020 but only 11% in the six months to October, so the market is paying for the category before the numbers arrive. That reaches well beyond one stock, because public prices set the ceiling for late-stage private rounds. A debut like this lifts what investors will pay for AI software companies still in private hands and opens the path for those preparing to list, which makes the next few quarters more telling than the first day.
Under the label, the prospectus shows a concentrated business. C3.ai sells a platform and ready-made applications that large industrial and public-sector customers use to run machine learning on their own data, mostly on three-year subscriptions. Its three largest customers brought in 44% of revenue last fiscal year, and many contracts include paid implementation work. A few very large accounts plus services on top of software is the risk a first-day price leaves out.
Founders selling machine learning to industry now have a public comparable, which helps in pricing private rounds, and a detailed map of the costs: long sales cycles with very large customers, professional services on top of software, and revenue that rests on a few accounts. The multiples the market now signals go to AI-first companies that reach repeatable deployment, with standard applications instead of custom projects. C3.ai already charges part of its fees as runtime by compute hour, a sign that pricing tied to usage is becoming normal in AI software. Fanaura starts its conversations with enterprise AI founders at exactly that line between custom work and product.
C3.ai's first quarterly report as a listed company is the near test. Growth moving back toward last year's pace would support the price; another half-year near 11% would not. The wider test is the pipeline of listings: how many AI and data software companies file in the coming months, and whether they also price above their ranges. Concentration runs deepest. If it falls as the company builds sales to mid-sized firms and through partners, enterprise AI can scale beyond large custom projects, and if it holds, the category remains a services business with a software valuation.
The debut confirms public demand for AI software, but a first-day price is not proof of a business model. Revenue that repeats across many customers is, and that is why progress counts in years rather than in trading days.
Building an AI-native company? Apply in five minutes. →
More from Fanaura
Building AI software that deploys the same way at every customer? Send us the project.
Fanaura invests its own capital and the capital of its limited partners (LPs), usually through a separate SPV for each deal. If we see a fit, we get in touch.
Get the news by email
One short note each weekday. No marketing, and you can unsubscribe at any time.