C3.ai IPO: What It Means For Enterprise AI

C3.ai recently filed its S-1 with the Securities and Exchange Commission and the public offering is likely to hit the markets in the first half of December. The company is expected to raise over $400 million and list on the NASDAQ exchange under the ticker AI. 

The CEO and founder of C3.ai is Tom Siebel, who is a legend in Silicon Valley. He got his start in the mid-1980s with Oracle and helped lead the growth of the relational database market. Then in 1993 he founded Siebel, which pioneered the CRM (Customer Relationship Management) category.

Yet Siebel is more than an entrepreneur. Keep in mind that he has written several important books, such as Digital Transformation (I wrote a review of it for Forbes.com).

Background

C3.ai, which Siebel founded in 2009, has a market opportunity that is much larger than his prior ventures. According to his^{shareholder letter: “Today, at the confluence of these technology vectors we find the phenomenon of Enterprise AI and Digital Transformation, mandates that are rising to the top of every CEO’s agenda. The global IT market exceeds $2.3 trillion today.”}

Of this market, the size that C3.ai is addressing is at about $174 billion this year and is expected to reach $271 billion by 2024. 

“The company should do well both during and after COVID as demand for easy-to-use AI / ML solutions in the enterprise is set to increase significantly,” said Yiannis Antoniou, who is an analyst at Gigaom. “We are still at the beginning of a AI-fueled enterprise digital transformation, and C3.ai is positioned well as a pioneer in this category. Given the quality of its offering, its long list of major customers and the relative immaturity of some of its competitors, the offering should prove appealing to investors betting on post-COVID economic rebound in the general economy, as well as those interested in expanding their tech holdings.”

The Future For C3.ai

The growth has certainly been strong. During fiscal 2020, revenues jumped from $91.6 million to $156.7 million or 71% year-over-year. One of the keys to the success has been the significant expansion in average customer contract values. From 2016 to 2020, they have gone from $1.2 million to $12.1 million, showing that the software is strategic. 

Yet there are skeptics. “Deploying exclusively or almost exclusively to a third party hosted cloud applications and platforms is not the ideal model for every enterprise scenario,” said Christopher Savoie, who is the CEO of Zapata. “Frankly, CIOs try to avoid any kind of vendor lock-in when it comes to their data and data analytics. There are many reasons for this. To name just one: many multinationals data sovereignty laws make such a deployment model tricky. The other thing to note is that deployment models such as this work best when the ETL portions of the data workflow are large but the computational tasks relatively straightforward, such as the k-means and decision tree algorithms offered in Ex Machina. Advances in neuromorphic computing, quantum computing and other exotic compute approaches will shift the focus to the compute side of the equation. In such cases, a more distributed, hybrid cloud platform that leverages on-prem compute resources will almost certainly be more performant and cost effective. The need to lift and shift data from a single platform in order to take advantage of emerging compute technology may pose real challenges down the road.”

But while technology roadmaps are essential, so is the need to provide short time-to-value for customers. And this is something that Siebel has much experience with. He estimates that his system has already generated billions of dollars in annual economic benefit. 

“AI will disrupt today’s business models because it will transform the core of what differentiates a company, like its IP,” said Andy Bane, who is the CEO of Element Analytics.  “Industrial companies will require a new Industrial Stack to deploy AI at scale that combines IT and OT systems and data through edge-to-cloud deployments, which is far more complex than the consumer or enterprise stacks. C3 positions itself as the single-vendor AI platform, with an underlying model-driven architecture, as the superior approach versus assembling best-of-breed solutions from an ecosystem of vendors requiring integration. Ultimately, users will have to determine their appetite for going with a single vendor or picking best-of-breed solutions to meet their needs.”

Tom (@ttaulli) is an advisor/board member to startups and the author of Artificial Intelligence Basics: A Non-Technical Introduction and The Robotic Process Automation Handbook: A Guide to Implementing RPA Systems. He also has developed various online courses, such as for the COBOL and Python programming languages.