C3.ai, Inc. (AI) Earnings Call Transcript & Summary

Good afternoon, everyone. We are almost wrapping up day 4 of the Morgan Stanley TMT Conference. I'm Sanjit Singh, infrastructure software analyst on the Morgan Stanley software team. I'm super thrilled to have Tom Siebel, CEO of C3.ai. Looking forward to having a conversation on the C3 AI story. Before I get there, we'll go through some disclosures. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley representative. Tom, this is true for me, this is the first time I got to spend some time with you on stage. You have a storied career in software. You've seen many cycles.

I wanted to kick off the conversation, having been at Oracle, you were -- started a pretty revolutionary company in Siebel Systems. Can you give us the sense of the story behind C3 AI? Was the premise on which the company was founded? And what was sort of the problem that you're looking to solve for in this era of software?

After we sold Siebel Systems to Oracle in 2006, we started thinking about what was happening next in information technology. The business had grown since the time I joined Oracle in '83 from about $200 billion globally to maybe $4 trillion. And it occurred to us that the next kind of big vectors on the information technology were going to be about elastic cloud computing, big data, the Internet of Things and predictive analytics. I understand at the time that we were doing this planning, AWS was selling about $1 million in software. So this elastic cloud hadn't quite happened yet. So we got together a team of people, sent out an e-mail on Friday, raised $20 million by Sunday, started up the company in January of 2009. And the idea was to basically build a software stack, a software platform and a family of applications that would allow companies to apply these technologies to solve problems that have never been solved before. So this is generally -- and these are all generally in the area of what we call predictive analytics or enterprise AI. So that was the idea. And now 12 years later, we've invested hundreds of millions of dollars in the development of this technology stack. We're addressing a market in enterprise AI that promises to be, according to IDC and others, order of a $600 billion addressable market in, say, 2025. We are, if not the largest, one of the largest companies in enterprise application software. And the game that we're playing, like we played successfully at Oracle and like we played successfully at Siebel Systems, the idea is to establish and maintain a global market leadership position in enterprise AI.

And so when you think about that first era of the company, call it, that first 8 to 10 years, you were really focused on serving some really large global enterprises in a few select industries. Can you give us the insight on why you focused on serving these kind of key lighthouse accounts for the first 10 years? As most kind of companies that are founded, they try and grow really, really fast with a small number of customers and expand in like the technology vertical. What was sort of the motivation behind focusing on these really strategic lighthouse accounts?

Well, we have a number of people who -- at C3 who I've worked with for many years, who are very experienced at very large enterprise selling. This is what we're good at, okay? And there's nobody in the world, from Morgan Stanley, to Goldman Sachs, to Credit Lyonnais, to Volkswagen, to Daimler that wasn't a very, very large customer of ours, either at Oracle or at Siebel Systems. So one of the -- a core competence for us were the relationships we have with these organizations and our ability to close those kind of transactions. So when we're a private company, the strategy was to form basically large lighthouse accounts, iconic companies in energy, in oil and gas, in manufacturing, in government services and what have you, and just establish these relationships. And the other nice thing about it is allowed us to fund the company through our customers rather than ring doorbells on Sand Hill Road. But in a private company, we're focused on very large enterprise agreements, and then now as a public company, we're in the process of kind of monetizing that into market share.

It makes a lot of sense. I have a 2-part question for you, Tom, next. And it's really around the platform itself and the customer journey. From the software stack that you've invested hundreds of millions of dollars in, can you talk about the essential defining characteristics of the C3 platform that makes building AI predictive analytics applications a more efficient process than stitching together a host of other point solution technologies?

Yes. So the alternative to us early on would have been IBM Watson, which is now, I think, shut down, or GE Predix, which I know is shut down, okay? I think that GE spent about $6 billion; IBM Watson spent scores of billions of dollars. And those -- so all the problems that they proposed to solving in the Super Bowl ads is exactly what we do. Now the -- virtually everybody, okay, every one of our large customers, 95% of my revenue, okay, last quarter, I don't think this is an overstatement, would have come from companies that spent years trying to build this themselves by stitching together all of these components from HTFS and Azure ML and DataRobot and whatnot. And you try to stitch those things together using structured programming. And companies spend years and hundreds of millions of dollars to billions of dollars on that -- at that. And to my knowledge, nobody has ever succeeded. We have something quite unique that we invented, and that is the idea of applying a model-driven architecture to this platform. And it reduces the level of complexity, and if you think about this model-driven architecture as a hierarchy of abstraction layers where we have abstracted this to the point where the level of complexity is, say, order of ten to the third rather than ten to the thirteenth. Ten to the thirteenth is a nontractable problem; ten to the third is a solvable problem. So our customers are able to design, develop, provision, operate these very large-scale enterprise applications very quickly, but they all go through a phase of trying to build it themselves. It's a necessary step in the process of becoming digital.

And that's been the trend in software, right? Like, CIOs have been somewhat risk-averse. They want to put all their eggs in one basket, so they go through the sort of...

Until they've tried that, they're not ready for us.

Right. So let's talk about the typical customer journey for a new customer onboarding to the C3 platform. What do customers typically start with? How long does it take their engineers to learn the platform? And how long does it take to roll out the first set of apps on average?

Well, I anticipated this question, so I looked at one of our large because you told me you're going to ask it. So I actually looked up the facts on this in the car on the way up. So this is one of our larger customers. We met them in December of 2016. We were introduced by AWS, and that was December of 2016. In June of 2017, we began a trial of, say, EUR 600,000 trial to -- for predictive maintenance. A little later, we started another EUR 600,000 trial for production optimization. Then in August of 2018, we entered into an EUR 18 million enterprise license agreement. In 2019, we extended that by EUR 1.5 million. In 2020, we extended that by EUR 2.5 million. And later in 2020, we extended it by another EUR 1.2 million. Then they decided that they wanted to expand this much largely globally, a bunch of number of assets, so we get to April of 2021. They extended the contract by an additional EUR 56.7 million. Then since then, we've done another couple of contracts. So I think this is now 7 or 8 or 9 contracts. But as you can see, this adds up to this land and expand, it gets to the point where it's significantly non-zero, okay? And so these are -- we're now into scores of billions of euros, and the economic benefit of this effort to this company last year that they have disclosed in the public audience was EUR 1 billion last year and it will be EUR 2.2 billion this year. That's what they've disclosed to the bankers is what they're going to get from this effort.

As you walk through that, Tom, I think the interesting thing about that to me was the EUR 600,000 starting with predictive maintenance, the EUR 600,000 -- then the EUR 600,000 with production optimization. It seemed like you need to start solving a single problem, go to the next single problem and then there's like this hockey stick of...

Then it becomes 17. You sell 1, you sell 2, like we announced this week a large expansion, I believe this week with LyondellBasell, which is a pretty large chemical company. And they skipped the trial phase a few years ago, built production optimization, process optimization, predictive maintenance across a number of their assets, and then decided to expand this globally. And I don't know if we disclosed the size of this transaction or not, but it was bigger than a bread box in terms of a very large expansion that took place last quarter. And we did 3 transactions last quarter in, say, the $10 million to $50 million range, and that would have been one of those.

So it seems to me that those first couple of years in terms of winning the hearts of the customers is particularly important. And so if you get those first couple of years right, you're probably going to be pretty successful over the life of that customer. So to what degree do customers require services help from C3 AI to help deploy or operate the platform in that additional year? And then how does that -- like how does the handoff happen in sort of year 2, year 3?

Customers today are becoming -- so our idea is that we're in the business of providing technology to customers so they can build these applications themselves. And this organization today, I think, has hundreds of people building projects using our platform across the entire global enterprise, and this is a pretty large company. Now the first step was build the technology, and we saw the very, very difficult technology problem that IBM failed to solve, that GE failed to solve, that Siemens failed to solve, that Schneider failed to solve, but we solved. Okay, now the -- so the further thing was get the technology and get it out there to work. Now associated with transferring the technology, there is some professional services work up for us. This is why professional service today varies between about, I think, 15% to 20% of our revenue. As this product matures, and you'll see this with the release of C3 version 8, which will be released later this month at our User Group Conference in Miami, you'll see the development of the whole C3 community, much richer documentation, much richer online support, the -- a new set of tools, integrated set of tools for deep code, low code, no code development, faster provisioning so that customers can get the applications up live faster independently of us. So that's where this goes. We've invested, basically rewritten the entire stack over the last 3 years, and we'll release that later this month.

Ahead of your Transform event that's coming up?

At Transform.

That's right. Awesome. So let's talk a little bit about Baker Hughes, one of your marquee customers, one of your marquee partners. That contract was recently restructured for the second time at the end of last year. Can you discuss the reasons why that contract was restructured? And then going forward, how confident are you that they're in a position now to deliver on what are some pretty compelling revenue commitments over the next 3.5 years?

I see that this contract restructuring is the subject of some discussion on the Internet. It's a little bit fascinating. Okay. The net-net was we extended the contract by a year, added how many millions of dollars to it? Added $45 million to it, but basically made $350 million, an irrevocable, nonrefundable revenue commitment to the company over 3.5 years. Now how that is not in the best interest of the shareholders of C3 is absolutely baffling to me. This would be an IQ test. So this was a win-win deal for Baker Hughes. It was a win-win deal for C3 shareholders. But tell me, what CEO, what responsible CEO of a software company is not going to agree when a partner wants to expand their relationship, okay, and offer you $350 million in irrevocable, nonrefundable revenue commitments over 3.5 years? What's not to like, Sanjit? Okay. So it -- I've seen a few -- anyhow, I'm confident that it's in the best interest of shareholders. It had a huge impact on RPO. So the RPO for the company is now, what, I think, $470 million, up 90% year-over-year. And I'll add the RPO is 160% of the predicted revenue for the company this year. Compare that to RPO as a percent of predicted revenue for any company out there, and I think it's like maybe 3x higher.

Got it. Let's talk about...

So this is like goodness and light.

Let's talk about some of the other partnerships that you've struck, because you've struck a number of them over the last 2 years with the likes of IBM, Microsoft. Google is the one that's been coming up more and more in your comments on the earnings calls. What partnerships seem to be generating the most momentum today? And what are the ones that you think will be more material contributors to the business over time?

Well, Baker Hughes, let's start with Baker Hughes. What gives us is an indication that's going well? Okay, well, none of my revenue last quarter was to Baker Hughes, correct me if this is not clear, but none of my bookings last quarter, thank you, were to Baker Hughes. 32% of my revenue was with Baker Hughes, okay, up from 17% a year ago. And that is an increase of 700% in revenue year-over-year. So I would say that is generally positive momentum. When you -- and I think there's some short seller that nobody knows out there that somehow is indicating and somehow that's an indication of a business relationship that is falling apart. Well, I wish all my business relationships would fall apart like that, okay? The -- I think to date, I've sold about $200 million with Microsoft [indiscernible] into the enterprise, so that's an incredibly successful relationship. We did announce a relationship with IBM. I would say we never got a lot of traction with that. Honestly, I'm not sure why. The -- so let me kind of give you -- but the answer was what's our going -- the one that I'm generally excited about is Google, Google Cloud with Thomas Kurian. He has reinvented that company. He has completely replaced the sales organization. He runs the company in a very non-Google like fashion. Like you can do your job anyway you want as long as it's exactly the way that Thomas tells you, okay, or else you can go to work someplace else in about 4 minutes. And these guys are really leaning in. So they're -- they want to go to market through the application layer. So rather than -- their strategy is rather than compete with AWS and Microsoft based upon CPU seconds and GPU and storage hours with arguably a better architecture, they want to go after the application layers, to get through optimization, supply chain, supply network risks, CRM, demand forecasting, fraud detection. And so the net effect is they're selling CPU seconds, okay, and storage hours. But if you want to find a company that has 42 enterprise applications for oil and gas, for manufacturing, for telecommunications, for defense and intelligence, and for financial services and banking that has 42 turnkey applications to run on top of the Google Cloud, there is exactly one door you can knock on in the world. And that's the door that you knock on. So that's -- these guys are really leaning in and I'm very -- honestly, we haven't sold a lot with them yet.

Yes. You see the ...

I think these guys are all in. They're -- he is reinventing that organization in an impressive way.

Right. Let's talk about a little bit, I mean in uncertain times, geopolitically, with the Russia and Ukraine situation. And in that context, right, you've signed a $500 million contract with the Department of Defense that allows essentially DoD to streamline the purchase of C3 technology. How do we think about, I guess more broadly, C3's opportunity within DoD? And with this $500 million contract, what do you think would be reasonable to expect in terms of business being booked against this purchase authorization?

Well, it's the second contract of this nature. So this is basically a production operation agreement, one for $0.5 billion. We have another one for $100 million, but this $0.5 billion agreement was initiated by the Missile Defense Agency, with whom we do a lot of work. And it's classified work. But this gives basically anybody in DoD, they can purchase C3 applications or the C3 platform without doing RFP and without doing a competitive bid. So this definitely accelerates business. It makes it very easy for people who want to use our business, our product for defense, for intelligence, insider threat, contact tracking, predictive maintenance, readiness, logistics, what have you, it makes it very easy. So it will absolutely accelerate our business in the federal sector. Now I will never -- I mean in a steady state, the U.S. federal sector might be 15% of our business. I mean we're not going to be a federal contractor like some other people who are in this space. That's not what we will do. But to the extent that we have the opportunity to -- we don't sell to China. We sell only to U.S. and U.S. allies. And -- but to the extent that we have an opportunity to serve, we're -- feel privileged to do so.

Yes. It's a pretty uncertain and dangerous time.

As it relates to exposure to Russia, we -- okay, we have virtually no exposure to Russia. Did we have 2 deals that we were forecasting in Q1 -- Q4 and Q1 that we expected to close in that period of time with large Russian companies? Yes, we did. And I'll bet the probability of closure of those transaction is about 0. I think we can forget about -- that isn't going to happen. But we did have 2 deals.

They just lobbed against that.

I think we could be pretty confident that those don't materialize.

Understood. So 2 quarters ago, Tom, you were really frank with us and with me that you need to make some changes to the sales organization to improve their productivity on the [ booking ] side of the house. Last quarter, I mean you just flat out said it was a great booking.

We killed it.

Yes. So can you sort of pinpoint the change that you made? And then how you feel about the confidence of that sales organization to go out and close deals outside of obviously these -- some of these Russian deals that were in the pipeline, but just more generally, the ability to...

I mean this is a net-net. We brought in a sales leader from, say, a multibillion-dollar software company who basically ran sales like a sales administration function, okay. Sales administration, that would mean you're holding sales forecasts with generals and lieutenants and colonels on Monday, Tuesday, Wednesday, Thursday, Friday, what's your forecast, what's your forecast, what's on your forecast, what's your -- he adds the number up and brings it into my office, okay. But this is not how we do it, okay? And I knew it was not working, which is why I went out with Ed and -- Ed [ Humanu ]. We closed a couple of deals to save the quarter. Come on, we grew the business by 41%. Could -- life could be worse. Now that being said, this is not how we sell. And so we went back to the way that we've always sold, which is a strategic selling model, and so we did not restructure the organization. Same people, same places, same -- some minor changes. We really did not restructure sales. We just returned to our strategic selling model and our strategic matter of models, name the account, say Shell, say Coke Industries, say United States Air Force Rapid Sustainment Office. And we surround that group with a salesperson, somebody from the executive team, somebody from the [ resource or ] services team, somebody from the engineering team, somebody maybe from McKinsey & Company who is an expert in whatever that problem has brought us to get to the optimization of supply chain. So we surround that account with a really smart, engaged engagement team, and we solve the problem and make them successful. So that's what we returned to last quarter and it was hugely successful. I mean we closed -- where did we close? We closed 12 transactions less than $1 million, 3 transactions between $1 million and $5 million, 2 transactions between $5 million and $10 million, 3 transactions between $10 million and $50 million for 42% top line growth. Now understand, the year before we went public, which would be last year, the growth of the company was 17%, okay? In the first quarter, the company grew top -- year-over-year growth 29%, second quarter 41%, third quarter 42%. The expectations, I think, that you and [ others ] published for this year is like 33% growth. So we have a company addressing like $0.5 billion plus -- excuse me, $0.5 trillion plus addressable market opportunity. We're, if not the largest, one of the largest companies in the space. We're growing at a 42% compound annual growth rate and the company is trading at an enterprise value of 4x revenue. So give me a break, okay? So this looks like -- I mean something is not quite adding up in that picture, is it? Even in this market, it's not quite adding up.

You mentioned something on sort of the...

Ask me about selling stock. Please, don't forget it.

Yes, I know.

Don't forget, Sanjit. We're running out of time.

We'll close it. One last question on the growth equation. So -- and something that you've not talked about going back to the IPO, but in a simple just arithmetic P-times-Q, with Q being kind of the number of customers, when we think about the long-term opportunity that you're executing against, is this going to be -- is C3 going to be the type of company that's going to have 1,000 customers paying, I don't know, $1 million or $2 million, or is it going to be a company that has 10,000 customers or 100,000 customers? Your former company got tons and tons and tons of customers. What's the right -- what does that growth equation look like when you look out...

It's a combination of that. I mean we will be doing large enterprise major account one deals like we used to do before we were public. We'll do large enterprise deals like we do now. And then we have low-end products like Ex Machina and CRM, which gets you to tens of thousands of customers paying hundreds to thousands of dollars. So it's P-times-some-factor-plus-Q-times-Y. And so we're going after all. The idea is to establish a market leadership position in the high end, in the medium end and the bottom end of the market.

Understood. All right. So last question...

Again, growing at a 42% compound annual growth rate. Okay. So...

So -- yes, so top decile growth.

Which would be safely in, what, top decile, [ Rapo ]? You're an expert at this, I'm not. Top decile? Okay. Good.

Okay. So why don't you address the stock sales?

Stock sales, Tom. Why is Tom dumping stock where the stock is like this low? Well, it never happened. Okay. Tom did sell stock. But the last time Tom sold stock was in November when the stock was a different price. Somebody tell me what the stock price was in November. It wasn't $20. It was being sold under a 10b5 program. Under that 10b5 program, Tom would sell 1.5% of his fully diluted equity position every month, 1.6%, which means I always had 98.4% left. So subject to a floor, and let's just say that floor is higher than where we are now. So that -- so did -- was Tom the primary source of capital for the company for some years? Yes. And did I sell 1.6% of my stock every month under a 10b5 program, I think, most recently last November? He did, guilty as charged. And I still have had 98.4% of my position left. And -- you get the idea, to adjust every quarter. And so that's the extent to which Tom was dumping stock at current values. Now that being said, let me also say, you know that the company has announced that we are prepared to go participate in the market and retire shares. We think there are some real opportunities here to retire shares at a screaming bargain, and you can expect that we are and we'll be doing that.

The $100 million repurchase...

Bingo.

I got you. We have time for one more question. And I think probably one of my favorite parts of the C3 story is kind of the application part of the story, particularly C3 AI CRM. Can you talk to us about the traction you're getting there because when I talked to...

Talked to you about what?

C3 CRM. The CRM app. What's the traction been there? Are there other kind of AI-infused applications that customers can easily deploy and get started or become...

We have, I believe, 42 enterprises app, turnkey enterprise applications now. I just opened a development center in Guadalajara, okay. Guadalajara has a -- Guadalajara is a pretty nice place. They have 13 major universities, 1 really good, Tec de Monterrey, 90,000 graduate students, a lot of tech companies, Siemens, Oracle, Bosch, others that are there in this community. The governor is behind us. So you expect that we're going to be moving our applications development. And then they speak English some of the time, same time zone and it's 3 hours away. So this works. This is not trying to like to do distributed engineering all around the world. So as you see -- and there are some human capital constraints in Silicon Valley that are well understood that we need to relieve. So the big picture is, going forward, we'll be advancing the platform and the core technology in Silicon Valley, and we'll be building the apps, okay, in Guadalajara. And so you can expect -- I would not be surprised if 3 years from now we have closer to 150 turnkey applications than 42. I think what we've done in CRM is very exciting, okay? We're just now initiating some trials out there. And I think one thing -- basically, what we do -- the idea behind our CRM offering is we're not competing with Salesforce, we're not competing with Dynamics, we're not competing with SugarCRM or any of these other -- Veeva, who's my friend, Dave Schmaier of Vlocity, okay? We sit on top of them. So these -- I mean these companies have like huge, hundreds and hundreds and hundreds of millions of dollars of investments in Salesforce, and we're going to sit on top of that or Siebel in the case of Oracle or Dynamics in the case of Microsoft, and make those applications immediately predictive. So AI-based forecasting, next best product, next best offer, customer churn. And so we can add a lot of value to their existing investments. And as somebody who's not entirely unfamiliar with the CRM market, I believe that will be a hugely successful product.

Well, we're out of time. Tom, thank you so much for attending the conference. Thank you for the conversation. I really appreciate it.

Thank you, Sanjit.

All right.