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

Awesome. Thank you, everyone, for joining us today. I know we're late the day here, but we are joined by the CEO of C3 AI, Mr. Tom Siebel. Tom, thank you very much for joining us for this fireside that we have lined up.

Thank you, Mike.

Absolutely. And just a real quick logistic guidance to, I guess, viewers or listeners out there, there should be a Q&A function on your screen that's available to you. So I want to make sure we're maximizing your time as well. If you have questions you want to submit, please send them through and we'll make sure that we get to those. But otherwise, I know I have a lot of question at my side that I'd like to get into as well. And I think, first, Tom, again, thanks for joining us here. Maybe just to help level set for the listeners on the webcast, can you just provide an overview of the history, the vision of C3 for the investors who may not be as of the speed of today?

Okay. So thank you, Mike, and thank you, everyone, for your time. Let me get to put this into perspective. We've been at this in about 13 years, and we have developed a C3 AI, and we have spent about $1 billion in cost developing a software stack that allows companies to build predictive enterprise applications. So I think this AI market is very confusing because we have Databricks and DataStax, DataRobot, H2O, and Cassandra and DynamoDB, and it's presumed that we compete with some or all of these things. And we actually -- we compete with none of those things. And we use a lot of them, and we partner with all of those guys. Now the story is really quite civil. Okay. We have -- roughly, when we took Oracle public, it was about a $50 million business. When we [ fix ] Siebel Systems public, it was about $5 million business. C3 AI is rough numbers, about a $0.25 billion business. We had rough numbers growing at about a 40% compound annual growth rate. And the rough numbers has $1 billion cash in the bank. I think last quarter, we year on cash in the third quarter we were about $50 million. So I think that's rough numbers. We have about 20 quarters of cash left. Should we -- at the correct -- now this AI market is very confusing. And I think we've contributed to that confusion and allowing us to get thrown in with this mismatch with Databricks and DataStax and data cloud and Alteryx all these other things. It's really simpler than that. Now when we I'm hearing your slide now, is that right, Mike?

Yes, that's correct.

And we built the enterprise application software market, and I was there, okay, this with Oracle, [indiscernible], we basically took storage technology, computing technology, primarily for management and [indiscernible]. And with different related to database systems, we both development tools on that. And then we built enterprise applications like ERP, manufacturing, supply chain, asset management, CRM, what have you. And we let the market as SAP, as Oracle, as PeopleSoft, as Siebel Systems and these applications would allow us to tell -- report with perfect 2020 hindsight, for example, what our inventory levels were of our end process inventory at [indiscernible]. And Boeing might have -- it used to be roughly a $60 billion commercial aircraft manufacturer and they just have about $50 billion within process inventory and a 777, for example, would have about 1 million components of the [ built ] materials going in with the supply going from chain going from South Carolina to [indiscernible]. And so we have millions of parts of resistors and transistors and [indiscernible] power units in all these bins all around the world, and we needed to be able to count them accurate. And then 767, 787, 757 737, 707, what have you. And with this enabled us to report with pretty good accuracy what our inventory levels were to basically meet FCC requirements or it would tell us what our cash balances were or would tell us what our rate of device failure was, be it aircraft, be it autofill, be it tractors, be it radios, be it a medical instrument. It would tell us with perfect 2020 hindsight, what our customer share was at Verizon in Bank of America, 90 days ago, 120 days ago. What our customer satisfaction levels were -- and I wonder, if we don't report at name the bank, okay, what our broad cases were or what I have no cases were then the CEO gets to rewrite his way away, maybe he or she goes to jail and get tens of billions of dollars of fines like Deutsche Bank for details, and we go to bank as a wrap sheet longer than already made of [indiscernible]. So it's important to comply with these things. Now what we did -- and so this is what ERP is all about, and this is roughly -- these applications are descriptive in nature, and it's roughly a $500 billion software market today. It turns out it's a bit pretty good idea. Now what we did is we built this AI application platform that uses all of this AI componentry that I mentioned before that allows us to take these ERP applications, the CRM applications, these manufacturing applications, these supply chain applications and make them predictive to work with SAP, it works with [ more ] or it works with Salesforce, it works with whatever it might be and rather than simply becoming descriptive telling us with 2020 hindsight, it will tell us not only how many parts we had 180 days ago, it'll tell us more importantly, how many of each part do we need any spend for the next 120 days to meet the demand function rather than simply tell us how many of our aircrafts failed in the last 6 months, it will tell us which aircraft are going to fail before they fail. So we can fix this to avoid the fail with AI-based predictive maintenance. If we get into [ AI anti-money ] laundering as part of that, it will identify the event in real time so we can avoid the fraud event just simply rather than simply report on it. As it relates to customer churn, it will tell us as a Verizon, at the Bank of America or whatever it might be at a Philips Medical, exactly which customers are going to leave us in the next 120 days. By name shows an intervene, and avoid the customer churn. So this is what happens when applications become predictive and prescriptive. And so we have then built this platform and now, I think, 42 enterprise applications that provide these predictive capabilities for financial services, manufacturing, aerospace. And this is what enterprise AI is all about. So enterprise AI is not buying some [ speaker toy ] for $20 million -- for $20,000. Enterprise AI is about [indiscernible] our enterprise so that you're predictive and you can change the future. And this is a $600 billion addressable market. And when you think about it, when you have this capability, I mean how many companies really if we think 5 years from now, are we going to be satisfied, knowing how many devices failed, how many customers left or what the inventory levels were. I mean in addition to that, they go out a lot to know which customers are going to leave, which devices are going to fail and what inventory levels they need. So this is the market we address, this is enterprise AI. We're roughly, I believe, we're the world lead provider of this kind of stuff, roughly $0.25 billion business. We're growing at roughly a 40% compound annual growth rate. We have a rock-solid balance sheet with roughly $1 billion cash in the bank. And our goal is to establish and maintain our market leadership position in this emerging market. and to do that, we will be one of the lead software companies in the world.

Great. Great. Thanks for the overview there. And I'm thinking about the specific verticals that you guys currently touched, right? I know you're talking about some of these specific applications for the verticals. Can you detail, I guess, which should be you see the most traction with, which are up and coming, degrees of success in attacking these various markets.

[indiscernible] so we're -- when we think about -- these are the customers that were penetrated so far. That means our large enterprise customers and understand that one of these customers, and I know that whereas these guys are focused, everybody is focused on how many new customers they opened to this quarter, selling them a $20,000 product, okay? And I would say that our story is quite different than that, okay, whereby what we want to look at is we look at a company like Shell, okay? At Shell -- we have, I think, 46 applications in production. They're generating $2 billion in economic benefit from this year from these applications. And this would be AI applied. When we're talking about an asset, 46 assets, and that's that would be like Pernis. Pernis is the largest refinery in Europe that would be the order of say, 10 aircraft carriers or Nigeria LNG. Today, they're on the path they have 65 applications of this scale. We've got in production this year. So this is an organization that [indiscernible] have started with about 7 years ago, with maybe a $300,000 trial. And to date, they've paid us well over EUR 100 million. So whereas other companies are about how many new $20,000 or $100,000 customers you get. Our story is quite different. Our story should be how successfully I would be entertaining these accounts, and it's very successful. At Shell, for example, we do -- almost 11,000 pieces of equipment in production, okay? And they've stood before their analyst meeting at Bank of America and said they expect to get $2.2 billion in economic benefit this year. If we look at the annual report for [indiscernible] which is a large grid operator in Europe, okay, they're telling their investors, they're getting EUR 5 billion [indiscernible] from these initiatives. LyondellBasell, a large chemical company, spoke at our Users Group conference, and they're expecting to get $1 billion in economic benefit this year. So it's a very unusual story. And so our story is very much about -- this would be, for example, the rate which penetrated the United States Air Force, AI-based predictive maintenance for Air Force. Air Force has -- sorry, 5,000 aircraft, E-3 Sentry, B-2 bomber, B-1, F-15, F-16, F-35 Joint Strike Fighter, we're bringing up a new platform every month -- every 3 months. We'll have 22 platforms in production. And this is a situation where we've gone from a say, $1 million or $2 million application to a blanket agreement for $100 million. So this story is about penetrating accounts, not about closing another account. Cargill. The Cargill we're looking at basically solving problems associated with true distribution around the world. And this is this, for example, will be a large integrated energy company where over a period of 8 years, we began with a $300,000 trial, and now we've done in excess of EUR 120 million in business with these guys. That's 400x. This is a large financial institution based in New York, where we began with 3 trials for $2 million and now, they've paid us well in excess of $40 million. This is another large oil and gas company that's paid us in excess of $90 million. This is a large chemical company based in Texas, where we began with $9 million and then $14 and now $56 million. So our story is about penetrating accounts deeply about where we are becoming, I think, the most strategic technology provider they have rather than selling $20,000 products to another 200 customers this quarter. So it's a little bit different than everybody else in the space.

Sure. So maybe in that vein, I know we're talking about the expansion of these existing customers, right? That's the strategy that you're outlining right now. And I'm thinking through that go-to-market, I know you guys have partnerships, but can you help us better frame out the initiatives in place, just again, if I'm thinking about driving deeper penetration with those customers. What is in place? Are there initiatives to continue to make sure you're walking down that road with these customers?

Absolutely. We go to market in a huge gateway of Google Cloud. Okay. We go to market with a huge way with Microsoft, okay, and Azure. In the oil and gas business, we go to market with a huge partnership with Baker Hughes that you're well aware of. In defense and intelligence, we go to market with Raytheon, in the [ international ] intelligence community. So this is what the ecosystem that we're deploying look like, where we have a geographic organization, sales organization, major accounts, enterprise accounts okay, middle market and then mass market accounts in North America, in EMEA and Asia. We have vertical market units addressing financial services, manufacturing, aerospace, defense, health, telco, oil and gas and utilities. In each 1 of these verticals, we're focused on partnering with a large [ leases ] like with FIS is a $13 billion financial services company that is selling our products to all of their banking customers around the world. Raytheon, we're partnered with Raytheon to meet the needs of the defense and intelligence community in Washington, D.C. With Baker Hughes, we made the global [indiscernible] needs of the global market for oil and gas companies. ENGIE is energy efficiency. And then we go to a huge market and we're selling with 4,000 people at Google and we're got a speed out with Thomas Kurian or [indiscernible] at Microsoft. And said, this is what so we're expanding the partner ecosystem, expanding the verticals, expanding the vertical partner ecosystem and expanding the horizontal market system. Okay? What's been the result, okay? Our average transaction value has come way down or the diversity of our pipeline is much greater, okay? The growth rate has been accelerating. I think last year, the year before we went public, our growth rate was about 17%. In the third quarter of this year, our growth rate was, I think 39% or 40%. And so we've been accelerating the growth rate, and we've been penetrating these markets in a big way. But we're penetrating deep, deep, deep into these customers, not thinly across lots of customers.

Great. And we have an inbound question from one of the clients who are listening. I'm sorry to make you go back to it. But there was a slide I think you pulled up showing cumulative TCVs with certain customers. I may have missed that. Sorry, I know I'm juggling a couple of screens. So the question is on those cumulative TCVs. While the TCV is impressive, the concern is, are those subscriptions growing on an annual basis? Or is it something that we should expect is increasing slightly or flat but you just have a long runway of existing uses with that customer?

I would say that our largest customer might be 10% benefited -- so I would expect to continue to see this to grow. I mean federal government might be 1 quarter of 1% penetrated in Defense Intelligence Committee. This customer, which is about a $46 billion chemical company. We're probably 10% penetrated. This is -- again, this as a large financial institution in New York. I think about $4 trillion in assets were probably 2% penetrated. So you can expect to see this to grow. So we're focused on these customers being hugely successful. And of the various other software companies that you've invested in, okay, how many of their customers at Salesforce, at Oracle, at SAP, okay? All of which are fine companies, okay, we'll stand up on stage and say they're getting $1 billion, $2 billion, $3 billion, $4 billion, $5 billion in economic benefit or more. With that vendor, ladies and gentlemen, 0 is that answer. It's 0. That's how many. And in our case, it's the rule, not the exception.

Got it. Got it. And shifting gears to products, right? I know AI version 8 was released in March. And it has been, I guess, teased out by management as a major release on the 3Q earnings call. What can you tell us about the new version with respect to customer reception? And then the follow-up is what sort of traction are you seeing with offerings such as CRM and Ex Machina?

Well, Version 8, we'll see Version 8 adoption across 100% of the customers. I mean it's not rational not to use it. it's completely upward compatible. In some cases, it's 1, 2, 3 orders of magnitude faster, okay? It has dramatically improved application development tools, dramatically improved DevOp tools. We basically have collected. This is a project, it's been in process for about 3 years where we aggregated the requirements of our customers and delivered them. In addition, we're kind of massively increasing the number of applications that we're delivering to the marketplace for financial services for manufacturing, aerospace, health care, more to follow. We're just getting started with CRM and Ex Machina. I think the CRM product is going to be huge. It's going to be huge. And Ex Machina is kind of a 21st century version of Alteryx -- and I mean it will be a successful product. Will it be as big as what we do with the platform in these applications, no way know-how. I mean is somebody going to pass $120 million for it? No, okay. But will it be a nice high volume, high margin, low-price business, it will.

Okay. Okay. And I know we talked to the go to market a little bit earlier, but one of your important partners is Baker Hughes, continue to be a substantial and important partner for the company. Can you help us think through the customer expansion that's being generated through this channel today?

Well, gosh, I mean, there's a lot of tailwinds in the oil and gas business right now with oil at or above $100 a barrel. And these people are making huge investments. And we're an active sales motion with virtually every oil and gas company in the world outside of China and Russia. So it's -- and that's a huge opportunity. If you look at our Q3 bookings, are we over concentrated oil and gas, I don't think so. Our Q3 bookings -- in Q3, I think, was a record revenue quarter. It was a record revenue quarter, right? 2.6% of our bookings were in oil and gas. But the Baker Hughes partnership is huge. And so we're an active sales motion with them all around the world, okay, in the Middle East, in Europe, in North America and in South America and Canada. And we used to have [indiscernible] would have been Russia, but that's not going on anymore. Those days are over. But -- it's -- I mean they're a huge partner. Google is a great partner. Microsoft is a great partner. Raytheon is a great partner. I think we're really doing a good job of developing these partnerships.

Great. And public sector came up earlier as well. I know that there was the announcement recently around the $500 million authorization agreement with the Department of Defense as well as the FedRAMP Ready status that was assigned in March. So can you talk to what needs to be done on the federal side of things. I know it's a bit of a trickier market but obviously holds potential for return is there based on the overall size of spend [indiscernible] segment?

[ Were they ] dramatically investing our investments in the federal community in both civilian and defense and intelligence. Our footprint is growing very rapidly. We expect that to -- that is a very healthy business. We expect it to grow. I mean, we're never going to be a defense contractor like -- but as our primary business. But -- could it be 15% of our business in a steady state? Absolutely. And the projects are important. Some of them are existential and it's really fun to be involved in them.

What about -- I know we were talking about demand earlier, right, and the need to drive predictive analytics. But there's been -- I guess, some of these large companies try to go it alone. They'll build homegrown and it fails for 1 reason or another then it has to be re-architected. Are you seeing more organizations turn around and decide they want to go to an outsourced model quicker? Or do you think that this is going to be the status quo where guys will try to tackle it a couple of times on their own before they turn around and say, "Hey, we'd be better off going to another vendor for this?

I think it might be part of right of passage, Mike, that you need to go through these stages of first, you do all these proofs of concept with SaaS and Alteryx, okay? And then you basically try to build it yourself on top of Azure or on top of Microsoft or you do something or you try to do a big bang with what was a project that IBM used to have, Watson where scores billions of dollars spent there. Tens of billions of dollars spent with GE Predix, all of that stuff keep crashing around. So then you decide to build it yourself using any number of open source components on the Google Cloud or these big science projects on the Google Cloud and the Microsoft Cloud on the AWS Cloud. And virtually every 1 of these customers spent between millions to hundreds of millions to billions trying to build this themselves, okay? And I think maybe that's a process you just need to go through, okay, before you're ready for us. And you're sick of it, you're sick of the CIO telling you that he or she can do it. And some of these companies like JPMorgan Chase has a $1 billion project to try to build it themselves. Well, I mean, there's no chance this is going to end well, they've already gone through 3 leaders. They were going to build a big facility in Palo Alto. Real estate in Palo Alto is pretty cheap right now. So this would be a good time to do it. But these things all come crashing down around themselves and then they come to us.

And if we're looking at the different market verticals you have here, like are any of those considered more -- I don't know if mature is the right word, but if they're thinking about like, all right, we've gone down this exercise, let's go over to C3. Like are some of these verticals more further down that path versus others?

Interestingly enough, probably most mature is utilities. Utilities historically are kind of the last to adopt and they were an early adopter in this area because of the big investment they made in censoring the great infrastructure with the smart grid that made them particularly well suited to deploy large-scale sort analytics to deliver safer, cleaner, more reliable energy. And so I'd say the utility market might be the most mature. Oil and gas is one of the most rapidly growing, defense and intelligence is huge. I think it's arguable that we are at war with China as it relates to AI. Right now, there might be a sense amongst a number of people in charge that China is winning that war. If you haven't read this book, AI Superpowers by Kai-Fu Lee, I recommend it. And so there's kind of people in Washington, D.C. are getting very attentive. So we expect that to be a -- and this is existential. If they don't get this right, we all need to learn to speak Mandarin. So I think the largest market is not -- it hasn't really started yet. The largest market will be precision medicine -- but financial services -- in the long run, just like ERP, just like CRM, it's going to be all markets that adopt.

Okay. Okay. And I know if we think about the components of the business, the AI Suite itself, the prebuilt applications offered, what's the typical progression in which companies are adopting and deploying these offerings. And I'm curious what your long-term view is of how the revenue mix plays out between those 2 pillars?

The typical progression is that -- we didn't use to have applications that we used to have some years ago it was the platform. So that was the bulk of our revenue. So people would do a trial, build an application and then license the platform. Then we built out these applications [indiscernible], people will do a trial and say, supply network risk, which is a very, very important issue in a lot of industries today. [indiscernible] and supply chain fraud, customer churn, whatever demand forecasting, whatever it might be and then they'll license 2 or 3 applications from us, and then they'll decide they want to deploy an organization of 100 or 200 or 300 people and build these applications themselves like Shell is doing, like Coke Industries is doing like Baker Hughes is doing, like LyondellBasell is doing. So that's the typical progression. I -- rough numbers, and this is just a swag. I would speculate, and I'm doing this analysis right now. I would expect 50% of our bookings today are from applications. I would expect to see applications to grow as a percent of our bookings prospectively. But the typical journey is, first, to do a pilot, you demonstrate the economic benefit, you deploy 1, 2, 3 or 4 applications and then you license the platform to build any kind of predictive applications that you might would like.

Okay. And the time -- I know it's going to be different customer by customer, but is there a way to put like a more generic time frame around it between when I first adopt the platform, start using some of those prebuilt applications versus like how long before I start building my own applications on the platform, like what Shell was saying in the example you provided?

Well, Shell started out evaluating predictive maintenance and then [ left ] for the platform right away, okay? So and it looks -- I don't know if this is [ then what, folks ]. So now they licensed both applications and the platform. Same with LyondellBasell, same with Baker Hughes, same with Coke. So it's -- when we have companies that are looking to -- for kind of complete digital transformation and want to be ready to deal with, I mean, how important is supply chain risk in this economy, it's everything, right? Or [indiscernible] optimization of supply chain, it's everything. Who can deliver on time [ in full ] today, I mean it's actually who, whether you buy a cabinet or a car. I mean, it's backordered, right? You can't get. And so I don't think these people are aware of this. This is the kind of technology that you need to use to address this problem. So I think that these kind of market dislocations that we're seeing accelerate in the next few years offer tailwinds for us.

Okay. And maybe 2 or 3 more topics in the final 5 minutes. The first, I know in the current macro environment, we've heard a lot about wage inflation. We've heard a lot about the maybe difficulty in acquiring talent. Can you talk to both of those topics from C3 side of the house? What kind of wage inflation you are seeing? The difficulty or non-difficulty, but I guess, in making sure that you're properly staffing this company for the -- to build it for growth?

Well, I think inflation is not specific to wages. I mean, rents are up 30% in many places year-over-year. Gasoline's up, what, roughly 100%. Protein up 23%. So we're seeing inflation period. The -- and have we seen upward pressure in compensation in the last 2 years, absolutely. Now some of this, I think, is mitigated by the following night that everybody gets to watch on their Bloomberg screen every day. We're now -- we have -- where we have the people who were setting the bar on -- who were creating the wage inflation, now hiring [ ceases ]. That would be Facebook, that would be Amazon, that would be Uber. That would be Twitter. And those were the primary movers in the space, now hiring [ ceases ] and Netflix is doing layoffs. They think the market is going through a new case on that. As it was access to capital, we employ rough number 700 to 750 people, Mike, and last year, we had 52,000 job applicants with their [indiscernible] or deep experience in the geophysical sciences or deep experience in aerospace. So we're -- I believe that last year, we screened 52,000 job applicants, interviewed 8,000 and hired 300. So where the C3 happens to be placed a very attractive place where people want to work. And so we've been seeing to be -- continue to do a pipeline of very, very well-trained, experienced and high-performance personnel who want to work here. So we're in very, very good shape. And I think that we're about to see a secular change in the capital markets as sources of capital dry up for companies that run out of cash, okay, and large companies continue to move from higher increases to layoffs. And then you've seen this movie before. I've seen this movie 7 times before. And I think that this one looks like a [indiscernible]. But net-net, let's look at the business. We have $0.25 billion business. We're addressing what the market suggests is a $600 billion market. Divide by anything you want, Okay, divide by 6, okay? It's still bigger than a [indiscernible] box. The company is growing at a 40% compound annual growth rate. We have a rock-solid balance sheet with roughly $1 billion of cash in the bank. And I think the company is trading at enterprise value of about 2. And then it looks like a valued stock to me, Mike.

That probably leads me to the -- probably the last question we can squeeze in before this is over. But there was 1 more question coming from one of the listeners. Just on capital allocation. And then can you weave that in with the share buyback authorization and how we should think about that?

We will be doing a conference call on third June, okay, where we will address this issue, okay, in some detail. And I mean we have a clear plan. I realized that the sentiment of the market has changed when it was not too long ago that I would have all the 24-year-old Vice Presidents from Goldman Sachs in my office, explaining that we needed to spend more. We needed to spend more. We needed to spend more and then you had [indiscernible] doing whatever the [indiscernible] was doing when we work and all these other companies. And we never listened to that. We did raise $1 billion in IPO to invest in branding and market leadership. I think we're doing a really good job at that. That being said, it's clear that in today's market, okay, the market is demanding that companies have a plan to run a cash positive business. And you'll find that when we do our conference call, whenever it is the beginning of June, that we'll lay out that plan in some detail to show how we are going to be operating a cash positive business. And so we know how to do it. We've done it before and we're going to do it now.

Great. Great. And I think we'll have to leave it there, but thank you very much to you. And I know Juho is back there, so please thank him as well.

Hi, Juho.

And thank you to all our listeners. Thanks again for joining us at the Needham conference. Any other questions, please feel free to reach out.

Thanks, Mike.

Take care.