Autodesk, Inc. (ADSK) Earnings Call Transcript & Summary

I'm Joe Vruwink from the Vertical Software team. Our next presentation comes from Autodesk. Autodesk is one of the names when it comes to design software for architects, manufacturers, engineers and digital media creators. Joining us from Autodesk today are Srinath and Simon. Srinath is Vice President of Business Strategy and Marketing; Simon is Vice President of Investor Relations. I will turn it over to both of you for an introduction to the company, and then we can handle Q&A.

Great. Thanks, Joe. I'm going to start by just reading Safe Harbor statement. We may make forward-looking statements during the course of this presentation. Please refer to our SEC filings for information on risks and other factors that may cause our actual results to differ. So good morning. Welcome, everyone. Just going to spend a couple of minutes talking about who we are and what we do. We are sort of one of the largest design and make software companies in the world. And what we're really trying to do is a very simple thing, which is to connect workflows, which historically have been siloed and have run concurrently to each other. We're trying to connect them together in the cloud to enable them to run simultaneously to each other to allow our customers to become more efficient and more sustainable. And we're doing that, really, across multiple dimensions. One is sort of vertically, which is to add more functionality, so you move from 2D to 3D to sort of building information modeling, and we can maybe talk a bit about that. But also crucially, and this is where we have a significant source of competitive advantage, is connecting workflows sort of horizontally in the cloud. So from an established position over time, over the last 40 years in design, in architecture, engineering and construction, we're moving from design down through pre-construction into construction, and then more recently, through digital twins into the operations and maintenance part of the market. And we're also moving upstream into the sort of planning phase of the market. And as we do that, we are adding new users into our ecosystem and new addressable market opportunity. But also we're bringing more value to our existing market opportunity by connecting those workflows into our existing users. Similarly, in manufacturing, and I'm going to pass it over to Srinath in a second to talk about that in a bit more detail, we're doing the same thing across design, make and, more recently, into the shop floor. And then in media and entertainment, we're kind of going the other way. We have an established position in post-production with 3ds Max and Maya, and we're moving upstream into the production side of movies and TV with recent acquisitions such as Moxion in the Loop. So as I said, as we do that, larger addressable market opportunity, connecting workflows to enable our customers to become more efficient and more sustainable. Srinath, do you want to talk about manufacturing just briefly?

Yes. So the first and foremost thing, I'll build on what Simon said, we help our customers make things. Our customers use our software. Obviously, our software is very visual and interactive, and it produces a beautiful pixels on the screen. But our job is not about painting pixels in the screen. It's about helping our customers take those pixels and get them to a point where they can actually realize the full potential of what they're envisioning into their life. And our software that you see here on the screen is right now about bringing the design and manufacturing disciplines together with backbone, so that we can address the end-to-end needs of our customers with collaboration at the center. And we are doing this with an open ecosystem in mind, so that we can continue to reach other parts of the market that we don't have access to right now with our own software, and that's where partnerships and connection to our assisted industries like AEC and M&A is going to be important. And one of the things we're also doing here is we are lowering the value for access both in terms of user experience, in terms of business models, and that's why we have new ways in which we are able to bring software to market with subscription as well as extensions. Extensions is there to add additional functionality, targeting specialized people or specialized personas on an as-need basis. And this is very much a new concept for the industry, and we're finding a lot of pickup in this because they are now unlocking portions of the market that were overserved.

Great. Joe, over to you.

So -- and I should say, anyone in the audience has questions, you can e-mail [email protected], and I'll get those on the tablet. This is maybe a bit of a key takeaway of everything -- [Audio Gap]

A bit about this on our Q4 call as well, which is -- it's a process of constant evolution. If you stand still, you not only missed opportunity, but you risk somebody else coming in and taking that new opportunity. So it's really the sort of shift from sort of 2D to 3D. It's a shift then on to sort of building information modeling, and all of those things incrementally add value to your customers, which we can then sort of charge more for. It's then also the shift sort of horizontally, so from design through make, through operations and maintenance, et cetera, as I was sort of talking about. It's then also -- as Srinath was talking about, it's enabling more people to come into our ecosystem. So users which haven't been traditional users, enabling them to come into your ecosystem. So partly, that's about -- Srinath was talking about disaggregating the platform from the vertical functionality. It's partly about shifting business model from perpetual license and maintenance to subscription. It's partly about enabling consumption models so the people who don't use the product full time but still need to use it occasionally can consume it. And then over time, as we shift from data in the cloud and we start applying AI and machine learning to that data, we will start being able to develop sort of usage-based models sort of Akin to sort of AWS and Azure, which you see today, those types of usage-based models. All of those transitions have been accretive for our customers because it has added value to their workflows and also enabled more people to benefit from those workflows. And also, it's been accretive to us because it's significantly expanded our market opportunity.

A bit of a turn event on the macro. Can you -- well, first of all, outside looking at what would you be paying attention to gauge the health of your customer? And then as a follow on, how does the health of your customer ultimately influence spending on Autodesk? And I guess, last fall, there were some new macro things that on the margin hurts customer spending, and then this year, the surprise from whenever the earnings report was as -- it seems like things are actually surprisingly resilient. So what would you be paying attention to?

Yes, so lots to unpack there. I mean, the key sort of point I think in terms of resilience is that the -- since phase of 2000s, we've gone through a business model transition from perpetual license and maintenance to subscription, and you saw the effect of that during the pandemic, which is -- whereas in the 2000s, we had multiple quarters of revenues down 30%. During the pandemic, we carried on growing double-digit comfortably throughout the pandemic. So just a very different world we're in, and the reason for that is that our customers are typically living their professional lives through our software. So if they are in business, they're typically subscribing. They have their workflow stored in our software and multiple projects over time, which they're building over time. So it tends to be a sort of much more sort of resilient business model. What we did flag, and I think we were one of the first companies to flag it last year, is that demand coming out of the pandemic, as you would expect, has remained very robust, and we repeated that last quarter. But what's been an issue is supply and really a sort of combination of factors, which has been supply chain disruption, the ebb and flow of COVID and labor shortages have really had 2 effects on everybody, really, which is it makes it harder to finish stuff and it makes harder to start new stuff. It's just sort of grit in the work. And it's a combination of those factors and different mix in different countries and the mix changes over time as stuff moves around, but really just makes it hard to get stuff done. So demand has been relatively robust, and it's the supply that's been harder. So in terms of our business, we have this big beast, which is the subscription renewal business, which tends to be very resilient, as I said. Where we would see and did see during the pandemic cyclical exposure is really, firstly, on the rate of new subscriber growth. It, not surprisingly, slows during tougher economic times and it gets faster during better economic times, and we'd expect the same thing to happen over the cycle -- in future cycles. The second area is in the sort of conveyor belt from entry-level products to premium products. That conveyor belt slows during tougher times and accelerates during better economic times. And then the third area is around multi-year build up front. If customers are trying to conserve cash, they will sign annual deals. Whereas if they're more confident, they will proportionally more of them will sign 3-year deals. That's not all bad news. For those of you who pay close attention to us, we're going through a process of essentially transitioning all of our customers to annual billing next year. So to the extent that they decide to do it early this year, would mean that actually next year is less bad than perhaps some people fear. So there's a silver lining to that cloud, if that does happen this year.

So let's focus on manufacturing. We have a manufacturing expert with us today. Maybe I'll -- a bit of an open-ended question, but competitively, I think when a lot of investors think about manufacturing, exposed design software, we tend to think more about Dessau, Siemens, maybe PTC. Of course, Autodesk as inventor, which I think is maybe the biggest mechanical CAD just stand-alone brand or one of the biggest. But how would you answer competitive differentiation between you and those other companies?

Yes. I have to start with the backdrop, I mean, the industry backdrop and the sign of today's times. If you look at the pressures our customers are facing, and I would say, in general, the market is facing, the product complexity is increasing. It's no longer sufficient to produce a mechanical gadget. Now, almost every product has mechanical, electrical, software, optics and the list goes on and on. Second thing is the time-to-market pressures are increasing, right? So the product cycles are getting shorter and shorter. And lastly, the consumer expectations increasing tremendously as well. And this is putting a lot of pressure on our customers and the entire market to collaborate more in terms of how they bring their entire organization together and the entire value chain together, and how they think about the product development process. They also have to start thinking about a bigger exploration than what they have done in the past and how they produce innovation. When I say exploration, they have to be exploring more ideas, and they have to do this faster and in more number of iterations. And this combination is where we see the biggest opportunity. And I feel like this is the opportunity that the existing vendors like Siemens and Dessau have not addressed properly. The market today is feeling overserved with monolithic, very specialized applications. And these specialized applications are very good at one single thing, and they're not good at solving these complex challenges. Like I said, bringing multiple disciplines together, solving the collaboration challenges. And what we have done at Autodesk is paying attention to the trends that I just mentioned, invested in cloud infrastructure for over a decade. They have been deliberately investing in this not just for getting presence in the cloud, but for solving these fundamental challenges. They are, first and foremost, brought all the disciplines together in Fusion 360 design and make convergence, as Simon said, to remove all the grid, remove the friction between all these disciplines, having to collaborate to build complex products. And secondly, with cloud, they're able to put data at the center. The data at the center is able to connect the entire organization. And when I say organization, I'm not talking about the people sitting within this virtual 4 walls of a company, I'm talking about the entire value chain. They're able to build the entire value chain around one single source of fruit, which is the data in the center. And lastly, because of our open platform, we are able to bring a lot of partners together, and all this in combination gives us access to new automation, automation that the market has not seen before. Generative Design is a prime example of this kind of cutting-edge, bleeding-edge, next-generation automation that takes the non-value add work out of the system and makes the engineer focus on the things that they're really good at, which is creativity and innovation. So that's I would say, in a nutshell, how we stand on.

This idea of convergence and exposing Autodesk to just a larger addressable audience. Is there a way to numerically quantify that? How many subscribers you might have today relative what the potential is?

Yes. I mean -- so I'll say the answer lies in the trajectory's perfusion itself, and so we uncovered the opportunity for convergence as we are evaluating the big trends, and we've invested in this over the decade. But the results talk for themselves or speak for themselves. If we look at the results as of our last Investor Day, our subscriber base is now 200,000. And it's about 53% 5-year CAGR growth for subscriber base. And it's not just the unit volume growth, right? So the exponential unit volume growth is one thing. Absolutely, we're picking up a lot of share, but monetization too. We are investing in new innovative business models pacifically to lower the value for access and also remove the friction because of all the monolithic applications and how they're over serving the market, and our monetization strategy is also working. Our CAGR for billings growth is 107%, so it's double our subscription growth CAGR. And these are kind of proof points for the validation of our path. We continue to pick up a lot of momentum, and I think that's the extent of what I'll say right now.

And just to add to that, looking sort of further out, and just to touch on what I've -- give you some example of what I was talking about earlier is that -- so all of you folks, when you go back to your office and if you want to print out the notes that you've made on your laptop, you're kind of hit Print and a piece of paper will magically appear out of the printer in the corner. What's happening when you do that is that your computer is turning your word processing into a piece of machine code, and then sending it to the printer heads and the print heads are printing it out on a piece of paper milliseconds later. That process in manufacturing takes 3 days from the time in the model where you hit Print, it sort of -- it takes 3 days to kind of generate the machine code and send it to the printer. So that's sort of one area where we're sort of focused on this to sort of automate that process. But then once you do that and you then connect the design model to the machine, you can then start tracking what the machine is doing and then feeding that information back into the design process. So first point is that we could eventually have machines subscribing to us as well as human subscribing to the software, it's the first point. And then the sort of second point is that the AI and machine learning that we do would then opens up to the next thing then seat-based pricing, which is value-based pricing, which is -- that is then what I was referring to with sort of AWS or Azure-type models, which is sort of usage-based pricing models where you're pricing to value because the value that you're generating from the machine learning is much higher. So that's sort of in the future, but the sort in terms of direction travel. The point is that it's not necessarily human beings, and it's not necessarily about seats.

Yes, that's a valuable point.

I don't know if you'll agree with how I'm about to characterize this. But if I step back and think about the Autodesk TAM expanders, construction has been obviously the biggest one. And there was a lot done over a number of years, both organically and making some sizable acquisitions. And then you use APIs to connect everything together, then you start moving some of the key functions into a common data environment that now, Autodesk Construction Cloud seems like a more mature. It's not all the way there yet, but a lot of work, and you can kind of see the fruits of the labor. Fusion seems to be following that path, but we're kind of in maybe the heightened mode of activity. When I think about adding up chain, there were some tinier acquisitions recently, adding exposure to what happens on the shop floor. Now, Fusion has been around for a long time, so it's not a new product, but how would you characterize kind of the full potential of Fusion? Is it relatable to what Construction Cloud have went to? Or how would you compare contrast, I suppose?

Yes. I mean, there are strong parallels. So let me start with a very value framework or a model for how people describe the digital trends. When you think about digital trends, there is a progression. It's a stepwise progression to the ultimate goal, which is autonomy or autonomous. But progression starts with information trend. Information trend, which actually replicates what's inside the model, meaning, a very descriptive way in which we can describe the model. Like, for example, you have a room like this and this room has doors and access points and everything else. In a BIM model, you can just look at the structure but do not have enough information that says, okay, this thing that I'm looking at is a door or this thing that I'm looking at is a window. So you see, our tandem application has started to tag assets so that you can actually start to do inquiries within a building where you can start to talk about like, this is a door, this is a window, this number of HVAC pipes and tubes and other things are in this building. And then from that information model, you get to a more descriptive model where you're able to not only look at assets, we're able to draw inferences about what these connected assets are able to do. For example, you can actually start to see the entire pipeline for HVAC and see the impact of an HVAC unit in a particular building. And from that, you can actually -- once you get to the descriptive, you get to predictive. The next step is predictive, where you're able to take this information and they're able to start thinking about where can failures happen? How can I predict these failures? How can I get ahead of the failures? And the last step, once you get to predictive, you get to autonomous after that. Autonomous is that the system is able to self-correct, this is what Simon said is very important. The bidirectional loop between machines and humans and activity is going to give the necessary pipeline for autonomy or autonomous. They're on the same journey with Fusion. If you think about where we are right now, we built the convergence pipeline so that we can actually start to go down the same path as what Tandem and the descriptive train that Tandem is building right now. So they have the pipeline. With the pipeline, they're able to capture the full value chain of activities. Once we have the value chain of activities, and as Simon said, they are also starting to extend the value of the activities to the shop floor, they're able to digitize more and more of shop floor, get the analog process into the value chain so that we can get the full descriptive trend. And once they have the full descriptive trend, they can help the manufacturers get more and more predictive value from the digital trend. And eventually, we'll get to a point where these algorithms are now able to help the manufacturers make the systems more and more autonomous. Without any manual intervention, a manufacturer can say, look, all I want to do is match this objective, which is 99.0 six sigma uptime. And then the system is able to cost correct based on whatever is the critical failure points. It's able to put some safety checks in place and cost correct. So this is [indiscernible], yes.

Yes. And what Srinath said is also -- points another important sort of competitive advantages is that everything flows from the design phase down through. So essentially, our historical position in design is really important. And it's important as we enable the workflows to flow down, that's the sort of use in the cloud to do that. But it's also important because the design is really, at its heart, is what it is structured data. And structured data is important is that once you put it into the cloud, you can then start applying AI and machine learning to it both in the design phase, but then also between the different silos as well. So having the historic strength in design and then enabling it to flow down the supply upstream and downstream. And then as you move into the cloud, applying AI machine learning, that's really the pathway ahead for us.

So I'll add one more point here. So the structured data that Simon was talking about, this is another thing that the industry has not addressed very well. Today, the information is locked inside files, and we are deliberately starting to unlock all this data and make it very granular so that we can start to expose this data at the points where it's needed. And why is this important? It actually gives us access to all the critical packs of the ecosystem that we can't touch out of. For example, the machines, and what Simon said earlier, when a person design's part and that part has to be cut on a machine tool like a milling machine, the milling machine has parameters. For example, what speed does it run and what are the cutting parameters, what are the tools, what kind of coolant you want to run on this machine. It's important for the software to know all the parameters of the machine, the cutting parameters and what kind of material it's optimized and cut. If we are able to bring all the OEM and the machine parameters into the software, it removes additional amount of friction. This is where the granular data becomes very critical. By making the data granular and exposing critical pieces of design information that can connect to the machine, we're able to make this bidirectional loop happen.

Simon brought up a good point that everything flows from design, so a lot of what we've been talking about so far is 10% of Autodesk revenues in the make revenue line growing very quickly, but it's not kind of the majority of the business. So if we do look at just the design of revenues of Autodesk and hone in on what's happening in architecture and building engineering, the question we get a lot is that, it maybe doesn't strike you as something that should be fast growing. I guess, the AEC collection at Autodesk has routinely been AutoCAD surpassed the last quarter, but it's routinely been the fastest-growing area. And if you look broadly over the peer universe, I would say all of the AEC's heavy design peers grow faster than their manufacturing counter parts. So of course, the question, is AEC just behind in terms of digital adoption, and we're seeing the catch-up? Or is there something else happening?

So it's a couple of things. One is that you've got some sort of vertical functionality going on. So you've got the move from 2D to 3D and crucially in architecture and engineering, the move to building information modeling. And that has been -- as it's been adopted in regulation across the globe, and it's -- you're getting some multiple waves of opportunity for us. Which is you're getting national adoption as a standard, you're then getting companies typically taking up projects as part of that, using building information modeling. And then over time, what you often see is companies standardizing around building information modeling as the sort of language of exchange internally. And as a market leader in that building information modeling with Revit, we benefited from that. Sort of first thing. And then the sort of second thing is really sort of adding -- as I was talking about earlier, adjacent workflows, is that there's still multiple workflows. So the one we've been talking about is construction, so you've got sort of architects -- and if you think about the life cycle of a building, you start with the year now on the sort of planning phase. You then move to the architecture and engineer, you then move into sort of pre-construction, which is where the architects and engineers and the construction sectors kind of meet. You then move into construction, and then you move into -- back to the operations and maintenance phase. And so what we've been doing is adding an adjacent vertical. So we've been talking a lot about construction, but even in Architecture and Engineering, there's still significant parts of opportunity. So one example, and just sort of a context here, we have about 6 million subscribers at the end of last year. Within Architecture & Engineering, there were about 1 million electrical engineers which is a market where we have relatively limited penetration today and which we're seeking to -- with our partnership with Schneider to develop products for that market. So even within what you would assume was the most mature part of the market, there are still pockets of opportunity, a significant opportunity for us. And then in addition to that, through Spacemaker acquisition, which we made a couple of years ago, we're moving upstream into the planning phase and then tapping into the owners market. And then through Tandem and the digital twin, we're moving through into the operations and maintenance phase, which is, again, the owners are part of the market. The reason the operations and maintenance phase is important is, if you look at the life cycle of a building, about 70% of the cost is post-construction. And so today, we have a significant market opportunity ahead of us. We're only trying to solve 30% of the problem. There's 70% of the opportunity for us to solve, and sort of the operating system to solve that problem will be the digital twin. And we hope that Tandem will be an important contributor to that.

A question from the audience. So we -- in speaking about CAD and CAM, the question is on more CAE for the simulation side of the opportunity, and what you're doing there? Is it an organic approach? Are you leveraging partners? This is the question.

Yes, so I'll start. All right. So this is going to be a combination of organic and inorganic as well as partnerships. Simulation is absolutely critical for us to advance on our pursuit of digital twin, the pathway that I just mentioned. And we do have significant simulation portfolio. If you look at the history of Autodesk, they have acquired significant simulation capabilities from structural, CFD and then also specialized manufacturing capabilities like injection molding simulation and also 3D-printing simulation. We realize that this domain is broad and deep, and there's also a need for us to be very strategic in how we think about our footprint in the simulation space. So we will be seeking out strategic partners, and case in point is Ansys. We partnered with Ansys as we were looking to build a very robust, broader simulation stack targeting the electronic stimulation. And those kind of partnerships will continue to be the way in which we expand our footprint. So they're good on the mechanical side, they're good on the fluid side, they're good on the manufacturing simulation, like I said, for both injection molding and for 3 different things. And there will be plenty of other domains where we will have to seek partnerships, and that's what we'll be doing like with our partnership.

Great. Maybe time for one more question. Free cash flow, what is it going to be in 2024?

Less than 23%.

Maybe we can just talk about -- you've spoken about removing a lot of the cyclicality from your business, cash generation is maybe the final piece, and how that factors into the 3-year upcoming view?

Yes. So -- what's the best way to answer this. So we're going through essentially a business model transition for our cash flow in the same way as we did for our P&L a few years ago. Specifically, we're moving a chunk of our business from multi-year build up front, so 3-year contracts where we take all of the contract value in year 1 and then 0 in year 2, and 0 in year 3, to a model where we signed multi-year contracts, but we get paid annually. So just to take you through the sort of thought process, and forgive me, this starts very simply, but you'll understand why I'm explaining the way I do by the time I get to the end of it. If you have a 3-year contract that's worth $1 a year, if you are paid upfront, you get $3 in year 1, $0 in year 2, $0 in year 3. The way that feeds into our business is you get, in year 1, $1 goes into short-term deferred and $2 goes into long-term deferred. And then in year 2, you have $1 left in long-term deferred because the other $1 has gone into short-term deferred. And then year 3, you've got $0 in long-term deferred and $1 in short-term deferred. So that's the way it sort of bleeds out the stack through our business. So what happens over the next -- this year and next year is in '23, in long-term deferred, you have the $1 remaining from the '22 cohort and you have $2 from the '23 cohort in long-term deferred, so you've got $3 in long-term deferred. In '24, you lose the $22 and you lose $1 or $23. So you go from $3 in long-term deferred to $1 in '24 to $0 in '25. So that creates a headwind, and that's why our cash flow in '24 will be down versus '23. That's the bad news. The good news is that in '24, we will get $1 from the '24 cohort. In '25, we will get the second $1 then from the '24 cohort and the first $1 from the '25 cohort. And then in '26, we'll get the third $1 from the '24 cohort, second $1 from '25 cohort, first $1 from the '26 cohort, and we've rebuilt that subscription stack. We're back to $3 by the time we get to '26. The reason I've gone through that rather long share of your dog story is it means that the '24 cash flow is not a fully loaded cash flow. It's an artificially compressed cash flow. So if you want to look at what a run rate cash flow is, you have to look at '26 and then discount back to what the cash flow is. So that's all I'd encourage you to do. If you're modeling us at least model us to '26, because that will give you what the business will look like once we've gone through this transition.

That's great. We are out of time. Please join me in thanking out.