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

Hi, everyone. I'm Joe Vruwink. I cover vertical software at Baird. Our next presentation comes from Autodesk, a leader in design software used by architects, manufacturers, engineers and digital media creators. Joining me on stage, Steve Blum, he is Executive Vice President and COO; Simon Mays-Smith, he leads Investor Relations. This is going to be a fireside chat format. If you have a question, you can e-mail Session 2 at RW Baird and I'll get that on the iPad. But maybe just to begin, I'll ask for an intro to Autodesk and overview of the current investment case.

All right. So thanks for joining us, everybody. We're glad to be here. As Joe said, Autodesk is focused on 3 primary industries, 2 very large ones, architecture, engineering and construction, and we do have that operate side of the business as well and also manufacturing. And we are also a key player in the media and entertainment space. We're a global company. And we are basically enabling designers, creators, artists to design and create a better world. And we're making everything real. We're helping them make everything they can conceive of, become reality and real. And we're an interesting company. We are growing and have some exciting transformations that we're in the middle of or at the tail end of and happy to get into some of the questions and dive into some of the details.

And to add to that, and that drives really the investment phase, which is multiple avenues to grow the business on the top line, self-help on margins, particularly right now on sales and marketing, which Steve can talk about to drive sales and marketing efficiency and margin improvement and deploying capital as our free cash flow stack rebuilds from another business model transition into repurchasing shares and reducing share count. And those 2 things as we come to the end of those 2 transformations, the new transaction model and the annual billing model, means less noise as well. So Autodesk has been quite noisy over the last few years, that noise will diminish as we get through those transitions and then you start seeing the benefits, particularly on the margin side over the coming few years.

I'll pick on Steve for a little bit because Steve is the master of transitions at this point, license to subscription to named users, annual billings. But this one right now is really interesting. Maybe you can just talk about why Autodesk is morphing its transaction model, taking control of what used to be a reseller-driven sales approach and kind of how that's all gone so far?

Yes. So we've gone through numerous transformations and as Simon said, sorry for some of the difficulty in modeling that, that's created for many of you are at the tail end of all of that, we are well on our way on the last significant transformation, which is moving from a 2-tier buy-sell model to basically an agency model, which is we've been calling it the new transaction model. I know some of you are very familiar with what we've done, some of you probably not so familiar. So let me kind of step back and kind of step you through what we're doing. So Autodesk has been around a long time. We've been in business for over 40 years. We're really probably the first software company that started in an indirect model. So most companies as start-ups, they start direct. Today, we wouldn't even have a sales team, they basically start with an online store, leveraging PLG, eventually, you start hiring some salespeople, maybe some success folks and try to figure out somewhere down the road when you're going to add a channel. We started the other way. We actually started selling indirectly. And in fact, for most of our history, we've sold through 2 tiers. We basically would take our list price of software, we would discount it to a certain level, we would sell to a distribution set of partners, so with distributors, who would then mark up that product and resell it to solution providers who were engaging with our end customers. So they would then actually mark the product up again and resell it to our end customers. And that's the model we've been in for a very, very long time. While that model works, and I'm not saying that it's bad to be in that model, there are trade-offs. And there are numerous that we decided weren't worth continuing. First and foremost, we didn't have great visibility to who our customers were. Going through 2 tiers of distribution, you lose fidelity in your customer information. And you also don't have the ability to track who they are, how they're entitling the software, who's using what, so that you actually have the data to be able to make recommendations on how they can be most effective and productive and start even using AI to give recommendations on what else they can be using to drive expansion. So we wanted to address that issue. It also meant that we were outsourcing the customer experience to our partners. And we decided that to deliver a world-class experience to our customers, we needed to own that experience. We no longer wanted to outsource that. So by changing from a 2-tier distribution model to a direct transaction model, but retaining our solution providers, which is really an important piece, in a different operating model, so there are local super power, so to speak, to work with customers and helping them drive adoption and implementation, that would be the best approach. So we began this journey of changing the transaction model back in, well, almost a little over a year ago. We started in Australia, just like with our transition from selling perpetual licenses with maintenance to subscription and cloud-based offerings. We started there in Australia. We began this as well. While Australia, it's a Western country, but it's kind of small, it's not that material and it's an opportunity to learn and grow. So we rolled out an MVP of this new offering. And the reason why I refer to it as an MVP is we have been remodeling and redoing our technology stack to run the business for the last 6 years. We started first with our financial systems and then went to all of our customer-facing go-to-market systems. And the new transaction model is now built on top of this modern SaaS-based tech stack. So we rolled it out first in Australia. We worked out some of the kinks. We added more capabilities. We moved on to North America, which went live middle of June last year. So we're approaching the first year anniversary of the North American market in another 1.5 weeks or so. We went then to Europe and the U.K. in September, Japan in November. And then the last piece was when we moved our construction offerings in North America to the new transaction model this past quarter. So that was the last piece. So we're now done in all of our primary markets. So the rollout is complete. It has gone really well. The technology has worked, and it's now even more capable because as a SaaS-based offering, we continue to add more value. And every time we put more features and capabilities in, it goes live everywhere. If you were paying attention to our fourth quarter earnings call, we did talk about a bit of disruption and friction in the system. And that was associated with the fact that as we move to rolling out the new transaction model in Europe, in the U.K., we went -- it was the first time we actually went to a region that had multiple countries. It had multiple currencies, multiple languages, multiple tax regulations and there were many more use cases than we had anticipated there and we needed to build more automation in, which we've been working on actively and are continuing to do, but a lot of that friction has come out of the system. We also had some noncontracted partners that were part of selling to our small, small customers that we had no visibility in and we kind of lost that connection to those customers, but we're attracting them back now through our Autodesk store, which is one of the reasons why in our Q1 results, we talked about really strong sales there. So, this is the last of the really big business model transformations. It's been years in the making, and we've kind of used each one as a stepping stone to -- to getting to this point where we now control the customer experience. We know who our customers are, we have great fidelity in the data, we're connecting better with them, we have much more self-service, automation and digitization that our customers really like and it actually will drive efficiencies in our go-to-market.

Can you talk about some of the financial consequences that result from this change? Like on the most recent earnings call you mentioned the online store, online store can be a very profitable channel, for instance, if more business moves that way. Just some of the things that now owning the -- pricing would be another one, but what should investors kind of pencil out over the coming years?

So first, there's mechanical changes. So if you think back to the old model where I said we were discounting our list price and then the next 2 tiers of distribution and partners we're adding it on, we are now setting the price. In a buy-sell model, a software company or any technology company is not allowed to set the price, can't even talk about the end customer price. In this model, we actually are now setting the price. And in fact, all quotes are coming out of our system. So our partners actually put the quotes into our system when they're talking with end customers and then we send them direct to the customer and then they transact directly with us. So, we're now actually recognizing revenue at our list price. So there's a mechanical increase to revenue at the top line. We're no longer paying or measuring -- compensating our partners through contra revenue through that discount mechanism that didn't show up on the P&L before. We're now actually paying them through OpEx. So actually, the cost of the commissions move on to our P&L. So, the mechanical changes of this before we get into some of these benefits are that we have top line growth, we have basically -- if we keep the commission system the same, our operating dollars stay the same, but operating margin takes a hit because we're putting more revenue on, but we're also putting more spend on and so it has an operating margin impact. Now over time, we will adjust and change the commissions, and I'll get a little bit to that. We're -- we expect over time this will be revenue positive, it will be operating margin positive, it will be operating dollar positive. So now back to the online store, as an example, our Autodesk stores. Those noncontracted partners, I mentioned that we kind of lost through this process and now we're getting the customers back through our Autodesk store. In the old model, we would still discount our products to distribution, who would then mark it up and sell it to that noncontracted partner, who would actually then sell it to the end customer. So we were recognizing revenue that was basically the commission across that board. Now when those customers come through the Autodesk store and they purchase like they did in the first quarter, they're buying at our list price. And we actually pay no commissions to anybody because that is fully self-service. So that ends up being a benefit to revenue and higher margins as well. And as we build out more self-service capabilities, and as we now move into the renewal cycles of folks that have gone through -- customers have gone into the new transaction model and more and more of those renewals are being done through self-service and automation, we'll continue to have that benefit that the cost goes down. Our costs for people purposes go down, our commissions to our partners go down in those areas. So we will end up netting more revenue and more margin as a result.

You said this was the last great transition, but there is one more thing I can think of that needs to happen, and that's desktop software going to cloud. Is that something that Autodesk doesn't need to force and that's why it's not necessarily a transition like these others? And what's the status of that conversion process?

Yes. So what I was referring to business model changes, systematic changes, this is the last one -- the new transaction model was the last one of those. But we do have a significant transformation underway, and it's a multiyear transformation, but it's not something that's going to be complicated, not going to impact models and things like that. And I would frame it a little bit differently from -- we're going to move from selling discrete individual products to actually selling industry clouds on top of Autodesk platform services. So we have 3 industry clouds focused on architecture engineering and construction. That's called Forma, the manufacturing industry, which is called Fusion and the media and entertainment space, which is called Flow, so 3 Fs. And they all sit on top of Autodesk Platform Services. And we see significant benefits to our customers to provide them with connected data across workflows that not only that span an industry vertical, but can actually go across industry verticals, because many of our customers actually participate in multiple different industries. They may be working in 1 industry, they may be a manufacturer of this desk, but that desk is going to actually go into a building and they want to be able to collaborate with the architecture, engineering and construction firms and pass that data along. So connected data and managing that data is really important. Now moving our customers to industry clouds is not -- it is a transformation for sure, but it's not a business model transformation or a system transformation, it's basically an additive selling process, where customers are either accessing access to our industry clouds through purchasing one of our collections. So as an example, Forma is included in the AEC Collection or Fusions included in our D&M Collection or they can just buy access to the industry cloud and then start working in that way. Many of our construction customers do that through Autodesk Construction Cloud. So, we see this as a growth opportunity. It's a TAM expansion opportunity. We actually believe that as a result of building out a platform that has 3 industry clouds will -- we'll be able to solve more of our customers' problems because of the access to data, which will bring more users and more TAM into the ecosystem than what we're currently serving as a discrete product provider. So there's upside in the model from doing that and more value we can add to our customers.

Just to add to that. When you're going from on-prem to the cloud, there's sort of 2 things you have to watch out. One is the transition from perpetual license and maintenance to subscription, so higher price points to low price point. So we've already done that transition. And then the second one is essentially the unbundling of the technology stack from a -- the entire technology stack to a sort of a segmented technology stack. So what I mean by that is, so in our world, I think Fusion and extension. So Fusion plus extension is essentially the equivalent of the full on-prem technology stack. So the challenge when you're transitioning to the cloud is that you're essentially risk turning apples-for-apples business model is turning dollars for a full technology stack into sense from a lightweight technology stack. But we've already got that problem solved because we have the model of the lightweight Fusion, Forma and Flow and then the vertical extension to make sure that we capture the dollar value as we transition. And then as Steve said, once you have the data in the cloud, you can then start expanding your offering into new use cases, into new users, into new workflows to grow your addressable market significantly versus the on-prem model.

Yes.

This is a hard question to answer, but how many customers use Autodesk the way you just described as Autodesk the platform company as opposed to Autodesk, the provider of AutoCAD and Revit and Inventor? And I say this because I think you've done a lot of progressive almost behind the scenes work on things like platform services, getting Autodesk docks ready, all the things that can build a much bigger business, but then we ultimately see customers that just think about the -- as you say, the discrete product they're using. So where does that stand in terms of, hey, we know it's a tiny portion, but here are customers that are starting to do it, do most customers still have to kind of go through that evolution?

This is a journey that will take time. Our largest customers actually see us as a provider of platform capabilities, but they also see us as a provider of platform capabilities that have platform solutions and also high-value desktop solutions. And that's okay. The desktop world is not going to just immediately go away. It may not go away. We may -- we'll see a hybrid model in this case. But there'll be expansion and growth that will come through the industry cloud. So there will also be -- I mean, for Autodesk, we have 1 million customers. So we have everything from the largest companies in the world to any one of you could actually be a customer of ours because you're doing landscape design and you actually are using one of our products for that. Those smaller companies don't have a need for all the capabilities and values of the platform. So there'll be critical products that they'll still use, and that's just fine, and we're happy to service them. But our largest customers and really even the mid-market, so a large number of really big companies are going to see significant value and they're already asking us for it. I mean they're already using things like Autodesk Construction Cloud as their platform for all their construction work. They are pushing us to add as much value as we can as fast as we can into Forma. By the way, for all of you -- for any of you that are going to participate either in person or perhaps virtually for Autodesk University, this going to be a lot we're going to be talking about the industry cloud and AI and things like that. So this will be a journey that will have really much a long runway of growth for us. And as we continue to add more value, more will move to the platform-based approach. And to Simon's point, our goal is not to replace a discrete product with something else. It's more of to start taking the key capabilities that would benefit from being in the cloud and being a part of the platform and moving them in as extensions or other things that could be monetized.

Want to ask about AI to may be connected into the topics we just covered. So platform services as an API framework, how does that benefit the AI strategy? Maybe different than what competitors have in place. Same with the sales approach, how does owning the channel and now directly engaging with customers and seeing how they're using your software? I mean, you have real-time insight on this now, how does that feed into the AI strategy?

Yes. So I would say -- I'm going to answer it in 2 different ways. The -- our API access. So we have taken an API first approach, which has been a big change for Autodesk, because that was not the way we were with perpetual licenses and things like that. So our -- that is enabling us more to support the platform move, because API access is how you can actually access the data that makes it extensible and connectable and actually can start driving other business systems and processes. API access can feed AI activities if you're going to try to do AI activities on your own. From an AI perspective, we're investing heavily in AI development. We've actually been working on AI and really machine learning first and then AI for 10 years now. And so our developments in AI are quite extensive at this point. And we really have 2 different kinds of AI. There's more of the productivity beneficial AI. These are like kind of agents that will do specific tasks that are repeatable, and -- redundant, scalable, maybe not innovative or high value, but you got to get the work done. And during our earnings call, we talked about one of them called AutoConstrain. It's a process that has to be get done by engineers who are manufacturing products. It's time-consuming. It's arduous. It doesn't require a whole lot of engineering innovation, but it has to get done. And -- so we have built an agent -- AI agent basically to do AutoConstrain. And we're seeing -- when it gives recommendations on it, we're now seeing customers accept it over 50% of the time, which is fantastic. So we're going to be finding ways to automate those kind of tasks and making -- giving our customers beneficial features that produce productivity enhancements and benefits. We are also working on some of the harder more innovative work for AI, where we are building foundation models. And this is where I'm going to tie back into the question that you asked, Joe. In order to build models for AI, you have to have access to data. You can be the most creative people to build a model, but if you don't have the data to train the model, you're not going to get the outcome and results you want from it. Because we were moving to the cloud far before anyone else, there's more information our customers have created in the cloud in the Autodesk environment than anywhere else. We have more than any one of our individual customers are going to have and we have more than any of our competitors app. So -- and know that having data immediately available, doesn't just make it usable to train a model. There's a lot of work that has to be done to be able to aggregate it and position it and prepare it so that it can actually be used in a trainable fashion. So we have access to a lot of data, and we are working through how do we train models that can do some highly innovative things. And this also is then when we start getting into conversations with our customers about how we're using data, how we're using some of their data to train models and there's a whole element that we have to work through on making sure that they're comfortable in trusting us in that process. But having that access to data in the cloud is really what's enabling us to be innovative in the AI space.

Autodesk does this great annual survey of customers, the state of design and make. And I thought it was interesting this year when you ask customers about how they're feeling about it actually kind of moved in the negative direction year-on-year, feeling like it's not ready or it's not going to benefit the industry, are just not the right products for the engineering community available yet? So are we just seeing kind of the lack of product fit and you're trying to address it? .

Not quite. So it's -- so this is -- we've released our third addition of the State and Design and make report this year. Last year was the first time we had anything about AI. Because 2 years ago, people weren't even talking about AI. So -- and I would say the compare and contrast was when we did the report a year ago, there was exuberance in the market about everything is going to be great, and we're going to use AI to do all these things, but no one really had any measurable, quantifiable thing. So the report showed a really, really high expectation level of what could be done. A year has gone by and has settled in a bit more about, okay, there's real value in AI, and it's going to be game changing, but it's also not easy, it's hard. We don't know exactly what it means to our business. And does it put me at risk with my data and my IP. So we did see the results go down where there's anticipation of greater disruption in industries than what was anticipated before. There's more customers, there was a growth in a number of our customers -- or users. It wasn't this customers, we evaluate thousands or talk to thousands of people around the world. The level of confidence that goals were being met through AI technologies dropped precipitously. But again, I wondered how it could be so high the year before, because there wasn't that much that was there, but everyone was highly exuberant. But what did come out in the report is that 2/3 of all people who are customers who were for surveyed said they absolutely believe that their industries will be disrupted in some way by AI. Most of them still believe that AI will be a benefit, although that number went down because some are trying to figure out how they play with their data and things like that. And they actually are highlighting some interesting areas where they think AI can actually play a really big role. And the single biggest play that they identified this year as a use case for AI is helping them meet sustainability goals and things. And so anyway, some really interesting outcomes. I'm really curious to see where it is next year after another 12 months.

Yes. The challenge our customer has that they don't have enough data themselves to build their own models and capabilities. And that's a function of not having enough data but also they don't have enough data in the right format. It's -- the tech environment is complicated, should we say. And so sucking that data in, in a usable way is hard. . And also, they don't have enough of the right data. Not all data is equal, some data is very useful for building models, some is useless. And so really, it's finding partners like us who can help them extract value from their data by putting it -- making it usable and then also helping them figure out what data is valuable and what isn't and then to generate value on top of that data. So that's really what we're here and to have those conversations with them.

Yes, the feedback from customers and the like coming out of Autodesk University and of course, there's start-ups focused on this, the text to CAD companies. But generally, I find that customers think about Autodesk as needing to be the company to bring it just because of your scale? No one else can kind of bring it all together at scale like you can.

Exactly. And that's -- and those are the conversations we're having with customers. There may be some areas where they want to do some AI innovation themselves, but in areas that we basically have the opportunity to be able to do it for them and say, let us do that for you and you go focus your resources in other areas.

Maybe I'll pivot to kind of a financial framework question. So again, an underlying basis, when you do all the normalizing for the accounting changes, Autodesk has kind of framed their business over time as net retention should be 100% to 110%. And in the most recent quarter, it was on the upper end of typically what you talk about. How do you typically grow at an existing account after 40 years in business? Are your customers still growing their seats at a faster pace than maybe the industry at large, just because you want to work at the big E&C firms? Or are there other things you're seeing that drive that type of growth?

Yes. So because we service all markets and all size customers, we kind of see the benefits or the challenges across the entire spectrum. So we're driving expansion. Expansion, by the way, is the single biggest growth lever that we have. So even though we've been around a long time, our growth is coming from selling more to our existing customers. It doesn't mean we don't add new logos, we do, and that's great. but the biggest add is coming through our existing customer base. But it comes in multiple places. First of all, we've introduced some new business offerings. Construction it was introduced about 6 years ago. It's continuing to grow. That's opening up new TAM. There's some new customers, but it's driving a lot of expansion in existing customers. Same thing with our water solutions. We were already working with customers that were in that space, but we didn't have water solution capabilities. So we have some new offerings that drive expansion. We continue to bring out higher value offerings. So -- for those of you that are familiar with the term building information modeling, it was -- it's a model-based or 3D way of doing architectural and engineering design. Well, that's been replacing 2D design that's been done with AutoCAD with a product called Revit, and that is all kinds of collaboration capabilities. So moving customers from those lower-value 2D offerings to 3D offerings just add more NRR because we're getting more per subscription. And then as much as we've been at it a long time, some of our customers are really large, and they have multiple divisions and there's still lots of white space within those accounts. And so our opportunity to go find those places where they're not using Autodesk software and help them adopt using our offerings just gives us the incremental growth opportunities as well.

What about on a margin front. There's examples in enterprise software. When you get monopolies or oligopolies 50% EBIT margins are seen. Autodesk on a normalized basis, I suppose, closer to 40% in a lot of ways. You'll accrue some benefit by doing this transition, maybe beyond that accrual. How do you think about needing to grow investment to support the growth ambitions of the company? Is this the type of thing where if you're growing 10%, the OpEx grows 5% just because at your size, it's hard to grow such big numbers you're already investing far more than most are in these types of problems?

Yes. So sort of a few dimensions of answering that. The first one, obviously, we're not a monopoly. So that's the sort of first answer. So we have some good large and small competitors, which we enjoy. Generally, we get the better of them, but we do have lots of competitors. The second thing is that we have different businesses. So for example, Cadence and Synopsys are the only 2 players in the industry they have about 1,000 customers each, whereas we've got more than 1 million customers. So you just have different intensities of go-to-market effort relative to the end customer. Similarly, we have much more of a technical sale than, let's say, Adobe. So that will always mean we have more sort of handholding in the sales process and sort of proportionally less self-service. So that has an impact on the sort of end margins as well. But -- and then like all companies, we have to make the same investments in things like AI, also as you put more of the business to the cloud, you have gross margin pressure, you move more of the business to the cloud. So -- so those are all reasons to sort of think about as you're thinking about margins. But to your question, there's also a bunch of stuff that we're doing, as you know, to improve margins, one of which is the work that Steve's been doing on the new transaction model, but it's not the only thing, it's an evergreen process of making the business more efficient. Efficiency is not a one-off thing. It's something you're doing day in, day out, every week, every day, every week, every quarter, every year, becoming more efficient and then that frees up resources to invest in your strategic priorities and also drive margin improvement. So if you look up and down the sort of P&L, looking at the sort of gross margin side, it's around how you manage the efficiency of your cloud usage. So for example, if you look at Amazon's first AWS AI announcement, they use 3 case studies, one of which was us using AI to manage cloud compute more efficiently within AWS. There's also where you do the compute on the desktop or in the cloud. We talked about sales and marketing efficiency. On the R&D side, we've talked about Autodesk platform services, reducing our technical debt as sort of Intuit type process, which we've been working on and also around data accessibility because if you try and extract data from files where the vast majority of data is stored, you're going to blow a hole in your margins to do scale. So again, making your data accessible in the cloud is becoming more important, there's a bunch of work going on in that. And then on the G&A side, we've been completely rebuilding over the last 7 years, our entire financial systems to support all the work we've been doing the new transaction model and a bunch of other stuff, which is one of the reasons why G&A is running a little hot at the moment. But as we scale, we'll get efficiencies from that. So it's really that sort of put and take of investing in the business for the future to make sure you don't become obsolete, but also at the same time, making sure you're driving growth in the margins and the flow through to margins over time. And we got sort of opportunity to do that. .

Great. With that, we're out of time, but please join in thanking Autodesk.