Autodesk (ADSK) Earnings Call Transcript & Summary

Okay. Hello, everyone, and welcome to Day 2 of the UBS tech conference. It's great to see everyone in the audience. For this session, we have Autodesk. So we have Andrew, who is the CEO; and Janesh CFO. So Andrew, Janesh, thank you so much for being here.

Thank you for hosting us.

Perfect. And before we dive in, in case you have any questions, you can submit them through the app, and I'll try to take a few of them at the end. But with that, Janesh, I know you have a big speech that you'd like to give.

It a very important PSA data. So 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 materially from these statements. And now that we all feel so much better protected back to you.

Well done. Well done. Okay. So maybe to kick things off, you guys just reported last week, strong 3Q results. There was a nice upward provision to the 4Q guide. So maybe you could just talk about the drivers behind that momentum and how we think about some of those demand trends going into next year.

Yes. Maybe I'll just kick us off Taylor. So overall, Q3 was a really good quarter for us. We felt that the momentum from the first half continued very nicely into the third quarter. We saw continued strength in many areas, in data centers and industrial and infrastructure. Those are, again, continuing teams that we had. Overall, the team executed well. We saw great strength not only in AEC, but if I look at it from a geographic perspective, we saw strength in different parts of the world, particularly in a lot of emerging countries. Also based it out as a highlight. So the quarter played out really nicely for us. Our go-to-market execution has been -- and the optimization of the go-to-market has been progressing nicely as well. And all of that allowed us to then have a nice healthy beat across the board in both revenue as well as the profitability measures. And that allowed us to raise guidance across the board for Q4 and the rest of the year as well. So, so far, I think that's played out really nicely. The important takeaways for us really around that consistency of growth and the diversification that we've built in the business over the last several years -- and the go-to-market transition that we've made has -- that's all helped us build a business that we think is very resilient for the future. And so we're excited about next year. The go-to-market optimization is ongoing, and that's something that we will naturally consider when we think about fiscal '27 guidance when we get to that point as well as the macro. But overall, we felt pretty good about Q3 and how the rest of the year is lining up.

Perfect. Yes, let's talk about that resiliency because I know the big question that you guys got from investors is around the potential growth durability. So Andrew, I know that there's been numerous transitions over these last several years. So in terms of where Autodesk stands today and whether or not they're positioned maybe better than they have been in the past seize a lot of these emerging growth opportunities. Can you comment a little bit about that? And on the durability question, I know there's a lot of questions around there's been products. They've been in the market for a while. Is there still sufficient runway left? What are your thoughts there?

Yes. Yes. So we have been through a lot of transitions. All of these are intended to modernize our business. We've been very intentional about creating a face to the market for Autodesk that looks like a modern SaaS AI provider. We've successfully done that. We're ahead on executing consumption algorithms. We've changed the way -- the relationships we have with our partners. All of these things are fundamental things that allow us to kind of accelerate the business into the future. When you look at the business moving forward, one of the things that Janesh highlighted was the consistency, right? And there is kind of a strong core consistency in our business. Fundamentally, our business is highly capacity constrained, our customers' business, in particular, right? There's not enough existing capacity to fulfill all the demand for what needs to be built or rebuilt in the world, both in AEC and manufacturing. So that capacity problem is one of the things that we're primarily focused on over the long term. And it's the idea of allowing our customers to execute on more projects with fewer people for project and actually bringing per project cost down, but increasing the portfolio of projects getting executed. Now if you look at the durability of our revenue that core basis of capacity, that's not going anywhere. That goes back to the consistency in our business. That core base isn't going anywhere. Our customers need to have that capacity to execute on what's out there today and in the future. What they need is incremental capacity. And that incremental capacity is going to be added -- everybody says agents, I say digital humans, I don't know which one is better. But we're going to be adding incremental capacity through automation. And that capacity is going to come along several vectors. One, let's just look at construction right now. Construction still has a long tail of execution here. We've got personas and construction that we've hit. We've hit field execution. We've got preconstruction planning. We've got resource manager. We've got a whole set of personas that we can hit in construction that are going to be durable growth opportunities for us. You look in manufacturing, there's a growing shift to the cloud in manufacturing. Fusion is arguably the fastest-growing application in its space. The durability of that opportunity in the mid-market of manufacturing in Europe and the U.S. is very strong. There's going to be continuing cycling from new applications in that space. And then you look at infrastructure. Infrastructure is not going anywhere. And the backlog and infrastructure projects is huge. So we're going to be rolling out in that area as well. So if you look at what we're doing for our customers, we're rolling out new technology to increase their capacity. It's going to be durable over years to come and just kind of play into the consistency of our business.

Perfect. I want to talk about some of the AI opportunities and infrastructure ones. But before we get there, Janesh, maybe to throw this question over to you because you've talked a lot about the potential for durable growth as we look ahead. But then on the last earnings call, made comments about the macro uncertainty and the potential to continue a guidance philosophy that embeds a conservative approach. So could you maybe talk about when you look into the future and the growth drivers that give you the most comfort in being able to sustain growth, what are those? And secondly, what are you trying to get through to the audience in terms of the messaging around 2027?

Yes. I think the starting point for our confidence for next year is this year's performance. And over the last couple of years as well as this year, we've had just a remarkably consistent growth that we've delivered on an underlying basis when your net LP affected the new transaction model. So that's the starting point for -- as we think about growth for next year. And then at the Investor Day that we had in October, we laid out a number of areas where we continue to see opportunities for us to drive growth. Construction is one. Fusion is another one, infrastructure. And some of these things have been delivering growth for us this year, and we see those as continuing into next year. We talked about adjacencies like operations that you'll get into over time. So all of those things give me confidence in the underlying momentum that we -- and strength that we see in the business. The main messages I wanted to make sure that investors understood coming out of our earnings call, there were twofold. Number one is that we see a consistent growth pattern in the business. But second, just a reminder that our go-to-market optimization is not yet complete. And so when we get to that point in fiscal '27, the approach that we took this year guiding prudently held us in really good stead. We still have a lot to execute on the go-to-market optimization next year. The macro will be uncertain. So we will take a similarly prudent approach when we think about our guidance for next year as we get to fiscal '27. In terms of exactly what the underlying growth is and exactly how we think about the specifics on go-to-market optimization and so forth, we've got work to do on that front until we get to that point, but those are the broad takeaways.

Perfect. And Andrew, let's go back to what you were talking earlier about being able to address capacity constraints with AI. So there's a lot of -- there's a number of exciting AI initiatives underway at Autodesk, whether that be generative design, what you're doing with AI agents, MCP servers. So could you just talk about the current pace of adoption that you're seeing amongst your customers? And when do you think we'll see industry-wide production use of AI.

Yes, you're actually seeing it already. And as we said at our Investor Day, our AI strategy has kind of three stages. We're starting with task automation, which is just basically making day-to-day productivity easy for our customers. It's very protective of our business. It provides a lot of customer delight. Then we're moving to essentially workflow automation, connecting disciplines together, automating the workflows with them. And then basically with systems-level automation, we're taking workflows and automating multiple workflows to provide systems level solution. So progressively, over time, we're going to be delivering more and more value into the ecosystem. If you look at the adoption today, we're super focused on making it easy for the customer to adopt AI. We're not trying to throw tools out there and say, "Hey, isn't this a great thing figuring out how to use it." What we're doing is we're tackling specific problems. What we did with Fusion, though it's kind of down in the weeds, but what we did with the constrained automation tool and Fusion is a great example of that. It's a very labor-intensive thing that our customers do, but now we're automating it 2.6 million constraints automated so far to date. And the acceptance rate is above 60%, which is very high for any kind of generated feature. Those kinds of things, you're going to see more and more of those coming out, and they'll get adopted very quickly, because we've specifically designed them to problems the customers have today or areas of bottleneck. So we're focused on some of the bottlenecks. Some of the things we're rolling out next year will be focused on bottlenecks as well. And then what you'll see is as we move into workflow and systems level automation, you're going to start seeing us monetizing some of those things differently. Task-based automation would the customers are just going to get it. Workflow automation system, I mean, we're going to start monetizing those in consumptive ways over multiple years. So our goal is to make sure we deliver things that are easy for the customer to use, understandable and provide value quickly. That's how you drive the adoption, and we're seeing it so far. .

Yes. Let's talk about the monetization strategy because you guys introduced API-based pricing around the platform services. So could you elaborate why when you spoke with your customers, why did this pricing model make the most sense? What is the overall customer reaction been to that? And could this be something as needle moving to revenue as next year? Do you expect this to take time? What's the pace around that?

So API monetization and MCP monetization, they're kind of related, right? And essentially, what you're monetizing is machine usage of our IP, our products or some of our cloud capabilities. To date, some customers, very few could run a suite of our products, construction cloud, Revit 24/7 and execute on real workflows, basically creating their own digital human in the process. We're going to monetize that moving forward, and we're also going to monetize connections to our APIs through MCP servers and other types of cloud APIs as well. That's just the way things are shifting as machine usage becomes a larger component of the capacity creation for our customers. Today, it's a small number of customers that do that. So don't expect anything to move the needle suddenly. But over multiple years, more and more customers are going to be having machine usage of our products rather than kind of human usage. And that's going to be how this capacity shifts around from the human-based capacity to the digital capacity.

Yes. Makes sense. And how important is the data layer in all of this? Because when I was at Autodesk University, it seemed like there was a greater emphasis on platform services than maybe in the past -- and -- which makes sense, right? Because if you think about AI, these agents, they need to have access right to all the information internally, it's -- it makes a lot of sense in that regard. But how big of a catalyst could that be? And then a second part to this, could that actually be a catalyst for cloud migrations to form a Fusion, Autodesk, Construction Cloud.

The short answer is yes.

Yes.

No. Look, as you noticed Autodesk has moved from a design to design and make company, which, by the way, increases the number of personas we touch. We're going to -- you're going to see us move to an operate company as well. So it's design, make, operate. the number of personas that cut across that entire light is very large. What's the glue, the data flow. The data flow is the glue from the design process through the entire kind of preconstruction process or manufacturing engineering process, all the way to where you cut metal or build a building all the way through to operations. So the data layer actually provides the fluidity across all these various personas and allows us to kind of automate workflows between personas. And actually automate entire outcomes across all those personas. So that's going to be a big part of how we get there. That's why you heard the customers talking so much about it, because they are aware that the data is the unlock. So -- in the past, we had lots of proprietary data formats. We're embracing lots of open formats to make sure that people are using our platform to manage those data flows, and we're using those data flows to automate and create the digital capacity. So it's a big deal, and it is an unlock.

Yes. Do you -- are you starting to hear that appetite increasing with your customers in terms of realizing that cloud is important for AI? Or is it still very early days?

Oh, look five years ago, we really didn't -- 5, 6 years ago we didn't really have a SaaS native compound or business. Now if you look at what we've done with Construction Cloud Fusion, we're 20% round of our revenue is SaaS native. It's growing rapidly, growing high double digits, it's doing quite nicely. So customers are realizing that if they want to get to that future, they have to embrace SaaS as part of that. Otherwise, the data flows aren't as easy to manage, because you're trying to connect lots of behind the firewall silos with various other places. You've got people that are outside your firewall. When it's all in the cloud, you can connect to everything and bring it back and federate it across multiple things using the cloud as a hub for that. Customers realize that and as a result, adoption has gone up.

Perfect. You mentioned the opportunity a little bit earlier, so I'd love to dive deeper in on that. At the last Analyst Day, you broke that out in the TAM, I believe, for the first time it felt like there was a bigger emphasis on the civil and infrastructure opportunity. I know you mentioned that on your last earnings call. Could you talk about what's driving that focus on these opportunities? And am I right in the read that this feels that there's more emphasis behind it than maybe there was previously. And that could be a question for Andrew and then Janesh, a follow-up to that. How do you think about this opportunity as it parlays into the durability of Autodesk Construction Cloud and the greater AEC opportunities .

Yes. So operations is a natural extension of what we're doing with BIM and modeling and manufacturing, especially around mid-market factories. Mid-market factories need to operate like big factories. So the little factories and the little machine shops need to have the same level of automation and control that large factories have. So there's a huge opportunity as people reshore and redo the supply chain around operating progressively smaller and midsized factories. So we're very much focused on that as well as buildings. So operations is a natural extension. It extends into a life cycle that's well beyond the design and make process. Design and make process plays out over years, asset management operations plays out over decades. Right? So it's a big difference for us. And the unlock is really clear. Tandem is the first play into this with the digital twin technology. We're going to extend that. Look for us to attack that market the same way we attack construction. Six years ago, we didn't really have a material construction business. We stood up a division. We did some acquisitions of new tech stacks, accelerated that. And over a 5-year period, we built a highly successful, fast-growing business. That's the same path we're going to take with operations. And it will increase the stability and consistency of our business long term.

Perfect. Janesh.

And I just think about the growth longer term across design make and operate. Operators, obviously, a new area for us and adjacency we are going to get into. So I think that plays out into the revenue model over a much longer period of time. In the near term, if I think about design and make, and you see the numbers from the earnings release, and Andrew mentioned a minute ago that the underlying businesses like Construction and Fusion continuing to maintain and deliver strong growth for us, roughly growing at about 2x the overall design business or 2x the overall rate of growth in the business, give or take a few points in any quarter. But we see that continuing. Because fundamentally, in those markets, we are not dam constrained. Construction is -- has a long way to go in terms of adoption of digitization. Andrew talked about the mid-market opportunity in Fusion where, again, we feel like we have plenty of headroom for growth. So I see that make part of the business continuing to sustain growth for next year.

Perfect. Yes. And let's talk about the opportunities in manufacturing and media and entertainment. Any big initiatives going into your fiscal year 2027. And any catalyst the audience should watch out for in those sectors? .

Yes. There's a couple of unlocks in manufacturing. One is the data management layer that sits in the Fusion Cloud. We've invested significantly in that, and it's been a major customer request in terms of taking Fusion inside of some of our accounts from say, 2 to 5 seats up to 5 to 20 or 30 seats. So the customer has been asking for it. We've invested in significantly. That's going to be an unlock heading into next year for multi-seat accounts with Fusion. So we expect that to drive some growth. We're also, like I said earlier, very much focused on mid-market factories. And we have a product portfolio that allows people to lay out those factories, but we're also very much focused on helping them manage those factories. And I think that's a good multiyear growth opportunity that will show some [ kernals ] of growth next year as well. Media entertainment, the shift has consistently been -- remember, we're in the high end, film. We make products for the directors and the high end. We don't make it for the consumer market in the same degree. Though we want to bring some of that technology down to the consumer market, but we make it for the high-end film and game market. And more and more, we've been moving to managing the whole production process from the capture onset all the way through into post-processing and special effects. So we're making it easier for the customer to capture the data early, manage the process of making sure it's right and pushing it into post production.

Perfect. And then maybe switching gears to 2026 calendar year and what you're hearing from your customers as we close out 2025. Andrew, how would you characterize overall customer conversations in terms of how they're thinking about spend going next year? Do you anticipate it could be similar to what we saw this past year? Are there any opportunities for growth unlock? Any pressure points that you're monitoring closely. What are you seeing there?

In generally, the sentiment is consistent with what we've seen this year, going into next year. Remember, we have large EBA cohorts that renew on a regular basis. These are 3-year contracts, right? So we get kind of visibility of how peak customers are looking at their 3-year investment. No one is downsizing their EBAs. We're seeing expansion in the EBAs across basically all of our customers, which is a strong signal that you're going to see consistent kind of investment in technology heading into next year.

Perfect. And then another big opportunity that you guys now get a lot of questions on is just the data center opportunity. I would imagine this ties in a little bit to what you guys are talking about with operate, right, and the digital twin opportunity. Could you just elaborate the tailwinds right you're seeing around there and how meaningful that could be to the business as we look into next year and beyond. \\

The digital twin opportunity in particular.

The data center opportunity. But I would say I would imagine digital twin plays into that in the part infrastructure.

Look, with our business, our business is highly distributed across multiple segments, multiple geographies, right? And remember, I always said there's not enough capacity to meet the demand that's coming in. So whenever something is getting executed on, there's something else waiting in the queue saying, "Hey, pick me, pick me." All right. So right now, data centers are consuming a large chunk of that capacity. In the past, there was commercial buildings, community consuming some of that capacity. That execution is going to continue into next year. But even if it starts to fall off, there's other things that come in to consume the capacity. So we don't see any concern about sector shifts from one build-out to another, but we also see data centers continuing well into next year as well.

Perfect. And Janesh, let's dive into some of the financials and some of the nitty-gritty details. So as we look into 2027, there's still a couple of moving pieces on revenue and billings. So one being we're still working through, the new transaction change. So first question for you is in terms of the tailwind that, that could be the billings and revenue growth next year, any qualitative color you could give there and how that might compare to the 600 basis points of uplift we saw on the revenue side. And then not to throw too many questions at you. But a second piece, we get a lot of questions on billings, because you had a big renewal cohort this year. So as we look into next year, could that possibly be a headwind such that we see more of a disconnect between revenue and billings growth adjusted for FX and the model change? Is there a possibility that billings growth could dip below right, revenue growth? Do you expect them to be more in line? Just any color you can give there, I think, would be really helpful.

Yes. On the first question, Taylor, what I would say is I realize that the whole new transaction model has thrown a lot of our metrics often. It's just made it a little bit harder for people to understand the business. But as you said, this year, that's providing about a 600 basis point tailwind to the as-reported numbers on billings and revenue. And this year, fiscal '26 is the peak of the transition. We will still have additional tailwinds next year, but they will be a lot lower than what we saw this year. And maybe one way to think about it is the tailwinds on the revenue and billings also imply that there's a headwind on the operating margin. And so maybe one way to sort of think about this in terms of a rule of thumb is, this year, we've got about a 300 basis point headwind on operating margins. And what we said at Analyst Day was we expect that in the terminal state that we'll be about a 400 basis point headwind overall on operating margins. So that gives you a sense of how much of the transition we are through and then you can sort of back into the fact that next year, we'll have a very modest tailwind on the billings and revenue side. And when we get to that point, we will actually spell it out for you like we did this year in terms of what the exact underlying math is and what the as-reported views will be, but you should expect it will be a lot lower than it was in fiscal '26. And then to the other part of the question in terms of thinking about the comps on billings on an underlying basis next year, I think a big part of that depends on how we close Q4. We've got a very large EBA cohort still to close. Most of that is loaded towards January. So I think that will largely dictate what the underlying business growth is next year. But -- so it's hard for me to give you any specific views on '27. At this point in time we are working through our plan ourselves. But ultimately, you should see billings and revenue converge.

Perfect. That's really helpful. And then maybe moving to the go-to-market transition. So we talked earlier, this has been a multiyear approach. We've seen a number of these transitions. More recently, you've talked about the move to self-service. That's been a tailwind to the Autodesk store. Next year, it sounds like there's going to be some changes to comp incentive plans for partners. So could you just unpack a little bit of that more? What changes should we expect going in next year? Andrew, that could be one for you? And then, Janesh, in terms of how we see impact to the model from those changes? Maybe you could talk through what you're contemplating there?

Again, we always said it's a multiyear go-to-market transition. The big part that we saw this year was getting the self-service ramped up, getting through the new transaction model changes. We've lapped out. We're in good shape. Heading into next year, what you're seeing is this now -- all right, let's turn that self-service around renewals and things like that into an advantage. So the partner comp structure is changing. The amount of money we're putting in commissions into the partner ecosystem isn't changing, but how they earn that money is changing. Significantly less money is paid out on renewals, a lot more is paid out on new business generation. And that's part and parcel of how do we turn the new transaction model, the intelligence we've got about the customer, the self-service we're building in around renewals into incremental value for Autodesk in terms of getting out to those new personas we've been talking about, getting some of those new products growing in particular ways. That's the first step of doing that. Now obviously, there will be some churn and kind of disruption associated with that. We factored that in, but that's the big change that's coming next year.

Perfect.

Yes. In terms of the model next year, I think Andrew covered the highlights already, right? For us, this has been a multiyear transition. We signaled to the investment community to our partner community, what we were doing long before we did it, and then we went ahead and did it. And so far, it's played out consistently with the way we would have expected. And there's more to be done. We signal to the partners that we're going to make the changes. Then we announced the changes in the third quarter, so well ahead almost 6 months ahead of when they actually take effect. And so we've been very consistently telling the investment community and the partner community that this is the direction we're headed. And so that's all played out quite nicely for us. So that just underlies that consistency of execution and momentum that I talked about earlier.

Perfect. And Janesh on margin. So you have the out-year target of 41%, 45% on an adjusted basis. So can you talk through the trajectory in getting there? Because you've commented earlier that, that's not going to be a linear path. So maybe you could elaborate a little bit more on that. And then as we think about the levers to getting there, I know on the AI side, you've talked about the potential to see gross margin compression. I would imagine the offset of that is operating OpEx efficiency. So how should we think about those moving pieces?

Yes. Taylor, maybe I'll take those in reverse order. I'll just describe what the levers are, and then I'll talk about the path to getting there. So in terms of the underlying levers, the biggest one is going to be our continued go-to-market optimization. And that continues into next year. It's been a multiyear journey for us. So -- as we said, that's been progressing nicely so far, but there's more to be done on that front. The other big lever for us is just the inherent operating leverage that we've got in the business. You've seen us deliver consistently against that this year. And as we continue to grow, the marginal dollar of revenue may not need the same marginal dollar of investments, and you will start to see some of that. And as we get that operating leverage and realize that, it's a conscious investment decision we make in terms of how much we invest back in the business, particularly for R&D and we think about our future from a long-term perspective versus how much we flow through to the bottom line. So that's played quite nicely for us. Embedded in that is the fact that as some of these cloud-based workloads increase, they will put pressure on the gross margin. It's gross profit accretive in terms of dollars of gross profit, but in percentage terms, it does create a little bit of margin dilution and we factor that in into the 41% that we outlined. So those are some of the big levers that we think about. In terms of the path to getting there, we've said it will be a nonlinear path. And the main thing there is when we talked about the incremental headwinds from the new transaction model on operating margin, most of that will get layered in next year. So it's more around just the mechanics of the new transaction model rather than anything underlying in the organic health of the business. In fact, this year, it's been a good year for us when we laid out that target at Investor Day back in October. At that point, our current year view was 37%. We've already bumped that up 50 basis points, so we're already well on our way there.

Perfect. And maybe in the last minute or so that we have left, we can talk about a couple of the pricing levers Autodesk has -- so I know you guys have spoken about there's a cohort of your customer base that's sitting on heavily discounted licenses, right? There's maybe the opportunity to monetize that a bit more as we look ahead. So could you just talk about how you're thinking about the opportunities from price going forward, how that could contribute to growth? And then also two, as you see the pricing model evolve. How should we think about that playing into the mix too.

Yes. So to start with that cohort comes up in fiscal '29, so it's a long time away. And even when that comes up, we give those customers a very natural glide path of some sort. For those of you that are thinking way out into the future already, I would not think about a step function change, but it's more of a glide path. But more broadly, in terms of pricing, we think about pricing largely as a function of value and I'll keep the response brief because we're almost out of time. But inherently, as we deliver more value to the customers, we think about how we monetize that. That could be in the form of price increases. That could be in the form of new products. But importantly, as we think about new monetization models, particularly in the world of AI, there will be some functionality that gets delivered in the existing products for the existing prices around task automation, as Andrew mentioned. And then as we think about more usage-oriented models, those will be more for workflow automation and system automation kinds of use cases where the value that we deliver to the customer will largely be metered differently and measured differently. So we'll use consumption-based models for those.

Perfect. So with that, we're out of time. So everyone, let's give Andrew and Janesh a round of applause, and I appreciate all your insights today.

Thank you, Taylor.

And Taylor as well. Thank you.

Thank you, everyone, for joining.