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

All right. Let's get started. Thank you, everyone, for joining us. My name is Keith Weiss. I am filling in for Elizabeth Porter, who just had a beautiful baby girl. But actually really excited to be talking with Autodesk, speaking with Janesh Moorjani, CFO of Autodesk. Before Elizabeth took over, I covered Autodesk for a long time, and it's always been one of my favorite companies.

Mine too.

Excellent. A long, long time ago, I came out of college as an idealistic young man, wanted to be an architect. So I've always been pining to be back in this field. And now you're settling into your role as CFO. You've had some time to take a look at the sort of scope of the business. I'm sure you're excited about it when you kind of came in the door. Now with some time under the belt, maybe you could talk to us about what excites you most about that longer-term potential at Autodesk?

I'm happy to, Keith, and thank you for hosting, by the way, while Elizabeth is out, so we appreciate that. Before we talk about the long-term excitement, what excites me the most in the short term is reading a safe harbor statement. So let me do that first. 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. That was so exciting. As you said, I've been here about 14, 15 months now. And when I first joined the company, it was on the thesis that we've got a stable, consistent, resilient growth business. And superimposed on that, we've got some terrific catalysts in cloud and AI and all the investments the company was making on the platform side as well as the opportunity to continue to create shareholder value by driving margin expansion. So it was a terrific trifecta of all of those things. And I think what excited me over the past 14 or 15 months, I think that conviction has largely played out. And what excites me the most is, again, just those same three things about the future. When I look at the market opportunity ahead of us, we've got some fundamental secular trends that are working in our favor. You look at the adoption of technology within spaces like construction. You think about the digitization of manufacturing, you think about all of the investments that are going into infrastructure and so many of the other spaces in which we operate. So we've got some terrific growth drivers in the industries in which we operate. That's working really well for us. And we've been investing heavily in cloud and AI for many years now. There was a point in time in the past when people used to sometimes think why we're making those investments. And clearly, those were the right decisions to make based on where we are in the present. And the future potential for those is, I think, incredibly exciting as well. And we've proven the ability to deliver strong margin expansion, and there's more to come on that in our long-term target as well. So all of those things still exciting about the future as well.

That's a great opening to the conversation. I want to start on kind of the near term, and then we'll expand into sort of the longer-term product vision. Last week, you guys reported earnings, and it's really a solid set of results across the board. You beat on billings, revenues, margins, free cash flow, all ahead of expectations. Maybe you could sort of help us dive into that and provide a little bit more color about the overall health of the business in Q4 and as you head into your next fiscal year, how are you feeling about both the environment that you're operating in and the efficacy with which Autodesk is operating in that environment?

Yes. Q4 was a great quarter for us across all fronts. When I look at the business from the standpoint of industries in which we operate or verticals or segments or geographies, we did really well across the board. Data centers did really well. Industrial buildings did really well. If I look at it from a geographic lens, emerging markets where there's tremendous infrastructure build-outs happening, we continue to do well on that front as well. So overall, no matter which dimension you look at, the business continued to do well, and that's on the back of strong expansion business as well as new business growth that we were seeing. And it goes back to the fact that in many of our businesses in which we operate, fundamentally, our customers are actually capacity constrained. And there's not enough people to do all the work that there is to be done. In construction, as one example, there's probably 8 to 10 months of backlog. And so when you see strength in one area like in data centers, it's coming out of some place else, but there's always a volume of work to get through. And over the course of last year, if I think about the economic backdrop, we had a lot of noise in the external environment, but we never really saw that flow through to customer buying patterns. Customers were concerned, but they still made the investments and they still wrote the checks. So we feel very good about how customers spend through that time frame just because they see the same underlying demand drivers over the long term. And we expect the same will continue in fiscal '27 as well.

Right. So if we dig into that, Autodesk has historically been seen as more macro sensitive than most in software. But it sounds like the customers are starting to understand the value proposition better, right? It's not a cost that comes in with new projects. It's a way of making your business more effective. It's a way of enabling you to get to that backlog. It's a way of augmenting a human capital. So is this making your business less cyclical? Is it enabling you guys to grow more durably despite what is -- it feels like always going to be a very noisy kind of macro backdrop.

Yes. That is my sense. The business has been a very consistent grower. If I look back in time over the past 3, 4 years, we've delivered a very consistent level of growth. That has been across different elements of noise or volatility that you see in the external environment. I think a big part of that is just the diversification of the business as well that we have. We've got strong presence in AEC, but also in manufacturing and a much smaller presence in media and entertainment, heavily diversified across geographies. But even within AEC, the Autodesk of 10, 15 years ago was an Autodesk that focused solely on design -- plan and design. And since then, we've extended into the make, which was construction and even on the manufacturing side with Fusion. We've got declared ambitions to complete that life cycle and go into the operate phase where over time, then our involvement with projects is not going to be measured in months or years, but it's going to be measured in decades because that's the life cycle of these assets. And that will then inform what happens earlier in the plan and design stage as well. So as we've grown the business across all of these different vectors, that's helped bring a level of stability to the business as well and the ability to withstand some level of external movements. No business is completely immune from the macro, but we feel pretty good about the business we've built.

It's a fascinating expansion of the product strategy and definitely going to dig into this. Basically, Autodesk is becoming almost the operating system for physical assets, whether it be a building or a product over time, and it's hugely expansive to the TAM. But before we go there, I want to talk about margins. You're the CFO, I got to dig into margins. It's my contractual obligation. So FY '27 -- FY '26 was a really good year for margin expansion, 200 basis points of margin expansion. You guys think you could add another 75 basis points in FY '27. Part of that is a restructuring that you talked about a 7% reduction in workforce. Could you talk to us a little bit about the puts and takes of that? Like where is the -- where are the heads coming from? How much of that gets incrementally invested back into the business? How do you think about that balance between pushing more leverage, pushing more margin versus going after that big growth opportunity that we've been talking about?

Yes. I'll put that in the context of some of the broader business model changes that we have been making because that's actually helped us drive the margin over a period of time. And you may not have followed the Autodesk story recently. I know you did some time back. But what we did was we fundamentally changed how we go to market in what we call the new transaction model. And with -- in many of the developed countries around the world, we now have a direct relationship with our end customers. The partner still plays a very important role, but we -- the customer places an order on us directly that allows us to build better integrations with them, better understand how they are using the products. We get great demand signals from them. We can use that back into funneling new and expansion motions. Now one of the mechanical effects of that model was it changed the accounting, which increases our revenue and increases our expense, which the net result is sort of a mathematical headwind on operating margin in percentage terms. It doesn't affect the dollars, but it affects the percentage. So this year, as we think about all of the changes that we made on the go-to-market side. We started off with the implementation of that new transaction model. We made some changes on the partner side as well. And then in preparation for being able to work more closely with our customers directly, we did one restructuring a year ago. And in that restructuring, we largely took out nonselling resources, customer success, marketing, sales operations, and we reinvested some of that back in the business to build new capabilities that we knew we would need for this year. And what you saw us do here in January in this latest restructuring, which is the final step of that overall business model transition and go-to-market optimization, as that final piece of the puzzle comes together, we did impact some of the sellers, but we also have new tools and capabilities that allow us to work with those customers a lot more efficiently. So where -- some of the savings from the restructuring are actually going to be reinvested back in the business. Some of it will be in selling capacity because we took some people out we need to hire in other locations for different skill sets, different capabilities. Some of that investment will also go back into the marketing functions. Some of it will go back into core R&D work that we're doing, particularly around AI and platform. And so if I think -- bring this all back to op margin and think about the various puts and takes, that 75 basis points of expansion at the midpoint, that reflects the savings from the restructuring net of the reinvestments that we're making. It reflects the inherent operating leverage that we've got in the business. And it also absorbs within that roughly about 100 basis points of incremental that mechanical accounting percentage headwind. So we're effectively absorbing about 100 basis points of that in this margin expansion that we've got as well. So there is core operating margin inherent in the model, and we're really happy that we're able to deliver that. And this sets us up very well to achieve our long-term margin objectives.

So within that sort of incremental leverage, I spent 10 minutes without talking about AI, and then that's as long as I could go during this TMT conference...

I brought it up.

You brought it up. I haven't asked the specific question. How far along the journey of kind of utilizing these technologies internally from a development standpoint, the customer service standpoint, improving go-to-market, are you guys -- have you guys gone, right, in terms of being able to run your business more effectively? And then we'll turn to sort of like what you're pushing out to customers.

Yes. We have been big users of various AI tools and technologies internally. When things first started to move, call it, a little over a year or so ago, we had 1,000 flowers bloom. We just encourage people to keep adopting it and using it in creative ways to see how they would actually be able to benefit from it in their day-to-day work. A lot of it was just basic productivity enhancements, but productivity enhancements, not in the form that you could immediately monetize. It made us -- it helped us better prepare for our earnings call, but that doesn't translate into real dollars saved anywhere. So that's just one example. But on the development side, now we've started to do a lot more. We've rolled out cloud code quite widely within the organization. Developers are a lot more productive. And the way I think about this is over the course of the long term, we're going to invest a certain amount of dollars into hiring additional engineering capacity, hiring headcount. And what this allows us to do is shift some of those headcount dollars to technology dollars. We don't want to use AI to limit the output of the organization and drive cost savings. But the way we're thinking about this is if we can use it to drive greater velocity and greater productivity in the development organization while staying invested in R&D, it will just help us accelerate that much faster, and it will help pull the future into the present that much faster. So we're using it as a way in which we can drive greater productivity. And in the initial stages, that also means a greater level of investment on the part of the company because we want to drive the adoption so people see the benefits and don't let the cost be a reason not to adopt the technology in the early stages.

Got it. Got it. And then shifting gears to the opportunity from AI. You said earlier that Autodesk has been talking about AI for a while, right? I remember probably a decade ago, you guys started first talking about generative design, which was utilizing a lot of these machine learning technologies. But now there's new capabilities that are emerging, new large language models. How does that expand the opportunity? Like when you look at these new capabilities being able to reason over unstructured data or whatnot and apply that to the frame of reference of what Autodesk is doing for your end customers, where do you see the white space? Where do you see the incremental opportunity of what Autodesk is now going to be able to do for their customers that 3 years ago, 5 years ago, we really just weren't?

Yes. It's a great question. You have a great memory, by the way. You remember things that we said a long time ago. So yes, we did start investing in AI almost a decade ago. A lot of that was with our initial research capabilities. We built out teams and expertise over time. The level of AI skill sets in the industry is -- it's a very rare skill set. 3D skill sets are very rare as well in the intersection of both of those, those are really hard skills to find. So we started to invest in those capabilities and build those out. The way we think about the opportunity is a couple of things. One is, as I mentioned, people in our industry are fundamentally capacity constrained, right? They just don't have enough people to take on all the projects that they want to take on. So as AI helps users become much more productive and reduce risk, it allows them to take on more projects with the same level of staffing that they have. It also means that if they can prevent mistakes that would occur downstream, which cost them millions of dollars, if they can prevent those upstream in the planning and design phase, it derisks their projects and again, makes them that much more productive. So fundamentally, when we think about the core customer problem we are trying to solve, it's about capacity and productivity for our customers. In terms of how we go about doing that, there's essentially 3 layers of monetization that we think about. The first is what we just call task automation. It's simply features in a product that makes your users more productive. It's a feature that allows you to do something faster and then you -- and take the mundane work away from you, and so you feel like you had a much better experience and got a lot more done. That doesn't make our own COGS meter spin any faster. That's just value we deliver to you through the subscription. So we would monetize that. And that also doesn't reduce the number of seats. So we would just monetize that through a traditional seat-based model where the price that we charge for the subscription is directly linked to the value that we deliver through that subscription. Above that, you have more consumptive types of workloads, which are essentially helping them automate entire workflows or even entire projects, what we call system automation or system-wide automation. So if you think about task workflow and systems, the workflow and systems automation will be much more consumption-oriented or usage-oriented. And then the way we monetize those will also be through usage-based models because those will also be more resource-intensive on our end. And so we had -- we described this all at Autodesk University, which is our user conference some time back. We've been working with customers along this. And we are very early, relatively speaking, in the adoption journey. A lot of what you see come out will initially be around the task automation side, but workflow automation and system automation is also on the road map, and we are starting to work with many customers who are in the early stages of trying those out. And then the important piece is that over the last 8 years or so, as we built the platform capabilities, we've built the technology stack to support it. And through the whole go-to-market transition that we did, ultimately, we knew this is the path we're going to head down. And so we've actually been building consumption-oriented capabilities in the go-to-market organization as well. Back in fiscal '25, 17% of our business was in consumption-based models. So we've already built the expertise and the knowledge to take these offerings to our customers.

And then is there a gross margin impact that as investors should keep an eye on? You've talked about a lot of these capabilities as more compute intensive. How should we think about that trade-off, the monetization that you just described versus perhaps higher COGS on the other side to power these solutions?

Yes. There will be a gross margin impact in percentage terms over the long term. These workloads will still be gross profit dollars accretive. So it's still the right thing for us in terms of creating shareholder value, but it does have an impact on gross margin. But that's one of the things that we factor in as we think about the inherent operating leverage in the model. And when we laid out our long-term margin target at our Investor Day some months ago, that was certainly one of the considerations as well. So it's one of the different factors that we manage in the overall mix for sure.

All right. So I want to switch gears a little bit and dive into product strategy, starting with the AEC side of the equation. One of the things that I've always found super interesting in the Autodesk story, there's always a lot of focus on the A side of the equation, the architecture and the design and you guys have had a pretty dominant sort of positioning in that, but there's so much white space in the C side of the equation, right? And now you guys are bringing together that design and make into a cloud-based environment. Construction, in particular, has been under sort of technologized or digitized, right? There's a huge amount of inefficiency and waste that takes place in construction. That's not new, but it feels like it's a more fertile ground for adoption of those technologies. So can you talk -- maybe to start out, talk about why is now the right time for bringing these 2 domains together? Why is right now the right time for getting better traction in terms of digitizing what's going on in construction?

Yes. It's a great question. I mean we started to think about extending beyond the pure design stage of the life cycle all the way back in 2018, as I hear Andrew sort of describe some of the history of why we got into make. And we started to invest in those capabilities because ultimately, to your point, when you look at construction, it's one of the least digitized industries there is. And the construction business is -- it's a really tough business. If you overbid on a project, you don't win the business. If you underbid on it, you may lose your shirt. And so it's a really tough business with wafers and margins. And so we saw this problem and said, look, if we can help these customers, the GCs, the subs and others think about ways in which they can manage the risk in their projects more effectively and make them more productive, that's a significant market opportunity. But also if we connect that upstream all the way back to the owners, where if you think about the entire life cycle, having the owners, having the architects, the designers, the GC, all on the same collaborative platform working through the life cycle of that project, there's a tremendous opportunity there, too. And we've been chipping away at it for some time. The diffusion of technology always takes a long time, especially the further away you are from key centers like the Bay Area or New York or what have you. And so we've been chipping away at it for some time, but we feel like these underlying growth drivers that have been in place for some time, we see them continuing into the future as well. So we feel very good about our position in construction and across the entire life cycle.

And it also matches kind of what you're seeing taking place with your end customers of that the design to build within one firm is becoming more and more part of the overall environment. And again, you guys become that system of record, right, for the data, the process, the workflows all along the life cycle of that building, and we'll get into the product side of the equation as well. How do you think about the Autodesk competitive advantage, particularly when we're talking about Construction Cloud, something you entered later. There's some -- vendors have been focused on that part of the solution for a while. How do you look to differentiate? And how has that competitive dynamic been playing out in the marketplace?

So when I step back and think about the core differentiators for us, right, we've got naturally a very strong industry presence and awareness and familiarity and a very high level of trust that we enjoy with our customers. But if I think about it from the standpoint of particularly in an AI world, as you think about the core differentiators of Autodesk as a platform relative to some of the other things that might be out there, we touched on this a little bit in the most recent earnings call, but it comes down to 3 fundamental things in our minds. One is the data. The second is the context and the third is the expertise. And so just to touch on some of those things, we've been -- as we invested in AI, we've been training our foundation models on the data that we've acquired from customers over many, many years and years of projects, hundreds of thousands of customers and years of experience with that. The customers naturally own their own data. We don't own it, but we do have the rights to train our models on their data, which allows us to build much more powerful capabilities in terms of model outputs than things that might be trained on data that's purely in the public domain or unlimited data sets because all of this data doesn't sit anywhere outside in the public world. And even it's scattered across all of our different customers. So it's just with Autodesk where we have all of this access to this data. That allows us to build much more powerful capabilities and models. The second is the context behind that. As we think about collaboration across the workflows, as we think about the recommendations that come out of it, the models need to be able to use the context of the overall project before you can actually make a commitment or not. The auditability, what the previous decisions were, what the implications are of decisions around everything else. Maybe to use an example, right, if you sort of look at a simple visualization tool and say, here's a drawing and I want to move this wall 10 feet from here to there, it's easy enough to just do that on a screen. But what do you do about all the things that you can't see? When you move that wall, what does it do in terms of egress requirements? Are you now still satisfying the fire code? What about the mechanical, electrical and plumbing, which you really can't see? Did you just cross over some pipes or wires on that front? So there's a lot of context and richness and awareness of the project itself, not just in terms of who all the collaborators are, but what all the downstream implications are. And using the data that we've got, we have that richness of context that we can then bring to bear in that example. And the third was the expertise that we talked about earlier. So I think those are the core differentiators for us as we think about the end-to-end workflows for customers as they -- particularly as they adopt more AI-oriented solutions from us.

Got it. So if we think about the Construction Cloud in particular, you guys have pulled together a pretty comprehensive suite. Are there parts of the suite that are getting more traction than others? Are there particular landing points that you've been particularly successful with?

Well, I'd say the landing points are more around wherever the nature of the demand is, right? We see a lot of strength from data centers right now. And the insertion point also might vary with the kinds of projects that you're landing. So in data centers, you have very sophisticated owner operators that want to actually dictate what the technology stack is for the entire ecosystem so they can centrally manage the entire project and also operate the entire project in the future. There may be less sophisticated use cases for other construction projects where the owners may not be as involved and they may leave the technology decisions to GCs. And in that case, that's fine, too, because we work with the GCs and we'll -- the sale to them is around how it makes their lives easier by coordinating across all their subs, making them more productive and helping reduce their risk. So I think that varies based on who the personas are that you're talking to. But the net result of all of that is they all see the opportunity to drive productivity and reduce risk.

Got it. Got it. Switching gears to Fusion. Fusion remains one of the fastest-growing products in the manufacturing industry. Can you talk to us about what's been the secret sauce enabling the solution to really sustain that high level of growth for really the past almost decade now within Fusion. And you've expanded out the portfolio as well. And as kind of a follow-on question, any particular strategies to improve like attach rates and get higher attach rates around Fusion of some of the extensions?

Yes. It's -- Fusion is really exciting for us as we think about the complexity of manufacturing and the opportunity to drive greater digitization in manufacturing. When most people think about manufacturing, they think about the complex manufacturing at the high end, where there's a lot of large incumbent players, highly complicated processes and so forth. The vast majority of manufacturing happens actually in the mid-market and the lower end of the market, where you're thinking about designing simpler products like a desk or like a flower base or what have you. And then you're thinking about how those will get mass produced. And so we see significant opportunity there. That's been our historical sweet spot. We've played in the mid-market. And for Fusion, a lot of our initial adoption has been in very small deployments, 1 and 2 seats, 3 to 5 seats. So they are not very complex, but we see a significant opportunity, particularly with moving into more multiuser situations, and that's, call it, 10 to 20 seat accounts for midsized manufacturers. And as we do that, that naturally drives significantly more seat count and growth. But also part of the keys to unlocking that is we also need to continue to invest in greater data management capabilities in the Fusion platform. So those have been road map items that we've been investing in for some time, and we had some releases over the course of fiscal '26. There's more to come in fiscal '27 and beyond. So as we continue to build the capabilities that are suited for that midsized manufacturer, we will continue to move upstream, and I think that will continue to sustain growth for us quite nicely. The other dimension to that is also the complexity of the products that are being manufactured as you go from simpler products to more -- there's tremendous opportunity in just core design and manufacturer of consumer products. And so that's another focus area for us. It's a different dimension or a different way of looking at the same opportunity set in terms of what products are being manufactured, but you'll see more of that in the future.

Got it. And it seems that there's also an element of that full life cycle that you guys have embedded into the Fusion solution or that the product data sits right with it is almost a database, right, of sort of that product data and the component that you could utilize not just in the design, but also in terms of the maintenance of that product over time.

Yes, absolutely. So when we talk about our ambitions in the operations space, it's not just limited to building operations as in the O of AECO, right? It also is all about manufacturing operations, where there's a very large opportunity set as well. The operations space is highly fragmented. And our approach to operations is going to be very similar to the approach that we took in construction, which is we are going to place a couple of big markers down with acquisitions and then build around them. But those will be focused on both AEC, but also on manufacturing operations because the life cycle of what you're manufacturing and then the closed loop to take it back to the plan and design and make phase, that's an incredibly powerful cycle that -- and that can be the gift that keeps on giving for a long time to come.

Got it. So looking forward a little bit, you guys have introduced Project Bernini, right, a generative AI model for 3D. You guys talk about it as a professional-grade foundational model for design. How should we be thinking about the time frame of taking a foundational new capability like that, starting to put into the product and then eventually starting to monetize it more effectively?

Yes. Bernini was the beginnings of what we did with foundation models, right? And it initially started off in our research labs. There's a gentleman called Mike Haley, who actually has been on the road with Simon and many investors have met him as well. And Mike and his team have been really the brains behind a lot of what we've done on that front. And since Bernini, we've invested and developed more in terms of additional foundation models that we've started to bring to market with some of our customers. And those foundation models are the ones that are trained on the data sets with our customers that I talked about earlier and are delivering much more powerful capabilities and frankly, more cost effective as well because they are tailor-made to our particular domains and use cases, so they're not only better outcomes, but more cost effective as well. So you're starting to see many of those things. We showcased some of the capabilities around that at Autodesk University back in September. Over the course of the next few months, you'll see many more launches from Autodesk in terms of product capabilities and specific products where they will bring some of those capabilities to life. So it's still very early stages, but we are super excited about where that can go in the future. And as we think about the workloads that end up, as I said, being more compute-intensive for us, those will get monetized over a longer period of time. The earlier adoption of that you see and the earlier manifestations you'll see will be more in task automation-oriented workloads.

So shifting gears again, I want to go back to kind of go-to-market. You brought up the transactional model earlier, which is another step on what has been really a long journey for Autodesk in changing the nature of your relationship with the customers. I would argue part one was moving to a subscription model in the first place. You talked about the accounting impacts, which is important for this crowd, but there's also the benefits, right? Can you talk to us about some of the benefits of having that more direct relationship being able to really have that communication or that to build, but still the communication with the end customer.

Yes. It's -- that was the whole intent. When we said we want to have that direct relationship, it was because we felt that knowing what a customer is using and how they are using it will give us much better signals to be able to then go back to them and represent the full breadth of the portfolio that we have and drive greater expansion in those accounts over time. That was the core thesis. That's why we went through all of the pain of the implementation of that model. That's why many investors went through the pain with us as we got through the numbers and the financial impact of that transition. And when we implemented that as there was some level of operational friction in the latter part of '25, in particular. And we knew that going in, and that's all subsided over the course of '26. So if I think about the benefits of that, there's a few things that I think about right away. First, that's actually helped us be a lot more efficient in our go-to-market model. And we've already seen that reflected in the form of the restructurings that we did and reduce the cost and improve the margin profile of the company. So that's a real and tangible benefit that we've already started to realize. And over time, what ends up happening is we can continue to drive greater new business growth. Again, we saw very strong results in fiscal '26. Some of that is naturally the benefits associated with the new transactional model. It's really hard to sort of unpack it and say this was a benefit and this was organic growth because ultimately, it's one engagement model that you have with the customer. And then over time, as we continue to improve on the signals that we're getting from the customer, although we rolled the model out in '25, we really started to see initial signals from the customer in '26, particularly as we lap the first renewals. So now we're in a much more stable state. And as we get -- continue to get better signals from the customers and can engage with them, we can develop the right campaigns, we can develop the right plays with the expansion teams that we've put in place with the restructuring that we just did of the sales organization. So we feel that that's now the setup for driving even greater growth and expansion through those customers in the future.

Got it. So if we bring this all together, multiple vectors of growth for Autodesk. We're talking about improving the productivity of your end customers. So your existing customers become more productive, you're driving more value from them and you have monetization avenues for that in terms of the consumptive. We're talking about going broader across the life cycle of both sort of the buildings and the products that's enabling you to get into more categories, more fully into C, more fully into O with the operations across, again, both building and the product side of the equation. And now you have a better relationship or a tighter relationship with the customers to better understand their problems. How should we think about this sort of all coming together in terms of what's the durable long-term growth that we should expect to see from Autodesk, not just for FY '27, but for the 3 and 5 years beyond?

Yes. I mean you summarized it perfectly, Keith. Those are the things that I think about. And if I go back -- and if I go back to where I started in terms of what excites me about the future, it's the secular growth trends that we see, all the work that we've done to be ready to capture that opportunity through the products, through the go-to-market changes. It's the investments that we've made in AI. Just as we are implementing this final phase of the go-to-market optimization, we just need to be a little bit careful because there's sort of near-term disruption in the sales organization associated with the implementation of the [indiscernible] and the expansion teams and so forth. That's part of what we baked into our guidance as we said we'll be a little bit cautious in the near term associated just with that one factor. But other than that, I feel really good about the long-term growth of the business.

Yes. Unfortunately, that takes us to the end of our time. But Janesh, thank you so much for joining us. This is a great conversation.

Thank you so much.