Loading
Thank you for standing by, and welcome to Autodesk Fourth Quarter and Full Year Fiscal 2026 Earnings Conference Call. [Operator Instructions] I would now like to hand the call over to Simon Mays-Smith, Vice President, Investor Relations. Please go ahead.
Simon Mays-Smith: Thanks, operator, and good afternoon. Thank you for joining our conference call to discuss Autodesk's fiscal '26 fourth quarter and full year results. Andrew Anagnost, our CEO; and Janesh Moorjani, our CFO, are on the line with me. During this call, we will make forward-looking statements, including outlook and related assumptions on products, artificial intelligence, sales and marketing optimization, go-to-market strategies and trends. Actual events or results could differ materially. Please refer to our SEC filings, including our most recent Form 10-Q and the Form 8-K filed with today's press release for important risks and other factors that may cause our actual results to differ from those in our forward-looking statements. Forward-looking statements made during the call are being made as of today. If this call is relayed or reviewed after today, the information presented during the call may not contain current or accurate information. Autodesk disclaims any obligation to update or revise any forward-looking statements. We will quote several numeric or growth changes during the call as we discuss our financial performance. Unless otherwise noted, each such reference represents a year-on-year comparison. All non-GAAP numbers referenced in today's call are reconciled in our press release and supplemental materials available on our Investor Relations website. And now I will turn the call over to Andrew.
Thank you, Simon, and welcome, everyone, to the call. We delivered strong fiscal '26 results today with billings, revenue, non-GAAP operating margin, non-GAAP earnings per share and free cash flow, all above the high end of our guidance ranges. As demonstrated at Autodesk University and our Investor Day and reflected in our results today, we remain well positioned to deliver for Autodesk customers and investors. We have successfully executed one of the most far-reaching transformations in enterprise software, redefining our business model, evolving our go-to-market, reimagining our products and scaling our platform.
In January, we completed the final phase of our go-to-market optimization. While initiatives like this are difficult and complex, they are making Autodesk more resilient and unlocking new avenues for growth and margin expansion. We're also enhancing our portfolio with cloud-based platforms and capabilities that seamlessly connect design, make and operate workflows. These platforms enable partners and customers to extend and customize our solutions, driving greater value creation and expanding our addressable market opportunity.
And we're defining the AI revolution for our industries to empower customers with new tasks, workflow and system automations, and capturing shared value to subscription, consumption and outcome-based business models that blend human and machine capabilities. In the coming months, Autodesk will roll out powerful AI capabilities built on a combination of frontier models and our proprietary models that understand 3D design and make. Autodesk is building the future and the path to it for our customers. We have been preparing for and working towards the cloud and AI for more than a decade.
It's why I believe our best days and greatest opportunities lie ahead. I will now turn the call over to Janesh to discuss our quarterly financial performance and guidance. I'll then come back to update you on our strategic growth initiatives and why Autodesk is best placed to benefit from 3D agentic AI.
Thanks, Andrew. Q4 was another robust quarter for Autodesk, capping off a strong fiscal year. Autodesk continues to demonstrate growth and resilience, investing to advance our leadership in cloud platform and AI while also expanding operating margins. Overall, the underlying momentum of the business in the fourth quarter was similar to prior quarters and better than the assumptions we built into our guidance ranges.
We again saw strength in AECO, particularly in construction and emerging markets, with sustained investment in data centers, infrastructure and industrial buildings more than offsetting softness in commercial. EBA and product subscription billings, linearity of billings during the quarter and upfront revenue were also strong. Total revenue in the fourth quarter grew 19% as reported and in constant currency. The contribution from the new transaction model to revenue was approximately $137 million in the quarter.
Total revenue grew 14% in constant currency and excluding the impact of the new transaction model. Please see the tables in our press release, earnings deck and Excel financials for details by product and region. Billings increased 33% as reported and 30% in constant currency. The contribution from the new transaction model to billings was approximately $185 million in the quarter.
Billings grew 32% in constant currency and excluding the impact of the new transaction model. As a reminder, our billings growth rate in fiscal '26 was skewed by the new transaction model and by the transition to annual billings for most multiyear contracts. Turning to margins. Fourth quarter GAAP and non-GAAP operating margins were 22% and 38%, respectively.
GAAP operating margin was broadly flat year-over-year, primarily reflecting a restructuring charge of $100 million related to our go-to-market optimization. The action we announced in January marks the culmination of our sales and marketing optimization program and reflects our sustained commitment to both investing in our strategic priorities and achieving the long-term margin goal we talked about at our Investor Day. Non-GAAP operating margin was up 120 basis points year-over-year, benefiting from operating leverage from our revenue outperformance as well as ongoing cost discipline, partly offset by the margin drag from the new transaction model.
Fourth quarter free cash flow of $972 million benefited from overall billings strength as well as the linearity of billings in the quarter. Moving on to capital allocation. We repurchased approximately 1.1 million shares for $333 million in the fourth quarter. For the year, our share repurchases increased to $1.4 billion, a bit more than 50% of our free cash flow and reduced our shares outstanding by 2.1 million shares.
Recent share price weakness triggered greater share repurchases under our programmatic share repurchase grid. This increased share repurchase activity above guidance, lowered our average purchase price and further reduced share count. Let me finish with guidance. Our guidance philosophy is unchanged.
Our guidance continues to be based on the range of possible outcomes in our bottom-up sales forecast, which is grounded in the momentum of our business. We've assumed the macroeconomic environment will remain broadly stable through the year. You will recall, our guidance at the start of fiscal '26 was shaped by a prudent assessment of a few key factors. First, as we entered fiscal '26, we anticipated potential disruption from our restructuring and marketing, customer success and sales operations.
We see the potential for disruption again in fiscal '27 and we believe that both the probability and the potential impact of disruption are higher given the focus of the restructuring this year is on customer-facing sales functions. We have embedded this risk discretely into our guidance for fiscal '27. Second, last year, we anticipated potential disruption from our appointment of a new Chief Revenue Officer. This did not materialize as Andy Elder settled into his role rapidly. And finally, last year, I was new to Autodesk.
Over the past year, I've gained a deep understanding of the business and its resilience, which has strengthened my confidence in Autodesk's ability to grow at scale, invest in its strategic priorities and drive expanded profitability. So the way to frame our fiscal '27 guidance is that we expect the underlying momentum of the business will remain strong. Like last year, our constant currency guidance incorporates prudence to reflect temporary risk to billings and revenue as we operationalize our sales optimization plan. Unlike last year, it does not reflect additional prudence for a new CFO and CRO.
And of course, we cannot assume that we will get any tailwind from currency movements this year in the same way as we did last year. Let me point out a few items in our guidance measures. As you can see from our guidance, billings growth in fiscal '27 no longer has a tailwind from the new transaction model or from the transition to annual billings for most multiyear contracts. We expect billings to be slightly more weighted towards the second half of the year, in part reflecting our assumption that there will be short-term disruption earlier in the year from our sales restructuring and in part due to the weighting of our largest EBA cohort in the fourth quarter.
On revenue, the noise from the new transaction model will significantly diminish during the year from a tailwind of roughly 3.5 percentage points in the first quarter to roughly 1.5 percentage points for the full year. We will talk about it less as that noise fades. And again, our assumption that there will be some impact on billings from our sales restructuring earlier in the year is reflected in the implied revenue growth later in the year. Non-GAAP margins will reflect ongoing operating leverage, savings from restructuring, sustained investments in our long-term strategic priorities, roughly 1 point of incremental headwind from the new transaction model and prudence to reflect risk as we operationalize our sales optimization plan.
Free cash flow will reflect 2 discrete movements in fiscal '27. First, we expect cash restructuring outflows of between $135 million and $160 million during the year. And second, we do not expect to pay meaningful U.S. federal cash tax in fiscal '27 due to the R&D investment provisions in the One Big Beautiful Bill Act. The net effect of these discrete cash movements is immaterial to free cash flow in fiscal '27. Our U.S. federal cash tax payments will begin to normalize in fiscal '28. Additionally, we continue to manage our stock-based compensation with discipline. We expect SBC to fall below 10% of revenue in fiscal '27 continuing the trend of the last few years.
Reflecting all this, for fiscal '27, our billings guidance range is $8.48 billion to $8.58 billion. Our revenue guidance range is $8.1 billion to $8.17 billion. Our GAAP operating margin guidance range is 26% to 28%. Our non-GAAP operating margin guidance range is 38.5% to 39%.
Our free cash flow guidance range is $2.7 billion to $2.8 billion. Our capital allocation framework is unchanged. We will continue to deploy cash to the highest return opportunities, prioritize organic investment in R&D and accelerate the realization of our strategy with targeted and tuck-in acquisitions. And we will maintain a healthy buyback program with the goal of reducing share count over time.
Over the last few years, we've applied approximately 50% of our free cash flow towards share buybacks, and we expect to do the same in fiscal '27. We expect our share buyback in fiscal '27 to be similar to fiscal '26 in total dollars with the precise amount determined by our purchasing grids. Subject to acquisitions, we expect to apply approximately 50% of our free cash flow towards share buybacks over a multiyear period. In summary, we remain disciplined and focused on the controllable factors that drive our revenue, operating margin, earnings per share and capital allocation, which are the key building blocks of free cash flow per share.
The slide deck on our website has modeling assumptions for the first quarter and full year fiscal '27. As I mentioned at Investor Day, we are updating and streamlining certain disclosures to simplify Autodesk's reporting. There's more detail on those changes in the deck, too. Andrew, back to you.
Thank you, Janesh. Autodesk is focused on the convergence of design and make in the cloud, enabled by platform, industry clouds and AI. Let me give you some examples of our progress in the quarter. Our customers in AECO, architecture, engineering, construction and operations, are demanding convergence.
Convergence reduces risk, increases quality and optimizes cost and resource use during the design and build phase of an asset, and it enhances efficiency, resilience and reuse during operations. All of this is in service of deploying fewer resources to every project so they can bid on and win more projects with the resources they have. To better serve these needs, we intend to deploy capital to extend our footprint deeper into operations, applying the same playbook that proved successful in construction. Forma for Construction, previously known as Autodesk Construction Cloud, is a fast-growing component of our make solutions and has strong momentum with owners, designers, GCs and subcontractors seeking to converge design and construction workflows.
For example, following a competitive RFP process, Prestige Group, a top 3 real estate developer and asset owner in India, selected Autodesk as its core design to delivery platform for its engineering, digital transformation. We won back a major U.S. utility displacing a competitive solution with Forma for Construction, further demonstrating the value customers see in our modern, scalable platform, common data environment and tighter integration across design and construction workflows. A major hyperscaler is expanding its partnership with Autodesk to accelerate time to operation, cost management and digital twin workflows across data centers and facilities while improving collaboration and coordination across an ecosystem of more than 2,000 companies, including GCs and subcontractors.
As part of its enterprise digital strategy, a global pharmaceutical company chose Autodesk to be a strategic technology partner for the design, build and operation of its facilities to connect data, systems and workflows throughout the entire project and asset life cycle. Arup, a global engineering consultancy is expanding and standardizing on Forma for data-centric workflows, real-time collaboration across disciplines and geographies, and AI-driven insights to drive innovation and better outcomes for clients in the built environment. Three ENR Top 400 Contractors adopted Forma for Construction this quarter, including Roebbelen, which is replacing a competitive solution to leverage the power of a connected construction platform across preconstruction, construction and virtual design and construction.
These stories have a common theme, converging people, processes and data across the project life cycle to increase efficiency and resilience, decrease risk and prepare for an agentic AI world. Our comprehensive end-to-end industry clouds and platform drive convergence and extend our footprint further into the larger growth segments like infrastructure and construction that we discussed at Investor Day. All this is reflected in our strong momentum in both infrastructure and construction. In manufacturing, customers are demanding convergence as they invest in their digital transformation to leverage granular and unified data and embrace AI-driven automation capable of industry transformation.
By consolidating our design and make platform, customers have the flexibility and connectivity across workflows to increase agility, innovation and resilience. For example, after successfully adopting Fusion for in-house design and manufacturing of spare parts on production lines, a global brewing company is expanding the deployment to additional breweries to deliver cost savings and improve equipment uptime. A special purpose shipbuilding and marine engineering firm with more than 2,000 employees is adopting Fusion Manage as a mission-critical system for every new vessel project, aligning project management and multi-supplier collaboration in one place.
Typhoon, a Belgian industrial manufacturer, selected our manufacturing solutions to replace legacy unintegrated tools, which were causing lost engineering hours to non-value-added tasks and inefficient collaboration. Seeking a unified future-ready platform, a multinational automotive manufacturer is replacing a competitive solution with Autodesk design solutions to standardize workflows and improve integration and collaboration across creative and technical teams. A diversified industrial manufacturer is transitioning more than 900 users onto Autodesk's design and manufacturing solutions from a complex network of legacy systems.
While Autodesk Platform Services will be leveraged by sales teams to rapidly modify and visualize product configurations in customer conversations, a global manufacturer of advanced lithium-ion batteries is leveraging Autodesk's manufacturing solutions for asset standardization, digital factory simulation and simulation-driven quality improvements to drive growth and efficiency. Converged data opens up new opportunities for Autodesk. As customers seek efficient innovation, attach rates of Fusion's extensions are growing strongly, and we've delivered meaningful productivity gains to customers where we deploy AI.
We have continued to see success with our AI-powered Sketch AutoConstrain in Fusion. Since its launch last year, the AI model has delivered over 3.8 million constraints, up from 2.6 million last quarter. Along the way, the model has been retrained and the UX improved. As a result, the acceptance rates by AutoConstrain suggestions to commercial users have now grown to almost 2/3 with 90% of those sketches fully constrained.
In education, we expanded our relationship with Vellore Institute of Technology, VIT, in India, where students are applying industrial-grade Autodesk tools to real-world design challenges. For example, engineering students are using Fusion to design, simulate and optimize a formula race car. While architecture students are leveraging Revit and Forma to design, simulate and visualize complex architectural projects, which embeds sustainability and constructability. And lastly, we continue to find new ways for our customers to consume our products and services in ways that work best for them.
For example, Sobha, a real estate developer that designs, engineers, constructs and finishes units in-house uses our Flex offering to scale usage dynamically across projects and regions while maintaining full visibility and control over consumption and cost and aligning investments directly to business outcomes. Let me finish by talking about AI. Building agentic AI requires data, context and expertise. Scaling and monetizing it requires a platform and next-generation business models that can go-to-market.
Let me unpack that a bit by talking about what's needed to build agentic AI capabilities. First, data. AI agents need large quantities of physical world data to learn how to drive design, make and operate decisions. This data needs to be high-fidelity, contextual, geometry-rich, span 2 and 3 dimensions and represent physics and engineering-based principles for the entire design, make, operate life cycle.
Few companies have access to large amounts of real-world data to train agentic AI for our industries. Autodesk does. Second, context. Autodesk operates at the intersection of digital and physical worlds.
This is one of the most complex real-world context in technology. When making inferences, agentic AI has to operate inside a live project where the correct answer depends on the design intent, current model state, regulations and standards, constraints, dependencies, permissions and approvals. Autodesk provides agents with this context. Decisions must be compliant, coordinated, traceable and reversible.
Autodesk is a system of record where authoritative project and product context lives and where changes are executed, check and recorded. This makes Autodesk the natural control point for agentic workflows. Autodesk AI can propose and enable humans to verify and safely commit. Few companies understand this complex industry context across every discipline in the process.
Autodesk does. And third, expertise. Specialized data and context are prerequisites, but so is specialized AI expertise. For almost a decade, Autodesk has been building a world-class AI team for 3D design, make and operate and cultivating a broad external ecosystem to support it.
Over that time, we have built a strong reputation for having access to the right data, undertaking cutting-edge research and solving the most complex problems in 3D AI. With those strong foundations reinforced by Autodesk's unique purpose and culture, we have been able to attract and retain top talent and develop our own 2D and 3D multimodal models that understand how the world is designed and made. Few companies have been building their 3D AI capabilities, talent pool and ecosystem for almost a decade. Even fewer have sufficient breadth and depth of 3D AI capabilities and expertise to build on.
Autodesk has and does. Data and context fuel the knowledge graph, which is foundational to any artificial intelligence. Data scarcity and context complexity makes the knowledge graph hard to replicate for our industries. Even with data and context, you need sufficient specialized expertise to generate unique and valuable intellectual property.
And then you need an ecosystem of partners built around that intellectual property, as you saw recently with our investment in World Labs. Few companies have all this. Autodesk does. Let's move on to scaling and monetizing agentic AI.
In preparation for the cloud and AI, Autodesk modernized its platform and go-to-market over the last few years, well ahead of industry peers. Few industry peers are ready for the business models and go-to-market motions that will monetize their AI-driven future. Autodesk is. We built a platform that provides the identity, permissions, geometry kernels, data models and compute infrastructure needed to deploy AI safely and at scale into design, engineering, manufacturing and construction environments.
Platforms enable safety, innovation and efficiency at scale. Autodesk has built Autodesk Platform Services, APS, as an open platform that is purpose-built for an agentic AI world. It means we can ingest and process data more efficiently, accelerate our agentic AI innovation and deploy agentic AI solutions at scale. Few companies have built infrastructure specifically for our industries.
Autodesk has. To summarize, building agentic AI for design and make requires data, context and talent. Scaling and monetizing it requires a platform and next-generation business models in go-to-market. Few companies have all these advantages.
Autodesk does. It is not a coincidence. We have been preparing for and working towards the cloud and AI for more than a decade. While some new entrants have some of the required capabilities, they lack the data and context needed to deliver value.
At Autodesk University and last year's Investor Day, we demonstrated how we're defining the AI revolution for our industries, empowering customers with new task, workflow and system automations and capturing shared value through subscription, consumption and outcome-based business models that blend human and machine capabilities. The pace of our innovation continues, and we have much more to share with you in the coming months. I've never been more confident in the long-term value we are creating for our customers, for the industries that shape the world and for you, our shareholders. Operator, we would now like to open the call up for questions.
[Operator Instructions] Our first question comes from the line of Saket Kalia of Barclays.
Andrew, maybe just to start with you. I want to zoom into your AI comments a little bit because I thought they were super interesting. When you think back to Investor Day last year, I think you talked about Autodesk's path to AI monetization, and today, you're talking about sort of the AI competitive moats. Could we maybe talk about where Autodesk fits into the broader AI ecosystem?
And maybe specifically, how do you sort of see your relationship with the LLMs evolving over time? And how do those competitive moats maybe help balance that relationship? Sorry, there's a lot there, but does that make sense?
Yes. That makes sense. Thank you for that question, Saket. It's a great question.
Look, at a high level, it's not our goal to compete with the core capabilities of what the frontier models are good at. What our goal is, is to ensure that the combination of what the frontier models do, what an LLM does and what our proprietary foundation models do is always better than what a frontier model can do alone. And the reason we have so much conviction about that is really kind of the things I highlighted in the opening commentary, but I'll kind of go through it again a little bit here, right? It's the data, the context and the expertise.
We are sitting on volumes, large volumes of data about real-world problems, real-world situations, real-world constraints that is simply very scarce and very hard to get access to. When you combine that with the deep context knowledge we have around design and engineering, into preconstruction planning and construction, into manufacturing and all the things that go into making something, you get this strong combination between data and context that's very difficult to replicate. And that's going to allow us to always kind of stay in front of what is going on in the horizontal and the base foundation models.
And that's really our goal. Most companies in our industry are really going to struggle to do that because they don't have the volume of real-world data, they don't have the deep design and make context. We do, and we continue to use that and will continue to use that to stay ahead.
Janesh, maybe for my follow-up for you. It was helpful sort of how you compare the guidance from -- for this year compared to last year at this time. Could we maybe talk a little bit about the absolute levels of prudence that you've incorporated into the FY '27 guide maybe compared to last year? And how you've sort of thought about that prudence over the year?
Saket, I'm happy to talk about that. And maybe I'll just start by saying that the underlying momentum of the business remains really strong, and we see the demand drivers from fiscal '26 continuing this year here in fiscal '27. In terms of the guidance, as I mentioned in the opening remarks, like last year, we've reflected prudence in the guidance to reflect the temporary risk that we see to billings and revenue related to the sales optimization plan. But unlike last year, we have not reflected additional prudence for a new CRO or CFO. On the modeling and how that plays out over the year, we've assumed that there will be a short-term disruption in the early part of the year to billings growth from the sales restructure and then that assumption flows through rapidly to the underlying revenue growth later in the year.
If I think about where those potential impacts might be, it will likely be out on new product subscriptions because on renewal billings, which is, as you know, the largest part of the business and then also on self-serve and EBAs, we expect those to be relatively unaffected. But absent that temporary disruption on new product subscriptions, we expect the underlying customer demand to remain strong throughout the year.
Our next question comes from the line of Jay Vleeschhouwer of Griffin Securities.
Andrew, for you first. There are obviously many ingredients to your product-led growth, and you spoke about that, and it's certainly baked into your billings guidance. But I'd like to ask about two parts of that, one quite old, one very new. The older part is how you're thinking about the relative positioning of and development of Forma versus Revit?
And the newer part is how you're thinking about what seems to be multiple opportunities to connect the World Labs technology into various parts of Autodesk, your data models, Tandem, Forma, et cetera, how you're thinking about that. And then secondly, as a follow-up, it's sad but true that you're going to be discontinuing the disclosure of direct and indirect percentages. So maybe take the opportunity to talk about your thinking of the role of the channel, the opportunities and priorities that you're setting for the channel from here and then perhaps also for the Autodesk Store.
That sounded like 3 questions, but I'll treat it as 2. All right. First off, let's talk about the relative trajectory of Forma versus Revit. As you know, our industries evolve over time. Even with rapid technological changes, projects go on for months to years.
There's a lot of complexity, older projects over their lifetime often go back on previous releases and all the things associated with that. So there's a rate and pace of technology absorption that's kind of unique to our industry. So when we look at the trajectory of Forma and Revit, there's a couple of things that are really important. One, Forma and the whole stack around Forma, from design to make, is going to be very much focused on the cloud and AI-enabled tools.
Everyone is going to have access to the workflow tools, the agentic layer, that's the Autodesk Assistant, but some of these deep kind of model -- foundation model-driven workflows are going to be built into Forma and already are. Revit is going to benefit from all the workflow enhancements associated with, like I said, the agentic layer of the Autodesk Assistant, but it's also going to be tightly coupled to the workflow with Forma. And that's always part and parcel of how we think about things. We want to bridge the 2 worlds for our customers, so that as they go through their technical transition over many years likely, they have a clean path between these products and these products work together in a clean and powerful way throughout that whole transition.
And it's very important to us, but look for a lot of the model-based agentic features to show up in Forma, the assistant-based workflow agentic layer will absolutely work across Forma and Revit. Now when you talk about World Labs, look, we're very excited about that investment. And we're excited to work with a deep technology company that's focused on something that we feel is really important. World models are important for physical AI because of their ability to spatially reason, reason about physics and also respond to real-world changes because of their awareness of what's going on in 3D.
We see this as a fun foundational kind of horizontal technology that will power lots of solutions. It's starting in worlds that are associated with games and media entertainment, but there's so much more there, Jay. It will go deeper into initial architecture design, it will end up going into areas associated with digital twins, with factory automation, robotics, all of these things associated with that. So we're partnering them to bring that technology, first, into the media and entertainment space and kind of create workflows between Marble and our tools.
But over time, look for us to engage more deeply with World Labs and connect their technology with our technology in all sorts of interesting ways, just like we do with the large language models today. And the last one, okay, sorry, that's right, a 3-part question. The direct, indirect piece, okay? So yes, you're right.
We're not going to disclose that split because most everything is coming in direct to Autodesk now. It's an agency model. We're capturing things directly, and I think it's important to recognize that. So in terms of the channel incentives and the things that we're doing to change things, we're doing the exact same thing with the channel that we're doing with us.
We're focusing the channel on new business creation, going after new accounts, going after expansion in existing accounts. Our channel is compensated that way now. Renewals are compensated lower. New business is compensated higher.
Our sales force is compensated that way. So we have tight alignment on new business growth, both from new accounts and expansion between our sales force and our channel. And as a result, that will drive more of these numbers higher as time goes on.
Our next question is from the line of Adam Borg from Stifel.
Andrew, obviously, you hit on AI in the prepared remarks and kind of the moats that you perceive Autodesk to have. And of course, AI has been on all of our software minds of late. But I'd like to just take a step back and when you speak to customers, where exactly are they in their AI journey? And what exactly is that they want Autodesk to help them with on this front?
Yes. Look, the customers are -- especially in the GC and engineering community, they're exploring AI pretty aggressively right now, trying to understand what it can do and how it can help them. Absolutely, they want to see the complexity and time of creating a model to be reduced over time. But really, one of the things we're working on very closely with them is wrangling data and bringing data together in intelligent ways so that they can actually get insights and action things in an agentic way on top of complex data flows.
And that's one of the areas we're engaged with a lot of customers. That's why you see engagement on platform services and people extending their environments and building, frankly, on top of our -- building very complex life cycle workflows on top of our APIs in our environment. So that's an area of strong engagement right now.
And then maybe as a quick follow-up for Janesh. Back at the Analyst Day, we talked about consumption mix of total revenue, I think, in fiscal '25, that was 17%. Any update you have for fiscal '26 and how we should think about this consumption-based mix over the course of the year?
Yes, Adam, I'd say it was about similar. And just as a reminder, when we talked about this at Analyst Day, we said that EBAs were roughly 15% and the pure usage based, which is largely Flex, that was roughly 2%. So in the aggregate, it was about 17%, and that was for fiscal '25 and fiscal '26, I think, was roughly similar. That's what we saw here in the year.
Our next question comes from the line of Joshua Tilton of Wolfe Research.
Huge congrats on a very strong end to the year. I have 2 questions. I'll ask them at once. My first question is, obviously, you can't help but notice that the revenue growth guidance for this year is starting at a higher point than you started the guidance for last year.
Could you maybe walk us through some of the puts and takes for revenue growth maybe actually finishing the year above what you grew revenue last year? And then maybe my follow-up to that is in regards to Saket's question, when we weigh all the puts and takes that you discussed around the conservatism in the guidance, we put an equal sign after that. Does that equal guidance that is more conservative this year than last year, less conservative, similar conservatism? Just what's the answer to the question of what do all the puts and takes mean for conservatism this year versus last year?
Josh, this is Janesh. I'll give you a single answer to both of those questions, which is, overall, when I step back and think about fiscal '26, we were very pleased with how the business performed. And you saw that across the quarters. That strength reflected the broad momentum that we've got and strong execution across the entire portfolio.
The current year guide primarily reflects the prudence for the near-term go-to-market optimization. And that impacts billings in the early part of the year with the flow-through to revenue over time. Ultimately, remember that the new transaction model and the go-to-market optimization are really both designed to improve our long-term new business capture. That's been the core thesis that we've shared before.
We remain confident in it for the long term, but we are staying disciplined about what we assume in the near term.
Our next question comes from the line of Jason Celino of KeyBanc Capital Markets.
I have 2 questions. Maybe the first one for Andrew. It's an AI question, sorry, but it's not about competition or moat. But maybe a scenario in which AI actually works in architects and civil engineers become efficient and so efficient that these customers don't need to grow headcount, how much has the industry grown head count historically?
And if that's able to be applied to Autodesk growth? And then what happens in a scenario where the AI efficiency is what we kind of think it might be and how might that affect your future growth opportunities, if that makes sense?
Yes. No, that absolutely makes sense, Jason. So first off, I just want to make sure that you understand, in our industry, we have a fundamental capacity problem. There is not enough capacity in the ecosystems that we serve to build everything that needs to be built and rebuild everything that needs to be rebuilt.
When capacity is sucked up in one area, it takes away from other areas. So remember, there's this underlying capacity problem, not enough money, people, materials to build and rebuild everything. When you look at the way we're moving forward, we absolutely want fewer people per project because we want our customers executing more projects. There's plenty of demand for projects out there.
So at a kind of a task-based automation level, right, the kinds of things that you're seeing us do right now around speeding up modeling activity and things like that, that's kind of improving the core value of a seat of software. And we don't expect seats to go away anytime soon. There will be a solid core of seats, but the task-based automation is going to add to the value of that seat. That seat is going to get more valuable as we enable one person to execute on more aspects of a project.
So again, it's fewer people for projects, more projects executed at the task-based level and at a seat-based level. The important thing to recognize -- wait, there's something else here, okay? I wanted to talk about the workflow automation a little bit, Jason, because as we move into workflow automation, basically with the agentic layer of the Autodesk Assistant, we're actually monetizing the project now, not just the task that individuals are doing, but the whole disciplines across the project. As you know, we already deliver project-based pricing around construction.
We deliver site-based pricing and consumption-based pricing. We're going to monetize more of that project activity through consumption as we reduce the number of people working per project will monetize other aspects of the cross-disciplinary nature over the project. And that's important to recognize because that expands our TAM. And the last piece, and if you have a follow-up, that's fine, is around the systems automation.
When you get to the level of systems automation, you certainly help with the individual and the project, but you also get into the wallet of the person paying for the project, the owner. And that puts you in a position where you've actually expanded your TAM deeper into the actual kind of total spend on the project, the total spend on the product, what the end user or the owner or the operator wants to get out of that product. So multiple avenues for us to monetize agentic AI across that entire process from task workflow to system.
Interesting. Yes, maybe we can explore it in another instance, but yes, it seems quite incremental. And then my one follow-up for Janesh is quick. It sounds like Q4 benefited from some better linearity.
Maybe is it possible to understand like what happened where some deals from Q1 closed earlier in Q4. It's just when I look at the implied Q1 guide ex-currency, ex-model transition benefit, looks like it's deselling by 4 or 5 points. So curious if there was any details on that.
Yes. Jason, I'm happy to provide some color there. Q4 was a very strong quarter, as you noted, and we're very pleased with that momentum. In terms of the exit rate of 14%, there's 2 factors that I'd call out.
First off, for fiscal '27, recall that we've got this near-term impact to billings that has some impact on revenue, not only for the full year, but it has some impact on revenue growth in Q1 itself. So that's something to consider. And also, Q4 benefited from lapping an easier fourth quarter from the year ago period. So that's also something you need to adjust for when you look at that 14%.
But overall, when I look at our momentum, and I look at the underlying strength of the business and the improvements we've made in the go-to-market model, we feel very good about where the business is today.
Our next question comes from the line of Bhavin Shah of Deutsche Bank.
I have 2 as well. I guess, first, either Andrew or Janesh. I think you guys have been pretty clear about the continued need to optimize the sales organization and changing up the incentives with the partners. And I know you talked this an account with your guide.
But maybe can you talk operationally, what are you guys doing to help mitigate any impact that might have any disruption there? How do you make sure that renewal activity is healthy as you kind of incentivize new business?
Yes, I'm happy to take that. We -- when we made the changes to the partner compensation plans for fiscal '27, that was for the partners that are operating in the new transaction model. And just as a reminder, this has all been part of our broader plan around incentivizing partners to focus on more new business, and it's really been a core element of our overall new transactional model thesis. And so the change that we made was to increase the incentives on new business and reduce the incentives on renewals starting fiscal '27 with the goal of keeping the total dollars unchanged.
And as part of that, we did -- we were aware that there might be opportunity for people to think about the timing of transactions, and we did put operational guardrails in place with the partners to try and avoid that kind of activity. And so we actually saw those work quite effectively. There are some early renewals that happen every year. But here in Q4, those were not remarkably different than what we typically see.
They were not a significant contributor to the outperformance in the quarter. And also, as you know, early renewals don't even impact revenue. So we really didn't see any impact from pull forwards or early renewals associated with that activity.
Got it. And maybe, Andrew, look, as you have a conversation with customers this quarter, particularly your larger EBA customers that are looking to expense Autodesk for multiple years. What are the types of things that your customers are asking you to help solve that might be different than what they were maybe a year or 2 years ago? And how is that helping inform your product road map into the future?
Yes. So look, as we've moved more across design and make and deeper into each aspect of those, customers are really kind of asking us for what we're classically calling convergence. They want us to kind of stitch the glue together between those things. They want fast feedback between a design decision and a make decision, between a make decision and a design decision and they want some kind of agentic layer that helps them sort through the noise and helps them get to what's right, what's wrong and how do we quickly focus on the thing that needs to be changed so that they reduce downstream risk later.
That's the kind of things we're digging into. That's why people are investing more and more in what we're doing because of the way we've connected design and make together across the whole life cycle.
Our next question comes from the line of Joe Vruwink of Baird.
I asked a long-winded question, so I might have ask 2 separately, but I wanted to start with Autodesk Platform Services. It's certainly being put to good use now within how a lot of customers are thinking about AI projects. And I specifically wanted to ask about the new monetization strategy you've launched around customers opting into their intended API use. When you think about just how those bundles are being priced, are they priced initially to drive adoption, so we might not see like a discrete uplift factor you're calling out in FY '27. But if you do your job that might very well contribute to growth in, let's say, FY '28?
So API monetization is very much specifically targeting machine usage of our IP. So someone that's doing a lot of 24/7 compute and access of some of our more complex APIs. We actually have customers that do that. Remember, when they call an API, they're not just pulling a piece of data, they're actually calling some functionality that sits on our cloud.
And some of our customers have been doing fairly sophisticated things running these tools over and over again through large periods of time. So we're targeting monetizing that machine usage. And frankly, in Q4, we already had some customers lean in and pay up on the consumption model around API usage that was driving a lot of value inside their accounts. So positive early signs on that.
But remember, there's not a lot of customers doing that yet. We're just getting ready for that future where people are driving a lot of machine usage. We don't want to get in the way of the people running regular usage, adopting, integrating and using our APIs in ways that are not agentic or machine-driven. So that's where we're at right now, and we're already seeing customers kind of engaging with us on some of these areas of deep machine execution.
Okay. And that kind of gets to my second question but a lot of your larger customers, they already have in-house development teams with Autodesk Centers of Excellence. I'm just thinking about the AI risk narrative that customers can go and build whatever they want. Well, you kind of allowed them to build a lot of what they want, but it's all complementary to Autodesk.
Does that change anytime soon? Or are you actually seeing this whole strategy drive usage into your kind of mainstay products. And so it ends up kind of lifting all boats, if you will.
Yes, it's absolutely the latter. What we're seeing is the customers are leaning into using these APIs to actually build solutions that they might have used behind the firewall solutions for in the past. And that's true in both the AEC base and the manufacturing base, kind of old behind the firewall, kind of rigid, inflexible implementation, they're kind of dying. I like to call them my next dead thing working.
I used to talk about files being dead thing, working now. Now behind the firewall implementations and complex rigid systems with dead things working. And I think our customers are exploring our APIs and our IP in really elastic ways to try to build solutions on top of what we've built so that they can kind of do things that they used to bring in large vendors to do with complex back-office implementations on our stack, which is lifting the rest of our stack up with it.
Our next question comes from the line of Ken Wong of Oppenheimer & Company.
Hoi-Fung Wong: Andrew, can you talk about the go-to-market optimization that you're putting in place in '27, the commission changes, the direct sales restructuring, like how much of these actions were originally in the blueprint? How much of it is because observations from '26, whether it was execution or the really strong results help inform these particular changes?
Yes, Ken, for the most part, these were part of the original blueprint, right? We knew that we had a series of goals we wanted to accomplish in terms of driving go-to-market efficiency. We knew that we were highly focused on renewal optimization, renewal automation, reducing overlap of resources on renewal activities so that we could more efficiently bring in the renewal dollars, and we knew we wanted to shift money to new business development and new business acquisition. So these were really basically baked into our thinking from the get-go in terms of how we were phasing out our go-to-market optimization.
Hoi-Fung Wong: And then, Janesh, just a follow up on that. In terms of the 7% reduction in force, how should we think about the reallocation of those resources? I'd assume a few people have done math and would have assumed a little bit more on the margin profitability side, but it looks like it's reinvested in the business?
Yes, Ken, we guided to 75 basis points of improvement year-over-year, and that's on the back of very strong outperformance in fiscal '26. That reflects both the underlying operating leverage and also the savings from the restructuring. And as you noted and as we mentioned when we announced the restructuring, we are making planned investments back into the business as well. And then also, as a reminder, we've got roughly 1 additional point of headwind from the new transaction model. So that's something else to factor into.
And then finally, I'll just point out on the overall operating margin is that we've also reflected the prudence that we embedded on the revenue guide. We've reflected that as it flows through to the operating margin as well. So that's something else to factor in.
Our next question comes from the line of Alexei Gogolev of JPMorgan.
Andrew, in some of the recent comments that you made, you were talking about the upside from data centers. I was wondering if you could talk a bit more where you think this upside could come from. Are the design centers increasing seats or are there new products being sold, new design teams being built? Because my thinking is that some of those projects are quite penetrated among the customer base and all these well-known architect teams, they probably already have Autodesk.
So where would this upside come from?
Yes. The data center projects are driven by very sophisticated owners. So a lot of what's happening is that the owners are buying more of our suite to manage the design and the execution of these projects. And that's actually getting deeper into the supply chain.
We expect the demand for data centers and tools for data centers to continue into several years from now. But remember, when that demand shifts, it opens up capacity for other types of projects that come in, like there's lots of infrastructure projects in the U.S. that I guarantee you are being stalled by the fact that capacity is being consumed by data center work right now. But the owners are driving a lot of the adoption of technology, and they're looking at the design through make cycle. Like I said, they're very sophisticated operators, and they buy very sophisticated tools and they go deep into the process across our entire stack.
Janesh, very quick question on free cash flow. So you mentioned that there will be a restructuring charge, partially offset by cash tax benefit from OBBBA. So these net effects of the cash movements, they seem to be somewhat immaterial for fiscal '27. But does this mean we should see a step up in free cash flow margin and free cash flow conversion in fiscal '28?
One of the important things to keep in mind about '28, Alexei, is that our federal tax payments will begin to normalize from fiscal '28. So you just need to factor that in discretely.
Our next question comes from the line of Tyler Radke of Citi.
So just a follow-up on sort of the underlying growth. So obviously, there was some upfront benefits and billings linearity benefits exiting this year. But as we think about that drop-off implied in the guide, can you just help us understand how much is sort of driven by the go-to-market changes in that restructuring? And I guess, specifically, like what are the biggest risks this year that you didn't maybe necessarily see pan out in Phase 1 of the restructuring?
And then are you thinking about sort of the data center contribution similarly for this year? Or should that be an incremental tailwind just given we're all seeing CapEx numbers go up into the right?
Yes. Tyler, this is Janesh. Maybe I'll take that. So a couple of thoughts on the 14% for Q4 and comparing that to the Q1 growth rate.
As I mentioned earlier, there's 2 things to keep in mind. One is that the 14% laps a really easy fourth quarter from last year. So that's just something that you need to account for. It's naturally a higher number than you saw in Q1, Q2 and Q3 of last year.
And then if I -- the other factor is that there is an impact to revenue growth in Q1 from the go-to-market changes that we have assumed. The reason that -- or the difference between this restructure and the prior one is that the earlier restructure that we had about a year or so ago, that focused primarily on non-selling roles, and most of those were in marketing and customer success and sales operations. In this restructuring, as we mentioned, when we announced it, the majority of the impact is actually in customer-facing sales roles. So when you have that, there's always a natural level of disruption that happens associated with that, and that's just something that we needed to discretely call out and consider here in the guidance.
And then I'm sorry, what was the second part of your question?
Yes. It was just around the -- obviously, the data center as an end market has been strong. Like how are you thinking about that for next year? Is that going to be a sort of a greater percentage of your end market or business?
Yes. It's -- we've been very happy with the momentum we've seen in data centers for a few quarters now. But as Andrew said just a couple of minutes ago, for us, the way we think about this is, our industry is fundamentally faced with capacity and productivity challenges. And when demand shifts to one part of the industry, if for any reason that demand erodes, then the capacity shifts elsewhere.
So we think of it in the aggregate and think of it as a shift more than anything else. Data center, the business has been great. We'll continue to see how that contributes to the business in fiscal '27. We saw some strong business close here in fiscal '26 by way of demand from -- for cloud from customers that are building out these data centers. And then longer term, that ultimately translates into an opportunity even for operations, which we've talked about externally before.
And then just a quick follow-up for Andrew. Post the World Labs investment, are there sort of other partnerships and investments that we should be expecting you to make? Or do you sort of feel like that sort of has you covered at least for now, obviously, a rapidly evolving market?
Yes. Look, it is a rapidly evolving market. In terms of big partners that we'll be engaging with, World Labs is a pretty significant one. But look for us to lean into the operations space in an interesting way, very similar to what we did in construction following a very similar playbook so that we can kind of move quickly and start to turn that opportunity into even a greater life cycle opportunity for Autodesk moving from like months to years of engagement on projects to decades.
Our next question comes from the line of Michael Turrin of Wells Fargo.
Just one for me, apologies for being on mute. Just the manufacturing segment accelerated to now above 20% growth. Can you speak to what's driving the strength there? And I'm curious if it's all correlated with some of the emerging markets commentary that you're making? Or just any other just kind of supporting details you can provide there?
Yes. Overall, what I'd say is manufacturing continued to perform well for us. The make business in the aggregate performed well. You saw that with 23% growth, which includes construction and Fusion.
Construction revenue actually accelerated during the quarter. Some of the trend -- the underlying trends that we've seen earlier in the year by way of the strengthened demand, those continued. We continue to focus on our strategy of continuing to drive multi-seat adoption in manufacturing accounts. And then we also continue to see strong adoption of some of the features like we talked about the AutoConstrain feature and provided some statistics on that.
So we're seeing great adoption there, too. Those are some of the things I would call out.
Thank you. And ladies and gentlemen, that is all the time we have for Q&A today. I would now like to turn the conference back to Simon Mays-Smith for closing remarks. Sir?
Simon Mays-Smith: Thank you, Latif, and thanks, everyone, for coming along. We look forward to seeing many of you over the coming weeks. If you have questions, please just ping me or the Investor Relations team, and we look forward to catching up again next quarter. Thanks so much.
This concludes today's conference call. Thank you for participating. You may now disconnect.