Let me just talk a bit, a few things this morning. First one is we're going to primarily talk about our two largest markets and the two biggest opportunities. We're going to talk a lot about AEC, architecture, engineering, and construction, and we'll talk a fair about manufacturing. We've upgraded the size of the TAM as we've calculated it. The other thing I want to make sure that everybody understands, most of the day, we will spend reiterating what we talked about last year, which we think of as 12, 20, and 50. In particular, the way to think about it in this equation is that what we're really trying to drive is the 20 and 50. We'll spend a lot of time talking about how to get more value from customers and how to get more subscribers.
That's what we'll be talking about most of the day. If we are able to achieve the 20 and the 50, the 12 is a natural outgrowth of that. It's easily accomplished by getting the other two. We are going to spend most of the day on the nuts and bolts of the 12, 20, and 50. Before we go down the rabbit hole here and spend lots of time there, I just wanted to back up and spend just a drop of time discussing what I think is really, to some degree, more important. At the end of the day, our ability to do 12, 20, 50 really depends on many of the strategic decisions we've made and many of the strategic decisions that we made over the past few years.
These are big things, and in many ways, I think they distinguish us from our competition. After all, strategy is how do you compete more effectively? Let me outline what these big choices that we made are, and that's really where our big bets are. The nuts and bolts of the 12, 20, 50 follow from these. The first thing is, about three to four years ago, we decided that there was a dramatic shift in the technology platform. The computing platform that would be used for design, engineering, and architecture was changing. It's changing computing platform was as big a change as any we've seen in our lifetime.
For those of you who remember, as we went from the mainframe to the workstation or the workstation to the PC, in some ways it was unimaginable that the new technology would take over and come to dominate. Does anyone know any engineering software running on mainframes today? Not that much. Not much even running on workstations anymore. In the same way, I believe all of the engineering software of the future will run on a cloud and mobile platform. We've made a heavy bet here. It is different than what I see our competitors believing and what they're acting on, and it's a really crucial part of how our business grows in the future, is this bet on cloud and mobile. The second big strategic choice we made is we've decided to have an open approach to the markets.
Many of our competitors have long felt and acted upon the belief that they were somehow going to convince customers to use a single stack of engineering software. That all of their engineering software would come from a single vendor. While in the short term, you occasionally see customers who consolidate around a single vendor, I've been in this business way too long to believe that that's a lasting trend. Over time, best of breed, best solutions, heterogeneous environments triumph. It is in the customer's best interest to do that. While many of our competitors spend a lot of time trying to figure out how to lock in customers through file formats or pricing mechanisms or a million other things, we've decided and acted upon the belief that allowing our customers to choose the best answers for their problems are there.
You'll see this in many of the things we do. You'll see that whether it's with our PLM products or our CAM products, many times the design system of choice may not be Autodesk. Given a choice of as they select a PLM system, would I rather be ours or not, regardless of what CAD system they use, I want that to be ours. I think we are distinguishing ourselves and gaining market share as a result of this strategic choice. The third thing that's also important is which markets you choose to compete in. It's a very important first-order effect is where you decide to compete and where you decide not to compete. We'll talk a lot about not only where we're competing today, but where we've put bets for the future and where we will compete going forward.
The last part of the differentiated strategy is a business model change. We've talked a lot about the business model. We are going to spend a lot of time on the business model today. We will go deep into the nuts and bolts. As I could tell by the conversation around the bagels, there's a lot of interest in subscriptions and perpetual licenses, we'll go through all of that. These are really the four pillars. As you go through the day and you hear a bunch of the people talk, I ask you to reflect upon these four things and put them in the context of them, because I think they really lead the way to understanding where we're going and what we're doing. What follows from that are lots of the tactics around the programmatic stuff that we'll discuss in detail.
Let me just outline the agenda for you. Andrew's going to come up and he's going to talk about the business model transition. It'll be followed by Amar and Buzz taking turns talking about the AEC industry, the manufacturing industry, and then our new cloud-based collaborative platform, A360. You'll get an update from Steve about sales and then an update from Sue about finance. I'll come back and have a few closing comments, we'll do Q&A. Let me bring on, at this point, Andrew Anagnost.
Thank you, sir. All right. Let me grab a glass of water here. All right. I'm the nuts and bolts guy. Let's get into the nuts and bolts. All right. Like Carl said earlier, our fundamental commitment to the 12/20/50 model hasn't changed. I want to talk about a few details, though, and some language changes so you can appreciate some of the things we're evolving moving forward. The idea around what drives the 12, more value from existing customers, more subscriptions, is completely intact. What we're going to be doing moving forward, though, is we're going to quantify the 20% for you a little bit more precisely. Last year, we talked about value per subscriber. You're going to see us moving forward talking about value per account.
Not this year, but next year, just like we did with Suites, we're going to start tracking that number to a baseline. If you recall the Suites transition, what we did is every year we came out and we talked about how are we doing towards the 20% ASP uplift as part of the Suites initiative. With this initiative, we're going to start tracking moving forward how we're doing with the 20% growth per account. That's one of the changes I wanted to highlight in terms of the conversation here. The rest is completely the same as we were talking about last year. We'll move forward with that moving forward. Let's talk about how we're doing so far. All right. If you look at how we're doing so far, billings are up 18% H1 over H1 from the previous year.
Subscriptions are up 11% H1 to H1 from the previous year. We're well on track to achieve some of our objectives, and we're starting off really strong in the transition. What I want to do is I want to give you some sense about how we're breaking some trends and some of the important metrics that are not only driving the business this year, but are going to drive the business forward year after year after year. First off, let's look at our historical subscription additions trend. What you can see is we're actually breaking the old trends around subscription additions. We're doing more this year than we've ever done before. The momentum you're seeing in subscription additions this year is going to continue into next year and the year beyond.
This presentation and the presentations as follows are going to tell you why that's going to be true. There's lots of confidence in this, and there's lots of evidence we're going to be maintaining that momentum moving forward. We're breaking from the trend. If you start looking at some of these numbers and some of the underlying fundamentals here, the maintenance subscription growth's been 7% year-over-year. You know that's off a large base. This is driving a lot of those subscription additions right now, and that's great. We're seeing a lot of strength in our core perpetual plus maintenance subscription business. We're also seeing some strength in some core areas around the fundamentals of this business. For those of you that track how these models evolve over time, you know renewal rates are important.
One of the things we look at pretty closely is the renewal rates around things like AutoCAD LT, where we have lots of sensitivity. Right now, year-over-year, we've driven that renewal rate up 5 points. I'll point a little bit more later in the presentation why this is so significant. What you get from this is that we have a lot of strength in that perpetual plus maintenance subscription business. Most of that's driven by upgrades going away. There's a whole bunch of other things that are happening in our business, much more forward-looking things that have absolutely nothing to do with upgrades going away. Let's take a look at some of those. First off, it's the desktop subscription growth. We rolled out desktop subscription in Q3 of last year, and it has continued to accelerate since we did that.
This acceleration is not slowing down at this point. We're driving more and more momentum here as we light up new products into the desktop subscription stream, and as we light up new channels. This is going incredibly well. It's also reaching out to 35% new customers. The customers that are coming in here, 35% of them are new. Why is it? It's pretty simple. No one's been able to get access to this kind of high-quality price point software at these kind of price points before. They're responding to the shift, and they're buying. Another thing I want you to understand about the dynamic here, because it helps you understand how the momentum's going to grow moving forward, is right now, currently today, 51% of those desktop subscriptions are coming from our eStore. Why is that number important?
I think if you know our whole business, you know our whole business, 51% of it does not come from our eStore. What this means is we're seeing this momentum and this acceleration, and we haven't even begun to light up the partner network with this offering yet. We've only started that journey. We're just in the very early stages of it, but we're showing huge momentum here. If you look at the other types of new offerings, let's take a look at cloud. We've doubled the number of cloud subscribers. This has nothing to do with upgrades going away. Nothing. All right. This is a significant momentum. You're going to hear a lot about this throughout the day, how we're doing with some of these applications. There's significant growth. Again, when you look at these customers, 30% of them are new to Autodesk.
They've never bought anything from Autodesk before. 30%. This isn't non-subscribers. These are net new customers to the company. We're not only going out there growing subscribers, we're bringing new people into the fold, into the ecosystem. Last year, you might recall I talked a lot about enterprise business agreements and the Token Flex model. The enterprise business agreements drive what we call consumption models into our large accounts. If you don't remember what these are, the consumption models offer our customers access to the entire portfolio, and they pay for what they use. They use more applications, they pay us more. They add more users, they pay us more. We've seen a 37% growth year-over-year in the subscriber base of these enterprise business agreements. That's a very significant metric because that's going to be a big driver of that 20% value equation over time.
Lots of strength. Lots of momentum. We're well on track to achieving our goals for this year, and we're planting the seeds to achieve our goals for the next year and the year beyond. What I'm going to do now is I'm going to talk to you about some of the things we're going to do both this year programmatically, next year programmatically, and what we're going to do continuously. I'm going to start with some of the things you'll see us do continuously year after year after year during this transition. The first thing we're going to do continuously is pay attention to this number. I told you last year that the number of customers that are one to five releases back on our software stays relatively consistent year after year after year.
We are going to exit this year with that being no different than previous years. All the execution we're doing, all the subscribers we're adding, we're still going to have 2.9 million non-subscribers in our base. That is a huge opportunity, and it's important for us to pause and talk about what are some of the attributes of these people. Why are these people so interesting? One, they are real customers. These are not pirates. Pirates are coming in from other places. We're getting pirates in with some of these new offerings, but these are real customers. These are people that pay us. They just happen to purchase from us perpetual licenses on an infrequent basis because they can. This really isn't really good for our ecosystem. This creates a large ecosystem of people off non-current releases.
We would like to see everybody on the most current release in our entire ecosystem. We think that's good for customers. We think it's good for them. The other little piece I want you to take away from this is 45% of them are LT customers. We're going to go back to that a little bit more. Over and over again throughout this transition, we're going to be looking at that base and looking at that number. Another thing we're going to be looking at is moving the business north of 90% renewal rates. Like I said, if you've modeled this business, you know how important it is to retain the subscribers. It's not just important to bring them in, it's important to keep them. We're going to continue to build beyond 90% renewal rates.
The next thing we're going to do, you're going to hear a lot about this from Amar later and others, is we're going to add cloud to just about every account. Every account is going to be purchasing some kind of cloud application from us, and this is going to be a huge driver of subscriber growth. We're also going to be using the cloud to reach out and capture a whole new set of subscribers that Autodesk doesn't currently engage with. The last piece I want to talk to you about in terms of the continuous activities that don't go away is this increased penetration in named accounts with these enterprise business agreements. We're going to do more and more in these accounts, and I'll talk a little bit about in this presentation.
Those things I just outlined for you, those are going to happen every single year. They're just going to be special changes in terms of how we address them. Let's talk about some of the special changes we're going to be looking at as we move into next year in terms of maintaining and driving the momentum for the business model transition. First, let's talk about leading with the new offerings. Next year, what we're going to do, all of our go-to-market, all of our marketing activities are going to be focused on the new offerings, in particular, the desktop subscription offering. We not only believe this is the right thing for our business, we also believe it's the right thing for our customers.
Not only does this provide the lowest cost access to the products that customers have ever seen, it also provides a superior experience. The download experience, the license experience, the deployment experience, the use experience, all of the experiences are superior in these new business models than they are in our existing business model. Because of this, because we believe in this so much and because we're going to focus so much on this, what we're going to do is over the next 12 to 24 months, we're going to stop selling new perpetual licenses. I want to be very clear here. Over the next 12 to 24 months, this isn't going to be an event. What this is going to be is a transition where we work on products and regions at removing new perpetual licenses from the offering mix. That's an important step.
It's going to drive a lot of activity, we're going to put a particular initial focus on AutoCAD LT as we begin this transition. Why AutoCAD LT? Let's talk about why AutoCAD LT. A couple of really important reasons. 48% of the new seats that come into the company currently are AutoCAD LT. That's 48% of our new seats buying the wrong business model. We concentrate on this one product, this one piece of distribution, and this one area, we're going to move 48% of the new seats over to the new business model right out of the gate. The other important stat I want you to pay attention to is that 76%, 76% of these LT users are non-subscribers. What that means is there's a huge opportunity to increase the value of this LT buying stream for Autodesk.
There's a huge opportunity to go out there, turn these people into having a recurring relationship with Autodesk. Why is that so important? Let's just look at how the value of a current LT subscriber looks relative to an LT subscriber in the new business model. The number on the left, that is the actual average value, annual value of an LT customer today based on all the behavior that our LT customers do. Some upgrade in three years, some upgrade in five years, some are in maintenance, but on average, they're worth $240 a year. You move them over to the new desktop subscription model, they're worth on average $310 a year. That is a 30% increase in value back to Autodesk with an offering that provides access to the customer at prices they've never seen before.
Remember, we price the LT offerings at 30% of a perpetual new seat. That is a huge shift in value for a very important stream that we're going to be focusing on, and it really is going to up the value at Autodesk, and it's going to up the benefits of this whole customer ecosystem. That's something I want you to watch as we move into this year. The next thing we're going to do is we're going to increase access to the new offerings. I've talked about some of the things we're doing to move away, target that non-subscriber base more aggressively. Now what are we going to do to provide broadening access to these new offerings so that we light up all these channels?
One of the things we're going to do is we're going to take these consumption-based models that we've been working on and move from just executing on dozens of accounts this year to hundreds of accounts next year. You're going to hear a lot from Steve's team about what he's doing to increase the focus here, build closer relationships with our customers around these enterprise business agreements and these consumption models. These are going to long-term drive that 20% in really significant ways. The other thing we're doing is looking at both our traditional VAR channel and our volume channel partners and getting them more access to these new types of business models. First off, next year, all of our channel partners are going to have access to monthly and annual subscriptions for desktop software.
Right now, they only have access to quarterly and annual, and the monthly is a significant feeder for some of the annuals. We're also broadening our engagement and access with some of our high-volume partners in terms of selling desktop subscriptions. Just last month, we lit up Amazon for selling desktop subscription. They're now out there selling monthly and annual desktop subscriptions for us, primarily around LT. They're doing a great job, and we're going to light up other partners that work in volume, CDW and Dell. This is going to allow us to execute even more effectively at the top and bottom of our market. Amazon, other volume partners like that are going to help us capture the bottom. CDW and Dell are going to fulfill at the top.
We're going to be broadening a lot of access to these offerings and lighting up more of that channel, which is going to create a lot of momentum, especially when combined with the transition away from perpetual purchases. As we do this, you're going to hear from Steve about how we're shifting the margins away from perpetual offerings to desktop subscription offerings. It's going to be attractive for our partners to sell these offerings, not just long-term, which we already know it will be attractive for them, but in the short term as well as they transition their businesses over. A pretty big shift in how we're providing access to some of these applications and some of these tools. The next thing I want to talk about is what we're going to do to increase the value of these offerings.
Not only are we going out there with a cessation of perpetual purchases in some areas over the next 12-24 months, but we're also going to increase what is already a highly valuable offer and make it even more valuable. One of the ways we're going to do this is simply add new types of things to desktop subscriptions. The first thing, and this is important because we think it's going to be a pretty powerful tool for connecting with that non-subscriber base, is we're going to offer multi-year subscriptions to desktop subscriptions, one-year, two-year, and three-year subscriptions. We'll promo some of these. We'll discount them. We'll target non-subscribers with some of these offerings, allow them to get on board, get engaged with us in a recurring relationship. We'll also offer more flexible ways to pay for this.
Buy three years, pay annually, new terms like that. We're going to introduce new terms. We're also going to have streamlined access built into these offerings, new ways of deploying, new ways of managing the users for these offerings. They're going to be superior to the experience we can deliver on the perpetual side because we're able to deliver a superior experience. We're also going to provide exclusive support offerings, things that help the customers get up to speed, help them get successful with the software that are exclusive to desktop subs. That's a nice direction we're going in terms of adding value to these applications. You're also going to see us introduce a lot more blended offerings that have no perpetual equivalent in our portfolio.
What I'm talking about is a net new offering that blends a desktop software on term and a cloud offering on term as well, and targets a particular customer need or a particular segment. I'll give you one example of something we're doing today. We actually already have a bundle of BIM 360 and Navisworks 360 that's completely termed, that targets a particular persona in the construction workflow and has no perpetual equivalent in our portfolio. You're going to see more of those. They're going to be targeted, and they're going to be high value. All right.
I've told you some of the things that we're doing on a recurring basis to drive the business model shift long term, and I've told you some of the specific programmatic actions we're going to be taking in FY 2016 to continue that momentum, see the same kind of subscriber growth next year that we're seeing this year. What I want to do now is kind of go back to those principles I talked about earlier and tell you how they're going to continue to change and evolve over time. The first thing, and you'll hear it again and again, is this doesn't go away. As we remove perpetual rights, this is going to fuel growth in terms of attacking and working with those 2 million non-subscribers, the 2.9 million non-subscribers.
As they come into the buying pool again, they're going to be buying a term offering from Autodesk, not a perpetual offering from Autodesk. That will provide significant long-term subscriber growth and significant opportunity for us to continue building up momentum. We're also going to be changing our culture around retention. We're going to be moving to a retention culture. We're already well on that journey. You're going to hear a lot of this from Steve, but we're building it around a couple of pillars. One, the way we engage with customers. Our whole method of touching the customers, talking to the customers is moving from offerings and information to being helpful and continuous in our engagement with customers. This is a big cultural shift that's going on.
We're also building out a pretty significant analytics infrastructure that allows us to know who's at risk, why they're at risk, and also predict well ahead of any renewal event what they're doing and what we can do to move them forward and keep them engaged. All of this is going to funnel into inside sales teams that really have the mission and the drive to keep renewing these customers. That's how we're going to continue to build up this culture of retention. If you run the numbers, you know retention is critical to making these numbers over the long haul. This cultural transformation is going to continue to drive those numbers for the long haul. The next piece I want to talk about is moving these consumption models down market.
Right now, we have Token Flex, we have these consumption, buy, get access to the whole portfolio, pay for what you use. You're going to see us introduce similar models down market into the SMB space as we move out beyond FY 2016. Other people will be able to access some of these consumption models that provide broad access to portfolio, but operate on a pay-for-what-you-use type basis, which is really powerful because we've seen tremendous results in terms of what people do with these and how much value it brings back to Autodesk, and you're hearing a lot of that from Steve. That's going to be a big long-term development that continues to drive momentum.
The last thing, back to those things I talked about earlier that are going to drive some of the momentum, is this whole notion of capturing new subscribers in the construction market and in very unique areas in the manufacturing market. You're going to hear a lot of that from both Amar and Buzz today in terms of what applications we're deploying and how they're actually bringing in completely net new subscribers that we were never able to reach before. This is going to be a massive engine of subscriber growth over the next two, three, four, five, you name it, number of years, and I think that's pretty exciting. When you look at where we're at, we're doing incredibly well this year, and we're going to be building momentum into FY 2016 and beyond.
There's a couple of things that we will do over and over and over again. We're going to target those non-subscribers. We're going to add cloud to every single account. We're going to reach completely new types of subscribers with the cloud. We're going to expand those consumption models and those enterprise business agreements, and we're going to drive to a 90%-plus renewal culture. All of that is going to continue the momentum and energy that you're seeing right now, well beyond what we're seeing this year. With that, I'd like to introduce Amar Hanspal to come up and talk to you about the opportunities in the AEC business. Thank you.
Thank you, Andrew. Good morning, everyone. A real pleasure to be here. I'm excited to share with you the opportunity we have in the AEC industry. Just to remind everybody what we mean when we use the term AEC, stands for Architecture, Engineering, and Construction. What we mean by that is all of the activity involved in the creation of vertical structures like buildings, skyscrapers, hospitals, and schools. Horizontal infrastructure like roads, rails, bridges, dams, and all of the infrastructure required to extract and process natural resources, things like refineries or mines. It's a very pervasive industry, and you all experience it every day when you go to work and go home. It's a large industry.
It's almost over $8 trillion in economic output today, and it's set to double in the next 10 years to nearly double to almost $15 trillion in terms of economic output. It's underpinned by secular trends, population growth, rise of emerging economies, rapid urbanization, and all of this leads to increased infrastructure spend around the world. The AEC industry is booming. It's creating a really compelling software opportunity. We estimate the TAM to be nearly $12 billion in software alone across all of these categories. We at Autodesk, we've really been capitalizing on this opportunity. To quote my sales colleague, Steve Blum, the AEC business at Autodesk is on fire. Why? Because we're at the right place at the right time, most importantly, we have the right offering. Let's begin with that right time, right place conversation.
If you just look around you, any macro index you look at, labor statistics, construction starts, any of the reports that you see from the Bureau of Labor Statistics or from ENR, from AIA, all these indices are up and to the right. Sure, there are a couple of pockets around the world where the economic activity is still restarting, by and large, the macroeconomic environment in the AEC industry is really, really good. You can do your own analysis of this if you look out of the window in any major city around the world, New York to London, Doha to Tokyo, the cranes are back. This time around, our customers are looking at more than just cranes.
They've been looking at some of the fundamental issues that plague the AEC industry around productivity, and they've been borrowing lessons learned in the world of manufacturing and applying that to construction. You hear the term lean construction a lot. That's very similar to lean manufacturing. It involves all those modular approaches, those things that you see used on ships and airplanes are being used now on large-scale building or infrastructure projects. All of that approach means, of course, there's sustainability and all of that kind of new imperatives for the industry. What all of that means is that this industry is automating as never before, and it's automating through software. If you look at that graph, software spend in that industry is accelerating faster than spend of any other category in that industry, including hardware.
Software spend is accelerating, and when firms in the AEC industry look to spend money on software, the software or the category of choice is building information modeling or BIM for short. Very similar to how people run supply chains using ERP or manage customers using CRM, BIM is now the way customers manage and run projects in the AEC industry. It is strategic. It's at that same level of an imperative technology choice for AEC companies as an ERP or CRM decision would. That's because it delivers real value, and I'll talk about that in a second, because we're going to use the term BIM many times today, I want to level set by explaining what it is that we mean or what is it that the industry means when they use this term, building information modeling.
BIM, it stands for the intelligent, model-based process that leads to the digital representation of the built environment. That's a lot of words. Let me translate that for you. BIM is the way our industry does what you'd call what you see is what you build. You've heard of WYSIWYG. This is the industry's equivalent in terms of driving the build process. It's more than the visual representation. It contains all of the rich metadata, costs, schedules, quantities, the I in BIM stands for all of that. Because it's an information-rich environment, it lets the industry coordinate all of the work across disciplines, all of the stakeholders, as well as across the life cycle of that project. That's what we mean by BIM. It's a fundamental change in the way people manage their projects by putting the digital representation of the built environment at the center.
It's more than technology for technology's sake. It delivers real benefits, you can read all of those cost and time-saving numbers up there. Let me give you a personal anecdote. There's a customer of ours who's just finished work on a major skyscraper project. In fact, it's one of the tallest buildings in the world today. At the start of the project, they were estimating the amount of waste that project would typically result in because they have to budget for that. A typical construction project has about 30% waste. 30%. I've said that number, shared that number with you all before. On this project, they were ordering 70,000 tons of steel for that skyscraper. They were anticipating about 3% of that going to scrap, somewhere in the neighborhood of 2,000 tons.
Because they used a different process this time around, only two tons went to the scrap heap. That's a savings of 1,000 times over, and that's a significant change in the cost equation for these building projects. Likewise, the glass panels, it's 20,000 sq m of glass panels, 19,000 individual glass panels on that skyscraper. Typically, again, they reject about 10% of these because they don't fit on-site. Not one of those 19,000 glass panels went back. Significant cost savings. This is why customers are voting with their wallets to make the move from the old way of managing projects or designing projects to Building Information Modeling around the world. There's a better result, there are better buildings as a result of using BIM, and that's a reason why around the world, governments are now mandating the use of BIM.
BIM is sort of unique in the technology industry, where it's moved from customer adoption to government policy. Whether it is the GSA in the U.S.A., someday somebody will make a song out of that, GSA in the U.S.A. or the U.K. BIM standard, or the BCA in Singapore or all of these various agencies around the world, they are increasingly mandating the use of BIM because they really want to reform the process, not just in terms of productivity and time savings, but with better results at the end of that process. At Autodesk, we've really been leading the transformation of the industry to Building Information Modeling. Not only are we continuing to do that with everything on the technology stack that I'm about to share with you, but we're accelerating that leadership position across all of the disciplines in various geographies around the world.
In fact, if you want to just look back or reflect on the BIM story, BIM really took root in North America and North Europe in sort of the building industry, and what I mean by that is the vertical structure of buildings. It took root in the architecture industry, and then the engineers involved in those projects started adopting BIM. That's still true today. That adoption continues, but we have plenty of opportunity around the world, as well as across disciplines with BIM. We're still in the early stages of BIM adoption around the world. Let's begin at the heart of it all with Revit, because Revit is our flagship product. It's the heart of the Building Design Suite, and it's what defines that what you see is what you build experience. Revit continues to set the bar for BIM in the industry.
As you can see this, it really helps describe the building and all of its detail, helps visualize it, put all of the information in place so you can analyze it, understand it, figure out the construction sequence, and Revit continues to just absolutely drive forward around the world and gets adopted by the who's who of the AEC industry. As strong as Revit and Building Design Suite has been, we have expanded the opportunity for BIM by bringing out a more affordable and capable solution with Revit LT. Revit LT is just like what we did with AutoCAD LT in the CAD market, is making that technology accessible and affordable to more players around the world.
When Andrew talks about net new subscribers, we're bringing on net new subscribers, subcontractors, and all of those kinds of companies that would not have been able to afford the full solution, but now can participate in a BIM-based process, expanding our opportunity with tools like Revit. Expansion is more than just a price point. We are going across disciplines. Revit MEP is a great example. I want to just celebrate the MEP guys for a second because you don't notice them when you're in a building, right? What we mean by MEP is all of the work that is put into a building to give you heating and cooling comfort, fire safety, all of the electrical conduits, all of the stuff that brings internet to your computer. That's put in place by engineers, contractors, subcontractors that work on these buildings.
This is a great example of where we've taken Revit and purpose-built a tool for that design process and brought those people onto the BIM process. Expands our opportunity. It also adds value because many of these customers, some time ago, were using 2D-based processes, and now they're using a full suite of tools from Autodesk. A great example of adding value on an account basis. Another example in that ilk is what we're doing with simulation. As we have these more computable building information models, you can analyze and simulate more. You can do energy analysis. In the case that I have up here, this is a CFD, a computational fluid dynamics analysis in a hospital surgical room where you can understand people have to do the trade-off between comfort and sterility, right? They're constantly trying to understand infection patterns and things like that.
Data centers, another example where heat transfer is a big design point. There's a lot of opportunity for us to add value in accounts with simulation. Sticking on that discipline front, here's another really key discipline that all of us should be grateful for, and that is the structural engineering discipline, because without it, all the buildings and bridges we rely upon would fall down. They provide the skeleton upon which these buildings are built. Structures underpin some of the most ambitious projects around the world, from stadiums to skyscrapers. In fact, the most recent project of the Kingdom Tower in Saudi Arabia, another huge, ambitious project, was all done using Revit Structure for the structural design of that. Revit Structure has matured a lot. All the structural engineers are using it. We also have inside Revit Structure concrete detailing.
I'm sure this is fascinating. It's a lot of cast-in-place and precast concrete. The thing I actually want to touch upon is what we're doing with steel. There's a lot of steel in these large projects. We've taken the next step forward in detailing steel with what we are doing with a product we call Advance Steel, bringing that level of detail into the BIM workflow. To get steel structures in place, there's a lot of fabrication detail, there's a lot of connection detail that has to be put in place.
We've added this to our portfolio, what we're doing with it is not just delivering more value to customers, we are winning new accounts because we go head to head with Tekla, we go head to head with Bentley ProSteel, and we're starting, not just starting, we have won every fight up to this point where we've taken them head on in accounts with Advance Steel. We're broadening our opportunity with this level of detail. A good example of customers using Advance Steel is this company called Steelway. They do all of the steel structures, hot-rolled, cold-rolled steel that goes inside a building.
More importantly, they are a great example of how manufacturing technology is being used in the construction industry, because they prefabricate all of the structure and they're driving sort of modular assembly of that structure on site, which is a major way to not just save time, but to save lives. It's way safer to do these kinds of modular assemblies. They're driving this upstream in the process, in the BIM process, all the way to the architect. Bringing it all together in our core industry between architecture, MEP, and structure, here's a good way to look at it. A project that recently finished in Singapore was a sports stadium they put together. What's unique about the stadium is not just that fact that they have a retractable roof, but the stadium is reconfigurable.
A whole host of sports from soccer, hockey, athletics, my favorite, cricket, can all be played here. The stadium moves around. Also what's unique about the stadium is, since this is in Singapore, air conditioning is delivered to every seat. You can imagine that MEP system I talked about has to reconfigure itself as the stadium moves around. There was no way this could have been built. All of the data, all of the coordination, all of the decisions, all of the dependencies could have been done any way other than with BIM. It's a great example also of an international project that is now starting to drive BIM adoption in a place other than U.S. or North Europe. That's the core story in the building industry. Now let's look at the word building in a different context in another discipline.
As we talked about earlier, the AEC industry includes the world of infrastructure or civil works. That world has been mired in mostly a 2.5D or a pseudo 3D process, right? Why? Because roads are flat many times, but they're complex, right? They involve lots and lots of details that you and I take for granted, from drains to overpasses. They're also huge, right? Roads can stretch for miles. There's never been a technology to do this all effectively in 3D or in a Building Information Modeling in a paradigm until now, with what we started to do with InfraWorks. InfraWorks is to infrastructure what Revit is to building, right? That's the way it's a replica of that. It's purpose-built for the world of infrastructure because it breaks those size barriers, right?
It understands not just what it's designing, but the constraints of that design. You have to design in miles. You have to include GIS information. You have to make it easily accessible to the public. You should be able to do environmental impact analysis. It captures all of that and expresses that for the world of infrastructure. We are off to a really promising start with InfraWorks. A major engineering customer here in San Francisco called it a shining star in their portfolio. It's another example of where existing customers are now adding InfraWorks into their workflow, driving more value on a per-account basis. Speaking of engineering customers, we've been democratizing the plant market for those engineering customers. Instead of complex, expensive 3D systems, we now have more affordable AutoCAD-based 3D environments for them to do their work.
Our story in the world of infrastructure keeps getting better. Every year, we tell you about how well we're doing in the world of roads and highways, how many DOTs we've won. This time, we can tell you in Canada and Mexico, we're also winning. We're winning around the world. Because of our broadening portfolio for the world of infrastructure, we're winning in rail now. We started taking share from competitors in rail, we're starting to win in the world of airports and aviation. These are really big projects that we're starting to penetrate, a good example of that here is the Denver International Airport. They use a whole host of our products, the one that you're seeing shown on screen right now is what we call ReCap, or reality capture, or scan to BIM.
You can see they've scanned the terminal to understand the exterior and interior so they can refurbish it. This is a great story. They've used InfraWorks, they've used Revit, they've used ReCap, AutoCAD, a whole host of our tools to do this new terminal design. San Francisco Terminal 1, when you fly out of there, they're using our tool set as well as the Panama Canal guys. In the case of the Panama Canal, all of the catchment basins, the visitor center, the core canal is using Autodesk products, but the locks that are being designed for the Panama Canal are also using Autodesk technology, specifically Autodesk Inventor. Selling our portfolio to our customers, this is a great example of how we're penetrating this world of infrastructure, really with a signature product like InfraWorks, but with the entire Autodesk portfolio adding value in that account.
We've talked about the core building market, we've talked about the world of civil engineering, I want to go back to something I said in the beginning, which is that the AEC market is doubling, construction companies in particular are increasing their spend. By far, as strong as our growth is in AE, the A and the E side of our industry, our biggest opportunity lies in the world of construction. Very simply put, the construction part of AEC employs the most people. It's more than double of what all the other A&E firms employ. They have the biggest needs because they manage all of the costs, all of the risks, all of the schedule in the construction project. They have to coordinate a whole bunch of work, they are the ones that are automating the most.
We got a foothold in this industry or this segment with what we started to do with Revit and the Autodesk Building Design Suite many years ago. These construction companies embrace that what you see is what you build paradigm and really drove a high level of detail and documentation using the tool set that we provided. A lot of construction companies have adopted our core BIM stack. We really cracked this opportunity open with what we're doing with BIM 360. BIM 360, simply put, connects the office to the field and the field back to office. People out in the field, instead of walking around with reams of paper and clipboards, have access to the building information model on their tablet device. When on-site they find something and they have to report an issue, that issue is not just reported on a piece of paper.
It's tied back to that digital model so that the people back in the design office can now understand what's being discovered on-site. That loop between what's discovered on the field and what is being designed in the office, it's a huge game changer for the construction industry. It lets people automate many steps of that process. It lets them automate production planning, quality control, commissioning, handover, a whole bunch of things. Our customers absolutely love what's going on with BIM 360. Here's an example. Mortenson Construction. They called it one of the best software rollouts. They saved a bunch of time and money. More importantly, they were, for the first time in their history, able to measure something they'd never been able to measure, which they call first time to quality, FTQ. Really key metric.
It's a huge change for them, a cultural change in their company, to be able to do things like this and connect the job site back to the design office for the first time. Customers like Mortenson, they're resulting in this rocket growth of BIM 360. Our current number of BIM 360 billings growth is over 200%. It is, simply put, one of the, if not the fastest growing product in our industry's history. We're really pleased with the level of adoption and take-up we're seeing with BIM 360. In fact, between BIM 360 and what's been happening with the Autodesk Building Design Suite adoption in the construction industry, we're winning the who's who of the construction industry. Steve and Andrew touched upon these Enterprise Business Agreements, and Steve will talk more about that.
Many of those are these kinds of firms, because they're really, again, back to that strategic value of BIM. BIM is at the core of what they're trying to do in terms of automating their entire process and embracing the new way of building things. We're really pleased with the speed at which we are capitalizing on the construction opportunity. It is and will remain our biggest growth opportunity. Just to summarize, we've got great strength with our leadership position in the architecture and engineering disciplines. We have plenty of opportunity around the world in terms of going to Japan, Germany, and many other regions to drive BIM adoption. We also have this increasing opportunity to drive BIM adoption in engineering disciplines. We talked about MEP, structure, civil.
Civil is one of those really promising segments as it starts to move from 2D into this BIM environment to really drive and for us to take a leadership position in there. By far, our breakout opportunity is in construction, and we're doing great. Construction is a huge opportunity for us. Another huge opportunity for our company is the world of manufacturing. To tell you more about that, I'd like to turn things over to Mr. Buzz Kross.
Thanks, Amar. Thank you. Got it. Good morning, everybody. Let's just start with a little talk about manufacturing. Everyone knows what the products are in manufacturing, the sort of things our customers do. The thing that's really interesting about this industry is it's not just about design. Autodesk traditionally has had its strength in that we helped our customers design their products, but there are many engineering disciplines that it takes to produce a product. We have specific solutions for the many different engineering disciplines. There are designers, maybe who are more artistic. There are engineers who do functional design. There are manufacturing engineers who do something very different. There are analysts. Again, many. Each one uses a different type of software and does a different sort of function in their company. There's lots of expansion capability in this market just by getting to new disciplines.
The market's big. It's about $11 billion in size, and it's a little bigger than it's been. It's been steadily growing. This index here is the Purchasing Managers' Index. It shows purchasing managers' confidence in the market. What it's shown is, it's shown rapid expansion for quite some time now. The market itself grows at about 8%. $11 billion growing at 8% means there's lots of expansion every year. Manufacturing Autodesk market, and revenue's grown at 17% during our last announcement period. That's about double, a little more than double the rate of the market expansion. It tells you we're growing share. If you look at things like billing, which we think are even a better measure of how our business is going, it's considerably north of the 17%. We're pretty positive about where our manufacturing business has been to date.
This presentation, however, is what we're doing for the future and what we're doing to, per Andrew's point, how we're helping drive 20 or 12, 20, 50. In manufacturing, we have specific programs and products that drive each. One of the questions is how do we drive more value per account, or how do we drive this 20% that Andrew talked about? There are lots of different ways. I'm going to go through a few of them. One is the automotive industry, really important for us. If you have a beautiful car, if you make a beautiful car, it was designed in Alias. It's just a leading solution for conceptual design and is used for every car company in the world. It's been used that way for a long time.
For us, this business has expansion in Class A surface definition, which is the step right afterwards. We're seeing a lot of growth in that solution, in that category, where we're taking conceptual design into the more technical phase of modeling it. We also have some very new types of products driving revenue in this business. Visualization is of huge importance to the car companies. They put so much investment in a car before they make it. They want to know exactly how it's going to look like, what it's going to look like. They want to know if they put this little notch in a corner there, how it's going to appear on the car. We have a new solution called VRED, which is really starting to take the industry by form.
Whether you use Autodesk in the front stage or the back stage. Auto companies are starting to use the visualization from VRED because of the high-quality surfaces and images you can generate. This is a pure digital image, although you certainly could not tell. They use it for big, high-quality stills as well as video as well. The other area that's becoming really important in automotive is interiors. Interiors more and more sell the car. Okay? Interiors are very hard to do, a lot of soft materials, a lot of light reflection inside the car. Making a good digital model really fast is difficult. Again, you would not think that's a digital image of that. That's an Audi interior, obviously. It really tells the auto company a lot about the car before they produce it, and that helps them make really good design choices.
VRED marks a really good way we can add value to customers, and a way we can increase the value we get from customers as well. I want to switch to mechanical design now. This has been the mainstay of our business. The product we use here is Product Design Suite. Inventor is the heart of this solution. We have a lot of advantage with our Inventor product compared to our competitors. Okay? One of the things we have is modeling. When you make a model, there's many different methods you can use. We have the ability to use virtually all of them at the same time. Parametric modeling, direct modeling, and freeform modeling, all integrated together into that one product. Mechanical design tools traditionally were really good for products like those first two images there. They're prismatic, they're very functional.
They don't necessarily have an elegant look and feel. If you're going to produce something like that third image there, the handle, how it feels in your hand is really important. I want to show you some of our modeling here. I'm going to run a little video here that shows you how Inventor modeling now works in this newest release. If you're going to make a design like this is an immersion blender. When you buy something like this, a consumer product like this, how it feels is really important. If you're a cook in the family, you're going to pick this thing up and just feel it. Very hard for traditional solid modelers to make designs that have that elegant look. You can see the dynamics here.
You put something like a fillet in, and you just stretch it really easy, so you really get a feel. Here's the real magic of it. When we're making the handle portion of it's not driven by complex equations, traditional engineering rules. Very much how it feels and looks. The user just grabs the surface and pulls it, so he gets something that looks very much like he wants. In fact, in this case, he's doing it over a set of sketches. An artist had designed the sketches, and he's able to model exactly to the sketch parameters without a lot of the complexity of the math that used to get used. It very much differentiates how you would use this product and how you could design a new sort of tool.
It's a big advantage we have in that you can design really complex tools like this. Even things like the very top here, really easy just to cut it off so it has that elegant shape along the very top of the curve. It's really, I think, very impressive of how far we've been able to come with these tools, that you can model a super complex shape like that. This sort of thing was virtually impossible with traditional modeling tools. All right. Let me go on and talk about another really interesting advantage. It's BIM for manufacturing. Amar did, I think, a really good job talking about the big advantage of BIM. Construction companies are switching to BIM for everything that they start to do. Many manufacturers make products that go into buildings.
Our Product Design Suite has a fantastic, the world's best connection between mechanical design products like this chair. If you were to design that really elegant chair, the mechanical engineer would design it in much detail and take a look at all the surfaces and wants it to look beautiful. It becomes really easy to put it right in Revit so you can have that really good connection. Manufacturers that design for building products have a big advantage with the Autodesk solution in that they can design things for a BIM environment. No one else can do anything like this, and it's a good way we can drive extra value from our customers and driving some new customers. I said before that there's lots of different types of engineers that go into designing a product. One of the really good, big segment is simulation.
It's a way we can add new value and add new customers in simulation. We have a very full range simulation solution. The big news for us is Nastran is the engine we now do to do mechanical design. The biggest portion of simulation is mechanical analysis, stress analysis. There's lots of different types you might do. There's about five times more users doing mechanical stress and three times more revenue than any other segment. It's an important segment for us. Nastran is by far the biggest name, the most respected name, and people just have confidence in the results that you get out of Nastran. This is an acquisition we did a few months ago. We've already integrated it into the Autodesk product, and customers are using it. We find our win rate is much, much higher now that we have a Nastran-driven engine.
Our solution in simulation is very broad range. From mechanical simulation, to fluids, to plastic injection analysis, we have many different ways we can derive value per account by letting them analyze their products a lot better than they could ever do before. There's lots of ways we can derive value with our existing products into our customers. Let's now talk about the 50% issue. How do we drive net new subscribers? There's quite a few different methods. Let me just talk about a few of them. One is fabrication. Okay, I mentioned there's lots of different types of engineering that goes on to make a product. An important engineering discipline is something like this. How do you actually make a product like this? Manufacturing engineers have a whole discipline in how they make those sorts of things.
This is revenue we used to really not partake in at all. We really only did design, then did design simulation. We've entered the fabrication or manufacturing engineering market. That market's big. It's about a $1.8 billion market by itself. Okay, again, market that we weren't participating in before. We now have dedicated products that work toward those things. We drive net new subscribers in this market with a set of solutions that vary based upon the complexity of design. From very simple manufacturing problems like the sheet metal part on the lower left side, to more complex two-and-a-half axis stuff, to full three-axis, five-axis machining, all the way up to specialized functions like composites. We have dedicated tools to do each one of these things, including composite tools that help you manufacture, lay out the layers in a composite model.
A very advanced area that's becoming hotter and hotter. We did an acquisition of Delcam, a little less than 6 months ago. Okay, Delcam's part of that market is right in the center, the complex machining. If you make complex parts of molds and dies, Delcam tends to be the solution for you. Delcam's pretty interesting in terms of driving net new subscribers. It's the world leader in NC. Everyone views Delcam as the top of the pyramid, the tool that does the best job in these complex machining problems. For us, it's incremental revenue. Again, I said we used to not play in the manufacturing engineering portion at all. We now drive about $100 million in manufacturing engineering revenue, which includes Delcam as a part of it. We have other tools that help there as well.
It's an acquisition we did about 6 months ago, often you see acquisitions have a little bit of a dip in revenue. We haven't seen that. We've been able to continue to grow it and grow it faster than before. I think we've done a really good job in terms of driving net new subscriber with our Delcam solution and our other manufacturing engineering solutions. A whole new set of users for us. You can't talk about making things without talking about additive manufacturing. Okay, we're really focused on this space. To be clear, it's not about this stuff here. It's not about printing little plastic crap out for the consumer market. Okay, it's interesting and all, but it's not going to change my industry. I know that's very popular. I think it's way overhyped.
This is what it's about right here, additive for industrial. That's sort of the quiet revolution that you don't read about in the newspaper, but is making a big difference. A really big problem in the manufacturing industries is the strength-to-weight ratio that you have to produce things. If you make something that goes into a car, a train, a plane, or even a common consumer product, what it weighs makes a big difference. The thing you see in this design here, and that's just a bracket, just holds three things together, is the open lattice in part of this design here. You see it right along here. That lattice structure makes the product dramatically lighter, but has as much strength as it had before. Unmanufacturable, unless you're using additive manufacturing. People don't even think about it this way.
You can take this far further, and Autodesk has. Autodesk leads the industry because we can do design like this. This might look like the same image you saw before, but it's actually really different. If you look at the lattice structure along the bottom side, it's a larger cross-section than it is along the top. We're able to produce designs that vary that lattice based upon the use of the product. There's more stress on the bottom part, so you add more mass and more strength, less where you don't need it. You can really optimize the sort of design that you have. Again, people won't naturally come up with designs like this. It takes software like Autodesk software to do this sort of design. It goes a lot more advanced. People can kind of visualize that.
That's a simple octagon or hexagon sort of shape. It goes quite a bit further. There's all sorts of complex lattice structures that can be produced. Shape optimization is impossible with traditional methods. We're really the only ones who are paying attention to this, and it's starting to make a big difference in key industries. This is sort of for the leading edge today, but it's very quickly going to go everywhere. You're starting to see it more. You hear more and more common products that are looking for these sorts of things. The next thing we do is the cloud. We have a really advanced cloud business, and I want to give you two examples of that are real specific about what we've done. One is PLM 360. PLM 360 is our PLM product, is now entering a growth stage.
The product's been out there for about two and a half years. I think we're really pleased with what we've seen with it. In case you don't know about it, let me give you a little reminder about what PLM 360 is. It is a unique product in the PLM market. We entered the PLM market about 10 years after everybody else. We waited until we could make a difference. In fact, it was the cloud becoming a real idea that we saw we had a chance to make a differentiated product. There was no point in us entering a product with something that was like Teamcenter or ENOVIA, it would just been another similar to. This let us make a product, the cloud made us let us make a product that was very, very different. It's cloud-based, which means zero deployment.
You don't have to buy servers, you don't have to spend the millions of dollars that our competitors' solutions require you to spend to set these things up. Because it's cloud-based, it's accessible on any device. So it'll work on your iPhone, it'll work on a tablet, it'll work on a PC, it'll work on a Mac. We see users doing it on all sorts of devices, wherever they are, any location. Very, very fast deployment. Up in minutes, not months or years as is traditional. Business for PLM 360, I think, has been quite good as well. We see solid growth in new seats. Last year was all new seats. This year, the new thing is we have customers returning. Customers that bought last year are coming to buy more new seats from us this year. That, for us, is a very, very positive sign.
A very, very high retention rate. I wanted to put exactly the number in there, I wasn't allowed to. Let me just say it's the highest I've ever seen of any product in my long history of making things. An excellent retention rate of customers continuing to buy it. This is one of these modern products that do the cloud-based, annual or monthly sort of solution that we want to help our customers move to. We've seen a 96% growth rate in billings and a 237% year-over-year growth in new customers. Pretty phenomenal results for a brand-new product.
One thing we see is there are a bunch of companies out there in the market that I call cloud-only companies, because the new trend that I see. When we first introduced these things, many of our customers said, "Not me, I'm never going to go to the cloud." Okay? They say, "Well, tell me more about it." We also have started encountering these companies who say, "I don't know why I'd use anything that's not a cloud-based application." They're trying to get rid of the things that are installed on these expensive servers they have to maintain. They want to do things that are a more modern way. I would say Autodesk, by the way, is one of those companies. We're looking to replace a lot of our old infrastructure with cloud-based tools, because we think it's just a modern system that works a lot better.
I got a couple examples for you of customers who have done just that. You all know Quirky, a very interesting company, very interesting products. I've got a bunch of them. They're interesting to try. They're trying to drive even faster than the fast pace they're on already. They're now trying to drive new product launch cycles to just 30 days. This is really an amazing accomplishment in an industry that used to take multiple years to introduce new products. They're trying to really drive them down, and at the same time, more than double the number of products they come out in that cycle. From 20 a quarter to 45 a quarter. They needed a PLM product to do this. They wanted, of course, a cloud-based product. They're one of these cloud-only companies.
They're looking around, they said the Autodesk solution is the only one that there is out there in the market. We're able to drive a solution to the customer that was really unique and that it's in the only category. Another interesting one is Memjet. Memjet is a NetSuite user. That's how we first met them. We met them at a NetSuite conference we were at. They're definitely saying, "We only want cloud-based applications around this place." In fact, what they say is SolidWorks Enterprise PDM is the last on-premises application they have. We don't think that's going to last either, frankly. Okay. They're starting to buy solutions like PLM 360 as well. We see this new part of the market that would only consider applications like this. I think it speaks well for where we can go in terms of driving net new subscribers.
Just a little other information about our PLM business. We have a lot of really new customers. In fact, all these companies here, in fact, a big portion of the companies we see, don't use Autodesk for their CAD solution. They have some other CAD solution. We've been able to go into new accounts in a new way and sell them a new solution to new types of users. In fact, what we see is an average of 130 net new users per account with our PLM product. Okay. That's far higher than our average CAD products. We're seeing a lot more net new users per account. We also see subscription billings to be much higher, 2x our normal subscription billings for our PLM product. Pretty phenomenal results from our PLM team.
I think the product's so differentiated, it gives us a new way to attract new customers. Last little thing I want to talk about, I call it a little thing, but it's actually, I think, an incredible shift in the market, is Fusion 360. Okay. Fusion 360 is a brand-new entrant. It's actually been in the market for about a year now. The way I would describe it is just like this: It is the first ever CAD in the cloud product. I think Fusion has now certainly hit its stride and is starting to drive real new business. Okay. It is a new, very fast, easy paradigm. Companies that are looking to do things in very different ways are drawn to Fusion because it is a very creative, new way to design products. I'll show you a little bit of it.
We have two types of users we see a lot of, those that are brand new to CAD, they haven't used a CAD system before. Maybe they've made products they didn't think you could design with a CAD system before. Generally you couldn't. It takes a product like Fusion to do that. The other type of user is those that are replacing multiple old CAD products. Rhino, SolidWorks combo we often see, where you need this combination of something that's aesthetically pleasing and something that's functional. Fusion does both those things together, without any kind of data loss. It's really ideal. Fusion is targeted at these new types of users, new types of companies. 21st century product development is just fundamentally different than what we've seen before. First, lots of rapid iteration goes on in this market.
Customers often design things in days instead of months. In fact, we have customers that say, "I design on Monday and Tuesday, I sell my product to Walmart or some distribution system like that on Wednesday and Thursday, I get my operations running." We see this cycle going over and over again. Old CAD, it cannot work to do those things. A product like Fusion is the only way. Now, one of the reasons is extended teams. Teams are not just in one place. There's lots of different disciplines. If you're going to design and make a product in a week, you have to have everyone involved in the design and manufacture together at the same time. Only products like Fusion can do that. The last thing is access anywhere. Of course, if you have extended teams, they don't live in one city, they don't work in one office.
They need to be able to get access to everything. Fusion makes it ideal because of its cloud base to let everyone gather at the data, which is the center. A few examples. One is Modbot, a real heavy-duty mechanical application. They make robotic systems. The magic they make is this little module you see up in the right-hand corner. It's a real smart system that lets you control joints. Completely configure robotic arms. The way they describe their engineering team is it's two friends, one electronics expert, and one CAD guru. Okay? Fusion is the first product ever that would fit for that sort of environment. In the old days, we would've been able to sell to the one CAD guru. We can now sell to all four of those guys.
They can all use together, they can all share it, even though they have very different views of it. This next one I think you'll find to be really surprising. Products like this, Creative Things, is also a Fusion customer. They make products like toys, very aesthetically driven, not function driven at all. Only Fusion can design products like this that have to have this exotic shape. Think about that mermaid. Design that mermaid, that thing right there is a mermaid. Or this is even better, this little monster or whatever it is. Just impossible in the old days of CAD. Fusion can do that really well, and it can do this in its extended team. One of the things that Fusion, it's not just that it's a CAD product that's on the cloud. It's a full product development solution.
It lets you do modeling, of course, like CAD always has, but it also lets you test the design. It lets you do things like rendering. It gives you documentation so you can make drawings. It lets you NC machine the molds or the product itself, and it's built into the collaboration engine that Amar's going to talk about after break, A360, which lets these teams share information even if they're not very technical. Let me show you Fusion in action, and you'll see why it's so good for things like toys and advanced shapes. We're going to make this little headlight here. It is unbelievably easy. You just grab shapes and drag them with this little triad. If you want to adjust how this thing looks, it does not take a real deep technical expert to do this. It takes a product expert.
He can explore all sorts of shapes and determine something that's going to be really ideal. Again, how this thing appears will make a big difference in your ability to sell it. Autodesk Fusion, the first cloud-based product ever, really easy, highly dynamic and fun, attracted a new type of user for us. One of the reasons it's so interesting is Autodesk Fusion runs on the Mac as well as on the PC. We just put it up on the Mac App Store a few days ago. We saw 30,000 new accounts in the first 10 days. Okay? It was ranked number 1 in graphics and design in the top countries worldwide. We've seen real good pull-through, and think about how important the Mac is to these new generation users. You just don't meet that many 20-year-olds that are going out and buying a PC.
They're using Macs. This is the ideal place to be for this new sort of market. Let me summarize all the things I've said here. We're really working on a 12/20/50. We're driving value from our existing customers through product differentiation. Products like Autodesk Product Design Suite really do things better and different than you've ever seen before. Things like VRED add a lot of value with new sorts of customers that we can drive. We're also adding new net users, new users via new engineering disciplines. Things like manufacturing engineering. Again, we used to get no revenue in that portion at all. Now about $100 million in manufacturing engineering revenue. Last, the cloud. The cloud's a real business today. It's not a future any longer. It's out there. We're driving revenue with it. We have thousands and thousands of users.
We've been able to produce the first CAD product on the cloud in history, and we've been able to produce the first PLM product in history. We think we're well on our way. Thank you very much. I think next is a break. I think we're due to be back at 10:30 A.M. Thanks, everybody.
Ladies and gentlemen, if you'll come back in and take your seats, our program will begin again shortly. Thank you.
Please welcome back Amar Hanspal.
Welcome back everyone. I want to begin this session now by reminding you of something that Andrew said at the start of the first session, and that was our drive to attach cloud to every account and drive value in every account by doing that. Buzz and I both touched upon in the manufacturing industry and in the AEC industry how cloud services like PLM 360 or BIM 360 are being added or being used by our customers. Buzz used the example of Memjet. I shared Mortenson. We're seeing wins and adoption of our cloud services in each of those industries. You ain't seen nothing yet because we have a real ace up our sleeve. We have this one cloud-based offering that we believe will drive broad adoption that every single one of our customers will subscribe to, and that's the solution we call A360.
In helping you understand how we're going to add cloud to every account, I want to spend some time talking about A360. Just to summarize where we are, every single product of ours is leveraging the cloud for computationally intensive tasks. We're doing things like BIM 360 and PLM 360, we have identified a broad-based opportunity that we are tackling with A360. That broad-based need really lies in what our customers do every day. No matter what they work on, cars or windmills, bridges or buildings, the one thing common in the design and engineering community is all of our customers work in projects. That is the single organizing principle for the work that they do.
The other observation is, despite this being sort of a universal need that exists across the design and engineering community, there is no good solution today to help them work in a project-centric environment. You contrast that with the software industry. The software industry uses source code control tools. Today with distributed development, we have tools like GitHub and lots of ways of organizing and connecting information and people. The design and engineering community does not have, until now, and what we're trying to do with A360. Our customers have tried. For many years they've used email, they've used SharePoint or FTP. They've even tried tools like Box and Dropbox to do their work. Each of these solutions is kind of like just one facet of the problem.
To have this project-centric environment, they need the ability to pull it all together, because design and engineering information is complex. The fundamental challenge in this world of design and engineering is that there are many players. There are engineers, there are designers, there are suppliers, there are customers, there are people involved in quoting, lots and lots of players in that whole process to bring a product or a building or piece of infrastructure to life. All of these people use a variety of tools, CAD tools, simulation tools, CAM tools, you've heard all of this when I discussed our AEC portfolio or Buzz discussed our manufacturing portfolio. Lots of tools creates lots of data. It's not just the size of data, it's also the fact that it's heterogeneous.
All these different CAD formats that are out there or office formats and CNC milling information and G-codes and all of that stuff that's out there. It's also the fact that it's complex. There's no easy way to understand that. You can't visualize it. You can't discover patterns in it. To just compound the problem, when you think of our customers trying to figure out what's going on, who's got the ball, when is it supposed to be finished, who's got the next action item, they now have to resort to all of this fragmented communication, email and phone and text messaging. These projects, there's no way to consistently understand what's going on. That's the problem that A360 steps smack into the middle of to solve.
To connect all of the people with all of the information that's relevant on the project, make that information digestible, make it understandable, index it, make it visualizable, so that even if you don't have the source authoring tool, you can see what was created. Also to provide an infrastructure for tracking all of the activity on that project. To make sense, to put tasks and milestones and notes visible, accessible, and actionable to the project team. To do that on any device at any time from anywhere. We're addressing a really universal need in our industry, a broad, horizontal, universal need across all of our customers and across even our competitors' customers. When our customers now log into A360, they can see all their projects at a glance. They can see what's going on.
They can see into the project, all of the activity in the project. They can initiate new projects, add information. I actually want to take a moment and ask you to look at the thing that's labeled two, Search anything. It just sounds like a feature, but think of this. We all today use Google to find information that we're looking for. Our customers have no way to look across all of their projects to find things that they may have worked on before. If they're looking for a component or a design part or something that they had worked on six months ago, or they knew another team had worked on, there's no easy way to do that until A360 comes along and indexes all of that information and makes it discoverable.
Much like Facebook has the social graph, we are standing up the design graph, the relationships, the patterns between all of the projects and all of the information in the project. Now, within any project, again, our customers can look at the people in the project, the activities specific to the project. They can look at the data within the project, addressing that universal need. We're combining all the things that they had to use, email and phone and SharePoint and Dropbox, all of that capability is now in one coherent experience. When they're looking for information, I talked about all of the complexity in the design and engineering world.
They just send a URL to somebody that is part of the project or invite them into the site to look at a piece of information just by clicking on that link, that person participating in the project is able to look at the data in all of its rich visual and information detail without ever needing the source CAD system. We support over 60 file formats. We continue to add that every day. We support our competitors' file formats. We just make it dead simple to share information no matter what the source data is, source system is with the extended project team. The information is indexed things can be found and discovered.
In the case of this beautiful engine, if you're looking for this one part number, this one component, you can just type in that in the search field, the system will isolate it for you. This real combination of making sense of a project, letting you discover things, letting you see things, letting you invite and understand the activity on a project, is a real universal need. We put the product out at the start of this month, out into commercial release. We have a business model. We charge $10 per user per month for A360 now, our customers really are giving us positive feedback on how this product is working for them. We're getting great responses, we're continuing to improve the product as we go along.
A360 is the way we are going to drive that add cloud to every account going forward. We're really confident that our customers will not just like what they see and use it, but that they will subscribe and add A360. We have a real focus in our company around A360 going forward in the product group, in the sales organization, the marketing organization. It's a key part of that 20 story where we're adding value in every account. It's also a part of the 50 story because as I mentioned earlier, some of the customers who are using A360 are actually competitive accounts that really have had no good way to look at SolidWorks information or any other kind of competitive information in this shared project environment. We're really excited. We're getting great feedback.
To tell you more about this adding value and getting new subscribers with A360, in fact, the entire Autodesk portfolio, it's now my pleasure to invite Steve Blum up here.
Thank you, Amar. Good morning. I'm going to actually come back to A360 toward the tail end of the presentation because we are very excited about it because it does attack both the 20 and the 50 part of the equation. More to come on that. Today I want to step you through basically four things. I'm going to give you a reminder of our sales approach by customer segmentation, just to make sure we're all on the same page about how we're going to market with named accounts and SMBs and very small businesses. I'm going to focus in on three areas that are of top importance to us, and I know from some of our conversations, high interest to you. I'm going to focus on what our partner strategy is going forward.
I'm going to focus in on our named account sales approach and strategy. Give you an overview of what our subscription sales strategies are, focusing on all three of our subscription businesses. Okay, using the same approach that Andrew had, we basically break up our customer segments into named accounts at the top of the pyramid, small to medium-sized businesses in the middle of the pyramid, and very small businesses at the bottom of the pyramid. For our named accounts, this is an Autodesk-led sales approach. I have dedicated teams, sales and technical folks that are assigned specific accounts, and they're responsible for driving business with those accounts. This is a direct fulfillment process in most instances, and we're leading with Enterprise Business Agreements, the consumption-based models that Andrew talked about earlier.
What's important is we're not just focusing on getting a licensing model in with those accounts. We're focusing on the full adoption, the full engagement process with those customers. I'll go into that in some detail here in a little bit, but it's focusing on services and how do we really help them be successful when they're given access to this software. We're also, of course, leveraging all of our cloud service subscriptions in this space. There's high interest there as well. I do want to make sure that you understand there are some instances where our partners play a role with named accounts. They may be able to help add value in either the pre or post-sales part of the engagement process. We've developed service fee models. I'm going to use the term service fee.
Some of you may actually think of it as agency fees or an agent program. I'm going to use service fees, so you can use that as an alternating statement. We have those engagement models in place for partners in the named account space where appropriate. Next, I'm going to talk about the SMB. This is where the majority of our business comes from. This is a partner-led sales approach. We're leveraging our partners around the world. I have territory teams that team up with them to help sell the solutions to our customers and ensure that they're being successfully adopted. Our partners take a lead in that capacity. They lead with Suites, with maintenance subscription attached. Of course, desktop subscriptions are also available for customers in the SMB space, and we are certainly seeing interest in our cloud service subscriptions.
This year we've invested in building out an inside sales team. I'm going to go through that in a little bit here, but we do have our inside sales teams focused on driving subscription renewals, especially when customers have expired and moved past when they were supposed to renew, and I'll step you through that in a little bit. In the VSB, very small business space, we really have either a partner eStore or e-tailer-based approach, depending upon the market and the best way to get to market based upon customer buying preferences. Predominantly, they're selling perpetual licenses, sometimes maintenance subscriptions attached, sometimes not. LT is a highly bought product in this space. Our different subscription offerings, of course, are available to these customers, and we're actually seeing strong traction for desktop subscription, specifically with LT, and to some extent, our animation products as well.
Again, that inside sales team I mentioned is also focusing on driving subscription renewals in this segment. That's our sales approach by each of the segments. Let me step you through our partner strategy overall. I know this is an area of interest. Our primary focus-- Well, first, let me make sure you understand. Our partners have been and continue to be a very important part of our overall go-to market. We do typically 80%-85% of our business through our partners. It's been that way. We expect it to continue to be that way. The partners are critical to our success. This is why we're focusing on helping our partners prepare to transition their businesses to support our new business models. I'm going to give you a very specific example of something we're doing in the fourth quarter in a few minutes.
As I mentioned, we have set up service fee programs to help support selling our cloud service subscriptions or selling opportunities into our named accounts with our partners. Again, we fulfill that business directly with Autodesk, we have service fee programs for when partners are playing a role in either the pre or post-sale part of the process. We're going to continue to invest in our high-performing partners. We're going to continue to reward specialization and the investment in resources that drive those specializations in particular industry segments. We know we've seen partner consolidation. Partner consolidation has been happening around the world for many years, and we expect it to continue. We don't see that as a bad thing, by the way. That's an opportunity for us to actually focus on working with fewer but stronger partners in each of the markets that we're focused on.
We actually then see many of our partners, the stronger partners, getting a greater individual share of the overall SMB space. You should expect to see partner consolidation continuing. That's not an issue. We actually recognize it. There's healthiness in that for our partners who get stronger and for Autodesk, and as well our customers. I do want you to know that we are expecting to bring new partners into the Autodesk ecosystem as well, specifically to help us focus on our newer offerings, our cloud offerings, as an example. We recognize that there are going to be specialized partners that want to come into our ecosystem and just focus on one of those new offerings. We're encouraging them to do that, and we're developing new partner models to support that. Let me step you through a partner economic example.
Now it's building differently. This is something we're rolling out for the fourth quarter. This is an actual example. In fact, we have been in the process of making our partners aware of this program for Q4 this week. What I'm showing you here are absolute dollars. These are margin dollars that our partners get on selling AutoCAD LT as a perpetual license. That's the bigger bar. That's what it looks like today, or a desktop subscription version of AutoCAD LT. This is exactly the margin profile in dollars today. In the fourth quarter, we want our partners to focus on desktop subscriptions for AutoCAD LT. As a result, we are lowering the margin dollars on perpetual LT seats, and we're raising the absolute margin dollars on desktop subscription LT seats.
In fact, our partners will make the same margin dollars selling a desktop subscription version of LT that they'll make selling a perpetual license of LT. The partners will see this as a plus because they can sell an LT perpetual seat once. They may be able to attach maintenance subscription, maybe not. They can sell the LT desktop subscription over and over and over again, and they'll make these margins over and over and over again. We think this is going to be an interesting test to see how we can start changing the economics to incent our partners to promote what we think is the best business offering for our customers, which is desktop subscription. We also have introduced two new partner models to support our new business offerings, the Advisor Program and the Referral Program. Again, these are service fee programs.
We're basically transacting business directly with the end customer and paying our partners a fee for the services they're providing. The Advisor Program is a program where our partners need to apply to become an advisor, and it requires dedicated sales and technical resources focused in the particular area of specialization to support the offer that they're applying for. There's a cost for them to do this. Once they're accepted, they're in a position to be able to go and identify opportunities. They register them. We, Autodesk, work together with the partner to sell the customer the solution. Again, transact the business directly with the customer. The partner stays engaged in helping drive adoption and use of the offerings. It will be involved in the renewal and will earn a service fee for doing that. We also have a Referral Program.
This program doesn't require anyone to apply. It doesn't require dedicated resources. It's an opportunistic program for any of our partners who may be engaged with a customer, and they may identify an opportunity, say, for PLM 360 or BIM 360. They can register that opportunity. We, Autodesk, take the whole lead and the entire responsibility of engaging with the customer, closing out a contract, helping them to adopt and renew that offering over time. We pay a referral fee when the contract's closed to the partner. The referral fee for this program is lower than you would expect that we would be paying to an advisor partner who is making a greater investment. This is how our partners can engage with cloud service subscriptions moving forward.
As I mentioned earlier, there are instances where our partners add value in either the pre-sales or post-sales part of the engagement process with named accounts. When they do, and they work closely with our teams, they can put together a teaming service agreement, and through delivery of a set of responsibilities defined in that teaming agreement, we would pay them a service fee. Okay. I'm going to transition on to our named account sales strategy. We're continuing to invest in dedicated sales and technical resources focused on our named accounts. We're leading with our enterprise business agreements. Again, this is what Andrew talked about earlier, our token-based licensing model that gives access to our entire portfolio to a customer engaged in this model. It's a consumption-based business model.
One of the things I want to make sure you're clear on, though, is that that's just one element of the engagement process. In fact, if that was the only element, I wouldn't be excited about the model. Included in an enterprise business agreement are Enterprise Priority Support requirements, a dedicated customer success manager focused on that account, and identified consulting credits and even consulting implementation projects to ensure that the customers are successful in adopting and using the offerings, sometimes even helping them get their work done in some instances. It's this entire reshaping of the engagement that is what we call an enterprise business agreement. We see that as critical to having this direct engagement with our named accounts and truly developing strategic relationships that are lasting and long-term and added value to our customers.
Again, as I mentioned before, but highlighted here as well, we do have partners engaging and helping us in this process from time to time, and in those instances, we set up a teaming agreement and pay them a service fee. We introduced the Token Flex model Q4 of last year. Since then, 10 customers, 10 named accounts, moved from our prior enterprise business agreement, which we called Multi-Flex, to the new consumption-based Token Flex model. I have actually shown here all 10 of those customers. What we're evaluating here are the number of active users that we have in the Token Flex model compared to the prior Multi-Flex model.
What we're seeing in the early days of our new model is a 2.4x, or said differently, a 240% growth in the average monthly users of Autodesk software within these named accounts, within these 10 named accounts. We're really pleased to see that we're driving deeper account penetration into the named accounts. They're taking advantage of more of the offerings because now those offerings are available, and we're actually reaching into other parts of the enterprise that before we weren't getting to. Again, these are early days, and if you're wondering, okay, why would one customer be on the one end or the other end, there's lots of reasons. These customers may have been in the Token Flex model longer because we actually see it growing over time.
Some of them may have had more licenses available to them in the old model compared to the newer model. They may be in different places in their implementation and adoption plans overall. The early results are very exciting and are absolutely aligned with our expectations. Now, when I visit with the executives from the named accounts that have implemented Token Flex, these are the types of things they're telling me. They're saying, "We love having access to the full portfolio. It's so much easier now to track and report how we're using that software." Many of these companies are actually using it to start bill backs, to actually track who's using the software and actually have the departments that are getting the most value paying the bill, and they like that. It's an easier way of tracking than they had before.
They love that they have the flexibility to access the software at any time, whenever they need it. They also recognize the level of support they're getting from Autodesk is dramatically different. In fact, it's game-changing different. The amount of comments about enterprise priority support or the value of that customer success manager come up every single time I have a conversation. Absolutely, they continue to say, "Hey, we are certainly viewing Autodesk very differently, and we feel like we're moving into a space where we have a strong strategic relationship." Okay, now, I'm going to step you through our subscription sales strategies, focusing first on desktop subscription. For FY 2016 and beyond, we want to lead with desktop subscription as the best overall economic offering or business model for our customers.
There's going to be a whole lot of emphasis and work done to drive customer awareness about why that is the case and how they can benefit by moving their business model to a desktop subscription model. Andrew talked about the 2.9 million non-subscribers. These are the active users of our software that have not become subscribers within our ecosystem. We recognize that the premier offering to turn those users into active subscribers is desktop subscription, and we'll be leading with desktop subscription as the primary offering to that 2.9 million users who are non-subscribers next year. We also see a tremendous opportunity to use desktop subscription as a competitive win strategy to get customers who are using alternative solutions to actually move to an Autodesk solution through the desktop subscription business model.
What I'm really excited about as well is that for our partners, we'll introduce the monthly option that Andrew talked about. More importantly for our customers, they're going to want multi-year desktop subscription options, which we also will be rolling out next year. We're going to align the incentives with our partners to ensure that they have a business model that's set up to support driving desktop subscription. I just showed you the example of what we're doing in Q4 with AutoCAD LT. I expect we'll learn a tremendous amount from that actual work, and that will give us guidance on how to continue to drive that evolution and evolving business model so that our partners embrace selling desktop subscription with all of our offerings.
We also want to maintain the focus on desktop subscription with our partners, and we recognize some of them haven't engaged heavily in desktop subscription yet. We're going to focus on helping them understand the value proposition. We're developing sales enablement tools for them and partner toolkits to be able to best position the value of desktop subscription as the best economic offering for our customers. As I mentioned, we have put in place an inside sales team, which we will continue to invest in and grow. Right now, from a desktop subscription perspective, those people are focused in two areas. They're either focusing on renewing expired desktop subscription customers, whether they purchased directly from us or through a partner. If someone hasn't renewed them and they expire, the inside sales team is contacting them to see if we can get them to re-engage.
We also recognize that quite a few, a large number of our desktop subscription customers are purchasing on an e-store, and there's no reseller involved. We want somebody responsible for driving the renewals there, and the inside sales team is focusing on that set of customers where no reseller has been involved in the sales process. I want to share something that's really interesting. Andrew showed you that 35% of our desktop subscription customers are new. That was based upon an analytical approach of us taking a look at parent companies and doing some data analysis. I'm going to show you a different point of view, which is also accurate. We actually had an independent company contact our North American customers and do a survey of, after you purchase desktop subscription, tell us, where did you come from?
This is from the customer self-selected point of view. What's interesting is that 56% of these customers who we spoke with say they were new customers to Autodesk, which means that 44% of them were existing customers. Peeling the onion back further actually tells us something quite interesting. More than half of the existing customers say they used Autodesk software, but they did not attach Maintenance Subscription. They were users, but non-subscribers, and they have now become subscribers in the model. That's goodness. That's exactly how we wanted this model to work. What's interesting about the 56% that say they weren't Autodesk users, a large chunk of them said they were using non-Autodesk design software. Good. That means we're penetrating into competitive accounts. That's exactly what we want to have happen.
A bunch of them said, "Well, I didn't use any design software." We see that as a weren't paying for the use of design software before and now have a business model available to them to monetize and actually pay for their use. If they didn't fit into one of these categories, we had other, but other was something other than being an Autodesk prior user. We're really excited about this information. Desktop Subscription is giving us the opportunity to grow the overall number of subscribers. It's giving us a competitive advantage to grow the overall base of business we can go after. It's hitting the mark of what we expected it to do. Let me move on to Maintenance Subscription. First, just a reminder.
Our partners who 10 years ago didn't feel so comfortable about Maintenance Subscription, do most of our Maintenance Subscription business today. They do a large majority of our Maintenance Subscription, and we need them to continue to focus on that business. In fact, we continue to use renewal rates and attach rates as requirements for the tiering of our partners. We are continuing to raise the expectation. We want the renewal rate to continue to go up, attach rates to continue to go up for them to maintain the higher tiering within our partner framework. This has been a key area of focus for our inside sales team. Like I said before, the first place we had them focus were on expired Maintenance Subscription contracts.
If a customer has gone through the process, their Maintenance Subscription contract expires, no one within the ecosystem has renewed it. That becomes an opportunity for our inside sales team. The inside sales teams go and contact those customers to bring them back into the ecosystem and to actually renew. We're getting great results there. We also recognize we don't have 100% attach of Maintenance Subscription on our sales. Any customer that purchases a new seat of software and does not attach Maintenance Subscription is now an opportunity for our inside sales team to contact to attach Maintenance Subscription.
We also recognize there are a large number of users out there, subscribers that don't have a reseller of record, either because they purchased on the eStore or perhaps the reseller they were working with has gone out of business, or they don't want to work with anyone. When we find those opportunities, we're also turning them over to our inside sales team. All these opportunities are being closed directly with the customers through our inside sales teams since the partners haven't renewed them themselves. This is something we're fulfilling directly. What's helping us as well as our partners drive renewal rates up is leveraging big data, this early warning system, and the analytics that Andrew talked about. By focusing on big data and by looking at analytics, we can identify early in the process where the high likely non-renewal subscribers exist.
We can start working proactively with those customers to ensure they are recognizing the value of Maintenance Subscription before the time of renewal so that they actually become renewing maintenance subscribers overall. This information, as it continues to expand and grow, will be used by our own teams as well as by our partners to drive those renewal rates up. Okay. We move on to the third type of subscription, which is our cloud services subscription. If you don't know this already, I'll make sure that you know we are investing and increasing our investment in dedicated product specialist teams focused on each one of our cloud offerings.
For BIM 360 or PLM 360, CAM 360, Sim 360, Fusion 360, InfraWorks 360, I have dedicated sales and technical folks that wake up every morning just focused on selling those offerings, engaging with customers, and driving these offerings to volume. We recognize our partners want to engage in some of these activities, which is great, which is why we introduced the Advisor and Referral Partner programs, as I talked about earlier. Here's where I want to check back in on A360. A360 is very exciting to us in that we see this as the cloud offering where we could attach it to every user in our customer base. That's all of our current maintenance subscribers and all of those 2.9 million non-subscribers that we want to bring into the subscription ecosystem. We recognize that by focusing this way on A360, we do two things.
First, we will drive deeper account penetration. We'll get to more users and turn them into subscribers within each one of our accounts, which of course adds more subscribers and attacks the 50 part of our equation. Each time we bring more users and subscribers in from each of these accounts, it works on that 20% metric of getting more value from each of those customers. We also recognize that this is a competitive differentiator for us, and we believe that by including A360 in our go-to-market approach with all customers sold by our own sales teams as well as our partners, that we have a better position, a more competitive position to compete in the manufacturing and in the AEC marketplaces. These new offerings are bringing in new customers. Andrew showed you a total metric of 30% of our cloud service customers are new.
I'm giving you some specifics here. I'm taking it 1 level deeper. Other than the first offering, which is HSM, this is a CAM offering, which is new. The others are BIM 360, Fusion 360, PLM 360, and InfraWorks 360. The green bars are the number of new customers, brand new to Autodesk. You can see Fusion 360, 84% of our customers are brand new to Autodesk. That's awesome. PLM is bringing in 36% new customers into the Autodesk ecosystem. HSM I put on here because the CAM offering is new for us. It's a new market. 76% of our customers are brand new. This is exciting. This is focusing on the 50. The growth of our subscriber base. I do want to highlight the gray bar. The gray bar are our existing subscribers who have now attached cloud services into their account base.
This now focuses on the 20%. We're getting more value from these accounts by having them attach and start using our cloud service subscription. I'm really excited about this. InfraWorks, by the way, if you're wondering, early days. We have a lot of traction in the infrastructure space right now, and we actually have been focusing on a lot of our existing customers because they're helping us work through scaling this and really solving some really big problems. I'm really, really pleased about the results so far. In summary, first of all, I want to say I'm really proud of our direct teams, my Autodesk sales teams, as well as our partners. The results they've driven so far this year have been fantastic. Our sales approach is best suited for the needs of the customers in each of the segments we're focusing on.
We're focusing on that partner framework, and we're focusing on ensuring that our partners are prepared to make the business model transition along with us. Our named accounts have great traction, and we're really moving to a space of having strategic relationships with these customers via leveraging the Enterprise Business Agreement. We have a long way to go. We're in the early days of that, and we have a lot of progress ahead, but great traction moving forward. Throughout this year, as you've seen, maintenance subscription has really led the day as far as driving net new subscribers into the ecosystem. We absolutely expect that desktop subscription and cloud services are going to become big areas for subscriber growth in FY 2016 and beyond.
Thank you very much again for joining us here, and it is my pleasure to introduce to you my finance business partner who's going to focus on the financials, Sue Pirri.
Thanks, Steve, and good morning, everybody. I have the distinct privilege of starting my presentation with something everybody's familiar with. The legal and finance guys have asked me to remind you of the safe harbor. This morning, we've made a lot of forward-looking statements about our business that's subject to risks and the uncertainties that are spelled out in our 10-Ks and 10-Qs. With that, this morning I'm going to cover the business model transition and how we believe that it's going to drive shareholder value for Autodesk. We'll talk about some of the key metrics that we think are important for people to focus on as we drive through the business model transition. We'll also talk about our capital allocation strategy.
That's something investors have a lot of questions about. We've spent a lot of time, we want to bring you up to speed on where we are there. We'll talk about this business model transition, specifically how it will impact our financials, then talk about the guidance for both this year, this quarter, as well as the longer-term guidance numbers. Before I get started, I want to remind everybody and do a quick level set on the strong performance we've had this year already. About a year ago, we set out some long-term guidance for ourselves around the business model transition here at Investor Day. As we entered FY 2015, we did what we always do, lay out our targets for you for what we think we'll do this year.
On several of our key metrics, billings growth, revenue growth, our net subscriber additions, we've already significantly raised our targets for the year. We think that that's a sign that we're off to a fast start with the business model transition and this fiscal year. One of Autodesk's key advantages in the market is that we're so diversified. We have diversification across industry segments. You heard Amar talk this morning about some of the things in AEC that are driving our strength. We've got new product offerings. Strength in the commercial construction business is coming back as well. We're entering new markets there. You heard Buzz talk about some of the things in manufacturing that are driving the business. Suite strategy has been particularly effective in the manufacturing market. We've got new markets that we're entering there as well.
When I think about Q2 performance, we had really strong results. Amar mentioned 23% revenue growth in the AEC space, 17% growth in the manufacturing space. I want to mention that platform solutions, which is our horizontal AutoCAD and LT products, grew 5% in the quarter. Strong growth for products that are our historic base. As well, we're well-diversified by geography. Now, these numbers are for FY 2014, but you can see that Americas and EMEA are about the same size, APAC's a little bit smaller, but well-diversified across the geographies. One of the things that was great in the second quarter is we saw double-digit constant currency growth from every geography. Really strong performance. Net billings is something that we really started talking about last year as we introduced the business model transition.
It's something that we think is an important metric for you to monitor the strength of our business. It'll show the underlying strength and velocity of the business in a way that's more reflective of what we're doing than revenue is right now as we go through the transition. This slide shows net billings for the growth for the first half of the year. We started talking about the growth rates in Q1. We'll continue to update you on growth rates on an ongoing basis. I think the exciting thing here is that you can see we had 18% growth in net billings in the first half. That's based on strong license growth as well as strong subscription growth. The subscription growth has been mostly driven by maintenance subscription because that's a big part of the subscription model right now.
We're beginning to see desktop and cloud come on and have some meaningful impact. The strong growth in the first half gives us confidence in our 10%-12% billings growth number for the full year. Subscription revenue is an important part of the business model. You can see we've had steady growth over the years. This is a five-year model. Our subscription program, the maintenance subscription program particularly, started about 10 years ago and has shown steady increases over that time. You see we grew 7% last year and 14% in the first half of this year. That's based on strong growth in, particularly, the maintenance subscription model. Actually, in fiscal 2014, we passed the $1 billion mark in subscription revenue for the first time, predominantly on maintenance subscription.
As I mentioned before, as desktop and cloud begin to accelerate, we'll see strong contributions from those, and we believe that our subscription revenue growth rates will increase. Switching gears a little bit, I want to talk about the balance sheet. It's a strategic asset for Autodesk, and it's as strong as it's ever been. In the second quarter, we saw strong growth in deferred revenue. In fact, we had record deferred revenue on the balance sheet. We also had DSOs of 52 days, which is a great result for a company that's as globally diversified as we are. On an ongoing basis, we've had strong cash generation. We have $1.4 billion of cash and investments net of our debt. Channel inventory is something that we've worked on over the past few years to bring those balances down through things like electronic distribution and different programs.
Our channel inventory levels are near their historic lows, and we expect them to remain low. Let's talk about some of the balance sheet metrics in a little bit more detail. I mentioned deferred revenue was at a record high. It grew 22% in the second quarter compared to the second quarter of the previous year. That's predominantly based on maintenance subscription, which is the biggest part of our subscription base. As I mentioned, cloud and desktop are beginning to have contribution there. As we successfully transition more of our customers and users to subscription models, we expect that deferred revenue will continue to grow. It's currently about $1 billion. Autodesk has been a strong generator of cash over the years.
You can see in this slide, in fiscal 2010, when we were having a bit of a business downturn, even in that time, we generated positive cash flow. As the business recovered, we saw strong growth in our cash flow, and we've maintained that over the years. We expect the cash flow growth will continue to be strong. We had 9% cash flow growth in the first half of the year, and we expect it will continue to be strong throughout the year. The exact finish at the end of the year will be a little bit dependent on billings linearity in the fourth quarter. One of the things that we watch very carefully and think is a really important metric to measure the success of our business is our cash flow margin. Operating cash flow divided by revenue.
It was 26% in the first half of fiscal 2015. That was an increase of 1% over the previous year, which was an increase of 1% over the previous year. We're seeing steady growth there, which we think is a sign of a healthy business and a positive attribute of the business. As well, our cash and investments balances, which I mentioned, were $1.4 billion net of debt at the end of Q2. This is one of the prize assets on the balance sheet. I should note that 78% of that is located offshore. I'm going to switch gears a little bit and talk about our capital allocation strategy. It's something we get a lot of questions from investors about and spend a lot of time thinking about. In addition to funding our ongoing operations, we have two main uses of our cash.
It's investing in M&A and our share repurchase program. As well, we've got a note here on the bottom. We believe that debt is a healthy part of every efficient balance sheet. To that end, nearly two years ago, we issued $750 million of debt, and we believe that debt will be an ongoing part of our balance sheet. If we can talk about M&A for a little bit. M&A strategy at Autodesk has centered historically around small tuck-in acquisitions or technology purchases that add features and functionality to existing products in a low-risk way. We don't like to buy big businesses or do mergers of equals. We believe that our sweet spot is with small tuck-in kind of acquisitions, and that's what we've done very well.
You can see from this slide, many of our major products have had some impact through features and functionality or even been acquired. If you look at the slide, you can see AutoCAD here, Inventor. Revit was an acquired product, 3ds Max, Maya, all in some ways were impacted by acquisitions, either through features and functionality or actually acquired. Our M&A strategy has had a lot of impact and effect on our growth and success in the market. It's really helped us drive a very broad and deep product portfolio. M&A will continue to be a part of our strategy going forward. We do expect to continue to do M&A, but I should note that fiscal 2015, the volume of M&A has been a little larger than it's been in the past. Buzz talked about Delcam.
It was a little bit of an anomaly in the typical kind of M&A that we do. It's a strategic asset in a new market that we thought was very important. Price was nearly $300 million. It was a little different than what we've done in the past, and we don't expect to complete many of these going forward. M&A will continue to be a key part of our strategy. As well, over the past few years, our M&A strategy has begun to focus more and more intently on collecting assets that will accelerate our move to cloud and mobile. As you can see in this slide, across the board in some of the cloud technologies that we've got and cloud offerings that we've got, we've been adding pieces of technology and new offerings through M&A that are driving and accelerating our cloud and mobile strategy.
This has been a key investment area for us over the past few years, and we will continue to invest here with particular focus. The next part of the capital allocation strategy really centers on our stock repurchase program. We have two main goals in the stock repurchase program. The first is to return excess cash to shareholders, and the second is to reduce the share count over time. As you can see from this slide, in the last two years, more than 80% of our free cash flow has been returned to investors through the share repurchase program than our operating cash. In fiscal 2014, 10.5 million shares were repurchased at a price just under $425 million. It's been something that we've been very focused on continuing to return cash to investors through that efficient mechanism.
We've had a lot of questions recently about currency volatility and what our hedging program is. I've got a couple this morning as a matter of fact. I wanted to remind, I know many of you are already familiar with this program. Our hedging program has been in its current state for several years. Our policy is not to be speculative when we do our hedging. It's merely to try to lock in FX rates to reduce foreign currency fluctuation risk and help with our forecasting process. We hedge net exposures in our major currencies, and we use a classic four-quarter rolling hedge process. It's simple and effective. Switching gears, I want to circle back on the business model transition.
We've heard a lot today from members of the team talking about the things that we've done to accelerate the transition to cloud and mobile and even desktop offerings. I mentioned earlier, we've already raised our revenue growth guidance for the year, our billings growth guidance for the year, as well as our subscription additions guidance for the year. We've seen meaningful increases in attach and renewal rates for our maintenance subscription program, which is really important going forward. The more of those customers that we keep in the Autodesk portfolio, the more customers we have, more services to sell to, we can move to cloud and mobile and get them onto recurring revenue streams. As well, I mentioned deferred revenue reached record levels in Q2, through all of this transition, we've been increasing our cash flow.
We've seen early success in the business model, the plans that we laid out for you last year at this time. That success has led us to have confidence in the guidance that we issued for Q3. We believe that's on track. We reiterated that this morning. As well, we reiterated our guidance for fiscal 2015. I want to call specific attention here to the net subscription additions, 200,000 to 250,000. We believe that's on track, and you heard Andrew mention this morning that we'll do at least that many next year. As we think about the Autodesk opportunity, we believe it's vast. We believe that growing double-digit billings growth will be attainable to us through a number of venues. We've got a large addressable market where we can offer best-in-class offerings in our core business. We're increasing services to existing users through products like A360.
We're also expanding into new markets. Steve just had a slide up here with PLM 360, Fusion 360, CAM 360. We're reaching all kinds of new users through extending our product portfolio. As well, we've been establishing leadership in cloud services. We were the first mover in our industry to cloud, and we believe that we have a significant advantage there, and we will continue to invest and drive that. That optimism translates into a final confirmation of some of the guidance numbers that we gave you last year, our long-term targets. You've heard a lot of talk this morning about the 12, 20, and 50, 12% billings CAGR through fiscal 2018, 20% more value per account, and 50% increase in subscriptions. We believe we're on track to deliver that. As well, I wanted to talk about three other metrics that we mentioned last year.
70% plus recurring revenue by fiscal 2018. We believe that's entirely possible. The speed and the final level of that number will greatly depend on the strategic decisions around perpetual offerings and those ends of life. As well, we believe we'll continue to see a significant increase in the percent of ratable revenue as a percent of total as we move to subscription models and more of the ratable models. Finally, we remain committed to our operating margin at 30% by the time we leave fiscal 2018.
We're on track for the guidance that we put out last year on the business model long-term growth targets. We think that the company is executing well against those and delivering faster than expected already. With that, I will look for Carl. Thank you, everyone.
Thanks, Sue. Okay. Hopefully, we've given you enough nuts and bolts, and I'm sure as we get to Q&A, we'll follow up more on the 12, 20, 50. Hopefully, we've given you sufficient detail so that people understand why we feel like we're on track with 20 and 50 and what we're going to do in future years to continue to drive the progression to our goals of 12, 20, and 50. I just wanted to remind you a little bit. Hopefully, while we've provided enough detail, if you step back and at any point want to understand what we're trying to do, I think it's important for you to look at what strategic decisions we've made. One of the ways we go about planning for the future is to take a point out on the future.
We form an opinion about it, and we go after that. Certainly, we course-correct along the way. Many of the decisions we've made have been in the works for a long time. Whether it's about our cloud and mobile and social strategy as the new computing platform, what we've decided to do to compete more effectively by a more open approach, the markets we choose to compete in or the business model are all long-term decisions. We don't wake up that way. As you go through and look at the things we're doing, it's certainly worth digging into details and understanding the nuts and bolts. I urge you also to use this other lens to look at what we're doing, and you'll better understand some of the things that are going on with Autodesk. Talked about at the beginning, there's the opportunity.
We think the opportunity is growing. As you heard, there's secular growth in these markets. We think there's ability to grab more market share, as well as in many of these industries, their spend on IT is going up considerably as the solutions provide more value. As we look going forward, one of the things that we've spent a bunch of time on is trying to choose the areas in which we want to compete. One of the most interesting areas for us is we've spent most of 30 years helping people design and engineer things. If you look at the complexity, if you look at these two images, one, a stadium on the left, the other is an airplane on the right.
The complexity of actually getting these built is where a huge amount of the money is spent and where a huge amount of money is lost and where a huge amount of value can be recaptured by doing this better. One of the areas we focused on particularly is how things are going to get fabricated in the future. We're really interested in how we can help our customers make things. You may have heard us talk in the past about additive manufacturing. Buzz mentioned it today. By the way, for those of you who haven't seen it, we announced what we're doing with our Spark 3D printing platform, as well as our Spark printer. One of the printers is out there, so if you want to see the Autodesk 3D printer, that's our first foray into hardware ever.
You can go out and see the 3D printer out there. That's just part of a larger effort, along with what we're doing with CAM, what we're doing in carbon fiber and composites to help our customers manage the process and manage the materials and be able to better realize the things they've designed and engineered. With that, we'll end the prepared remarks. Let me invite everybody on the team back in here, and now we'll spend the last bit of time with some Q&A. I think Dave's going to organize the Q&A. Otherwise, I'll start calling on people. Let's just wait for them to bring in the chairs. A little late. We're good. It's a little worrying with musical chairs here.
All right. Good morning. Jay Vleeschhouwer . Two questions, perhaps mostly for Steve and Andrew. First, could you share your thinking about what your annual new license volume opportunity or equivalent thereof might be, particularly as compared with your historical volumes and specifically for LT? When you last disclosed your annual volume about 6 years ago, you were doing about 600,000 or more new licenses a year. That's probably dropped now to maybe 500,000 or so if you think about the drop of AutoCAD and the mix shift to Suites, but still a lot of volume, and half of that volume, as you said, Andrew, was LT. Maybe talk about how you think about the new volume opportunity.
Secondly, could you talk about the operational issues pertaining to the wider availability of the usage-based model beyond just the largest accounts, and why that model wouldn't be the prevalent post-perpetual part of the model rather than rentals? When you think about the break-even periods, for example, for monthly and quarterly rentals, they might not be that compelling for a lot of your customers where the usage-based model might be more compelling. Thanks.
Okay. That was probably eight questions. All right, let me dodge the first question and answer it at the same time. One way we actually get a lot of insight to what we think the volumes are going to be is what we've been doing with some of our direct experiments, especially around the eStore. Let me give you a number to ponder as you think about what might happen to our seat volume. As we rolled out desktop subs onto the eStores, and let's just use the microcosm of the Americas eStore because it's our most vibrant and active eStore right now. We saw a doubling in seat volume. Basically, a kind of a steady state volume of perpetual licenses, and then an additional uplift of desktop subscription licenses being sold.
That's not going to presage exactly what's going to happen to the total seat volume as it runs out moving forward. Because we're able to bring in new buyers like this, and we're able to attract new activity, you're going to see those numbers change pretty significantly. Rather than try to guess what that change is going to be, I refer to these experiments because they're real in-the-market experiments that we do to understand exactly how the dynamics are moving forward. Yeah, some of those numbers, as we've learned from some of these experiments, end up baked into our models and what we view as possible. There's kind of a half answer to your question. Okay? The other question was operational challenges around the consumption model.
Yeah, the other question was.
Yep.
The issues between rolling it out and what point does that buying pattern become the larger part of the model?
Well, it certainly has. When you roll out an offering like that, it has a huge potential to be a very popular, very prevalent offering. It could be the next generation offering we have. Maybe next year, we'll start talking more about what we've learned in terms of rolling it out and trying to drive that to volume. It's absolutely got the potential to replace the buying pattern when you roll out a model like that. Not everybody's going to want a model like that. It all depends on how it's packaged and delivered. Yeah, it has huge potential to become a new model moving forward and a new preferred buying model. Yeah, operationally, there's absolutely challenges because if you're deploying an access and consumption-type model like that, you're going to have to be able to track the consumption on a very large scale.
We're definitely working to do that.
Hi. I got a quick question on the 20% portion of the 12/20/50 plan, which is when I look at the AutoCAD 360 offering, which has a potential of increasing $10 a month versus the $47 a month you guys are doing today. When I look at the customer lifetime value slide you put for AutoCAD LT potentially going up by 30%, then I look at just potential price increases that you could have on traditional license and maintenance model. Why do we think about 20% price increase for a customer? Why can't it be larger than that? What are kind of the pressing factors?
You're right on the math. I think right now, these are new experiments, these are new offerings, and we'll see how they go. We've put out what we think to be as reasonable guidance given where we are, but as we learn from them, as we've shown already in the first year, we've already upped the guidance considerably. If it proves out that these models work as well as they should, we will continue to do so. We'll continue to not only meet but exceed the guidance we give, and we'll correct it as we go. But yeah, most of the activities there speak to an ability to drive more revenue, in particular billings, and drive more to the bottom line.
Awesome. Thank you. This is Keith Weiss from Morgan Stanley. Thank you guys for having us, and definitely very impressive progress in the business model transition. Just a couple of definitional questions for you guys. One, just an understanding, if we're getting rid of perpetual licenses over the next 12-24 months, why isn't the FY 2018 target 100% recurring revenue? What else is going to be in there that's not recurring? Part 2, I'll just throw it out there, is in terms of the renewal rates, definitely when we do our models, renewal rates is a key factor in terms of driving that billings growth. Can you give us an idea of where renewal rates are today in your subscriptions? How close are we to that 90% level to help us sort of benchmark and sort of level set in terms of our models?
Sure. I can do number 2. You want to do number 1?
On the recurring revenue, we've got our plans on how we're going to roll out the perpetual changes, and we haven't announced those to anyone. We're not ready to do that. We don't want to change our guidance because we don't want to update and give out those plans. Certainly, with an end of life of perpetual 24 months out, it could certainly be more than 70%, but we weren't ready to give a more updated number at this point in time.
I know you shook your head, so Keith, yeah. Yes. If it is completely gone at that point in time, it will exceed 70% by a considerable amount.
Yeah. Okay.
Yeah.
If perpetual licenses are done in 20-- There's nothing definitionally that we're not understanding.
No.
There's no other revenues.
We just didn't have a better number. You can really get close to 100%, because if you think of the offerings available at that time, every one of them should be recurring.
Consulting.
Consulting's the one which we've just kind of given you a rough figure at about $100 million. You can take that off. There may be some other nits and nats around, the majority of the revenue will be recurring, the ratable still hangs a little bit on the new accounting standard.
Renewal rates?
What's that?
The question on the renewal rates?
The renewal rates. We haven't given out, but in most places, you should think of them in the mid to high 80s.
Thank you.
Thanks. It is Brent Thill with UBS. Just as it relates, Carl, you have talked a lot about the channel differences between you and Adobe. When you look at the model that is progressing, Steve, thanks for the color on how this is going to build out. It just seems like over time, there is no question that you can drop the % of sales and marketing, which has been pretty heavy relative to other software companies we all cover, down as a % of revenue. Can you just walk through your assumption on sales and marketing, where you find the efficiency in this model as you make this transition? It seems like if you do have resellers drop off, it feels like there is more revenue that is going to go direct to you, which means better margins over time. How do you see the progression of this?
I think there are a couple things in there. I think that was Brent who was asking somewhere. Yeah. Yes, there are a couple of things to think about as we make this transition. One is I think overall, just our sales and marketing numbers are high. Regardless of any model change, I think they are high relative to the business that we run. I think if you look at how much we spend ourselves plus our partners spend, I think it is a number that we can manage down over the next few years. Other things in there that will contribute are things like increased volume makes the hugest difference. Channel mix will make a difference. New models will make a difference.
There are a number of factors that give me lots of confidence that we will be able to lower the % of spend on sales and marketing.
I would not expect I was just going to add something.
Yeah.
Don't read into my statement that if partner consolidation continues, that means that everything goes direct.
Yeah.
Consolidation means that as some smaller partners get consumed by those larger partners, the business stays indirect but goes through those larger partners, too. I didn't want you to leave with.
Yeah
A misinterpretation of that point.
Great. Thank you. Heather Bellini with Goldman Sachs. I had two questions just to follow up on Brent, because you did have a big jump in OpEx this year for reasons that you guys have discussed. How do we think about the total, your thoughts around total OpEx spending levels when you think about it on a growth basis as you look out, what are your thoughts on that? I have a follow-up.
I think the first one is, I think you should think of the growth rate this year, we talked about a number of reasons.
Yep
Including Delcam, bonuses and commissions coming back from down last year. There are a number of reasons. The other one has been investment in cloud infrastructure. I had a little discussion before with folks, and I said, the way we think of our OpEx growth rates is that kind of a nominal growth rate for us is around 4%, that all other things being equal, you could start with a bogey of 4% and work from there. I think you'll see our OpEx growth rate moderate over the next few years, as well as our M&A spend. I think you'll see both of those come down in order to hit those targets.
Okay, great. The follow-up would be, given the initiatives around the elimination of perpetual over the next two years, plus the rapid adoption, it seems of the cloud offerings that you discussed today, is there any reason why the sub adds can't be actually considerably better than the 200,000 to 250,000 you've guided to this year and you've mentioned as a target for next year?
I think they can. I'll let Andrew on the hot seat.
Okay.
At least that level.
Certainly, they can be better. All right? There's no reason for us to Look, the model's solid. We've got a real sense for where the money's coming from, where the customers are coming from, the hit to 12.20.50, so we believe in it. Do we look at these now and look at the performance and look at some of the things we're doing and say, "Can we exceed some of these metrics?" Yes, we absolutely feel we can exceed some of these metrics. As we progress, as we learn more about what happens next year, we may start talking to you about revisions or updates to some of those. Absolutely, there's potential.
Yeah. As we get to our planning for next year, we'll start figuring in. We felt it was just a little premature now. I think the first half has been stronger than expected, we'll update you as we move into next year.
Okay.
Hey, thanks, Carl. You guys are arguably well-positioned better than just about anyone out there, but you don't operate in a vacuum. Could you talk as best as possible, how you see what's going on, your competitors kind of a range on, everyone's doing a range of reaction and response to what you're doing because it feels like you're first mover in a lot of areas. Just at least at a high level, think about that because you have to care about what other people think too, and do.
Yeah. Tell us about what they think. It's a personality flaw. If I was to look out there, I'd say let's start with manufacturing. My friend Buzz will jump in. What has surprised me most in manufacturing has been the hesitancy for any of the companies to do anything dramatically different.
I see most of the companies doing more of the same old thing they've done before. In a couple of cases, you see them reaching out to slightly new markets, but with an old business model and old technology.
Yeah. I think they're trying to protect an old-fashioned, expensive, labor-intensive business model.
Yeah
It baffles me why they haven't.
The interesting thing is at a time of transition like this, you can always get signs that tell you that the old model is good. You can always find the recalcitrant customer sitting there going, "I'll never move to the cloud. Mobile is stupid. These devices are insecure. We'll never do anything." There's a million reasons why not to look forward. You can get some great comfort in talking to your most stubborn customers about why the past really is the future. I think some of them are just having too many drinks with their customers and talking about the way it used to be.
Current spend also goes to the old model.
Exactly. A huge amount of the current spend. You can confuse yourself. I think Buzz's examples of some of the Fusion customers, some of the ones we're looking at, these are new companies. Even inside the mainline old companies, they're all looking to these new companies as consultants and to bring in the techniques that they have to bring products to market. I don't see a lot of move there. My guess is what will happen in manufacturing, in particular, is that there will be a number of startups that come into the field. People will do analysis online, people will do simulation online, they'll do design online. A bunch of these companies will wake up kind of late, realize that they better do something, and they'll be forced to overpay in some acquisitions to buy something.
I don't take great comfort in we're going to be ahead of the pack forever because people will go out there, and they'll wake up, and most of them have relatively strong balance sheets and can go spend lots of money. I think it will be a game of catch-up. I think the one thing that I feel confident in is that we've proved that the direction we're headed is the right direction. Whereas three or four years ago, you could've been a little bit more suspect about whether or not engineering could move to the cloud. AEC, I think, is a little bit more interesting. It's always been a lot more fragmented regionally. The only things I see different there are lots of new entrants, just like in manufacturing, but doing very different things. I think that's what's exciting.
It's these companies who are coming in and saying, "There's money being spent on the maintenance and operation of buildings, and we're going to put sensors and software, and we're going to put in data collection and analytics to understand how to run these projects more efficiently." That's where they're focused. In some ways, it makes the most sense. There's new opportunity. It's kind of a green field to go into. There are a handful of people like Trimble, trying to buy some of, I'd say, last generation's technology. Look, I don't want to take the high horse too much in that case, plenty of companies have proven you can sell old technology for lots of years and make money on it. That's never been what's interesting to us. Amar, you have others?
You said it really well. I don't have more. I think the other, you mentioned Trimble, there's Bentley, that's another company. That's another model.
Yeah. I haven't seen a lot of movement there. What is the most interesting to me is the fact that there are lots of startups coming, because there's so much money in the venture community right now. What do they say? First, it's the disruptors get funded, then the copycats get funded, and then the stupid gets funded. We're somewhere between the copycats and the stupid right now getting funded. The good news for us is that there are a lot of experiments getting funded in both manufacturing that's new and different, as well in stuff for the AEC industry.
That's good.
Thanks, Matt Hedberg, RBC. LT is obviously a huge component of the 2.9 million active customers you're trying to convert to maintenance. I'm curious, certainly appreciate the 30% increase in value on an annual percentage. That's very helpful. I guess from modeling purposes, can you help us understand what the contribution, from a margin perspective is from an LT customer paying a perpetual versus desktop subscription?
Margin?
Margin.
We don't break out margin by product at all. If you think about it'll be a more direct model, although we'll start with partners, you can sort of apply the same principles, we don't break out specific information by product.
Maybe a quick follow-up, given the new Token Flex pricing, which seems interesting. I think you have 10 now active on that. Can you give a sense for the relative size of that opportunity inside the base? What could that 10 turn into? It seems like it's been very successful thus far.
I can take that. We think there's an opportunity to move all of our named accounts over time to a consumption-based model. It's just a matter of time and their needs growing. The more that they start using more of our products, the more they benefit from that overall. We have a goal of moving all of our named accounts there. It's just going to be a matter of time now.
You want to quantify how many named accounts?
Yeah.
Hundreds.
In our named account program, we have varying degrees of focus. I would say there's several hundred named accounts that are targets for us right now with our current Token Flex consumption-based model.
All right. I'd just like to ask about how the cloud may play a role with your core desktop customers today. How does the cloud get involved in their lives in the future, and do you see that ultimately migrating to the cloud?
Yeah. I think the cloud becomes the predominant platform for our customers. Kind of full stop. For anyone who thinks it's some kind of cute add-on or addendum, or it's not where the real stuff happens, like I said in my opening remarks, that's what they said about the workstation, it's what they said about the PC, nothing real would ever happen here. I still remember my trip to this automotive company that insisted they had to build their own workstations in order to be able to do engineering work. The first thing I'd say is that the cloud becomes the platform of the future for people doing architecture, engineering, and design. If you want to understand the reasons why, there are two main things that I think are really important about the cloud. Possibly three, but there's two principal ones.
One is the ability to have totally scalable, elastic computing. As much computing power as you want at any point in time. As I've talked about before, it requires a change in tool set and a change in mindset. Most people don't go about their daily job of saying, "What if I had all the computing power in the world? How would I do this differently?" I think for those who do, there's huge opportunity ahead in recognizing that we're totally underutilizing the power of computational design to accomplish what many businesses want to do. If you just understand that at any point in time, I can scale up to really almost an infinite amount of computing at relatively low cost, and I can scale down when I don't need it. This first thing is this notion of infinite computing.
The second idea that's really important, it probably manifests itself best in what we're doing with A360, is this idea that almost none of our customers build, design, engineer things by themselves. Every project that we're involved in or our customers are involved in is a collaborative project. This idea of having a central coordination place where you can understand as well what's going on with your construction project, with your manufacturing project. If you can understand that as well as you understand your neighbor's kid's soccer game, as we do on Facebook and Twitter. It only makes sense that this is going to become a primary way people organize and coordinate their projects. It's this combination. The third one, which is interesting, was touched on by both Andrew and Steve, is the old model of delivering software is going by the wayside.
People don't want behind-the-firewall installed software, whether that's at the desktop level or at the enterprise level. It's just a model of the past. Nowadays, I don't have anything on my machine. The two requirements for me is that I don't install software, and the second one is that it's not behind the firewall. I think if you're working with applications that don't fit that, then it's one foot in the past, one foot in the future, because I think the applications of the future all look like that. I think the cloud is an enormous contributor to what our customers do going forward.
I have just a quick follow-up question on your eStore initiative. Where are you in terms of your capability there? That's something you're spending money on. It sounds like it's still under development. What role do you see that playing in the business going forward?
Yeah. The eStore is a couple of things to us. One, it's a direct channel that allows us to talk directly to our customers. Two, it's also a way for us to test, deploy, and push out new business models. Three, it's also a way for us to drive behavior and interact with the channel in specific markets. In our wildest dreams, if you look out, how many ever years you want to look out, maybe it's a $300 million channel. Its impact on the business will likely be more significant than that because of what we use it for in terms of testing new models, deploying new models. Jay, you speculated on consumption models going down the line. Where would we end up testing those first? Likely through the direct channels.
The direct channels also allow us to build out certain things in our back office that enable us to do more interesting things with our partners around digital distribution in the future and other aspects of that. It's actually a pretty strategic aspect in terms of the moving forward, regardless of what its scale is ultimately.
Walter Pritchard, Citi. I think you answered a lot of the question I had there on the first part of my question, Andrew. When do you get out of the business of versioning software and Adobe sort of moved to a version-less, every couple of months drops features, people pick it up as they just open the applications up. That's quick question number one. Question number two, just on the disclosure. If I think about this modeling, billings are more helpful than revenue, probably still not that helpful in next few quarters or next few years because of upgrades and perpetual licenses and all the shifts going on around that. Subscribers, you've got so many different kinds of subscribers that one subscriber isn't really the same as another, we can't really count subscribers.
I wanted to see if you guys have thought about ARR as a metric and recurring revenue and potentially disclosing that. It feels like Adobe is kind of shifting away from subscribers towards ARR and wondering if you're kind of thinking the same way.
Yeah. One question I'm going to defer to.
Yeah, there's at least three or four questions there.
I think the first part of your question is how far along are we in terms of developing the always updated experience. We've started that process, and as the desktop subscription experience rolls out by product, by region, you'll see it reflected in products. We made a first stab at it with what we did with desktop subscription last year. We have lots of ideas on how to improve it, how to expand it, make it more seamless, make the upgrade just like what you experience on mobile devices. We're well on our way, and next year, I think we have a lot of work to do on making that a superior experience that Andrew talked about come to life. I think we're going to do it across products as they roll into desktop subscription.
The other thing to do, I was going to say, is we have a handful of products. You brought up Adobe. One of the things to recognize is we're doing something fundamentally different in that many of our products are actually cloud-based products. What Adobe is doing is digital delivery of desktop products, and the updates that you talked about are updating desktop products. If you were to look at our PLM products.
Autodesk Fusion products
Autodesk Fusion products
A360.
A360.
Right.
Those are all true cloud-based products in which the drops look more like weekly, or every few weeks than even monthly, and certainly not annually.
They're really version-less.
I think you need to understand that. Now, in the end game, I think us, Adobe, Microsoft, Intuit, we'll end up in the same place in terms of business models and how we deliver software. At any point in time, we may not be traversing the two axes at the same pace. Mostly what they've been working on is digital delivery of desktop software. To the extent that people continue to get desktop software, we'll improve that experience, as you mentioned. The other one is to recognize we already have lots of customers getting true multi-tenant cloud-based software.
Right. It is a huge benefit for customers, too, that they never have to worry about versions of data. There's no backward/forward compatibility issues ever. There's one version in the entire world.
Let's see, what other questions?
Well-
ARR.
Yeah.
ARR.
I'll just take a stab at that, and I'll turn it over to Sue. Obviously, as the business matures, we're going to look at things like ARR. I want to challenge your statement about billings and subscribers not being meaningful. By the way, Adobe just started transitioning to look at ARR as a driving metric. Absolutely, if you want to look at the health of our transition in the next 12 to 18 months, you're going to be looking at the subscriber numbers, and you're going to be trying to understand the subscriber numbers and see that subscriber growth. That is by far going to be the best indicator. As we move beyond that era into other eras, you'll start looking at ARR and other things like that. I will not discount the importance of looking at those subscriber numbers over the next 12 to 18 months.
They're critical.
Yeah, like always, it's not just the nature that we have three different kinds of subscribers. Even within one bucket, not all subscribers are equal. Sue will kick me if I get into my ovary and testicle jokes again. I won't talk about averages again.
Just had to work that in, didn't you?
It's just an average. Even if you were to look in maintenance subscription, the difference between an LT maintenance customer from a Product Design Suite Ultimate, Yep, is pretty substantial. Yep.
Hey, Saket Kalia from Barclays. One question and one follow-up, if I can. Lots of talk about the 12/20/50. I think you gave a lot of reasons why you could grow that 20 part, the value. I guess, what is the line in the sand that we should use as the value per customer, or however metric you define it, that will grow 20%? Because I think we all have a great handle on what the subscriber number is and how that's going to grow 50%. What is that line in the sand to grow 20%?
I think you need to look at value per account. That is the best indicator and the most reliable metric we can calculate. We've spent a lot of time looking at this, we think if you just look per account, if the value goes up by 20% per account, that's the best way to understand it. That's what you should hold us accountable for. To the extent that's the line in the sand, yeah, subscribers are equal, easy to count given the previous caveat. On this one, it really is value per account, and we'll report on that.
Yeah, we'll start reporting a metric that defines that. If you remember when we went through the Suites transition, it wasn't until the second year of the Suites transition that we started actually showing where our progress was to the ASP targets.
We'll do the same thing with this metric as well, and we'll actually give you a line in the sand, a benchmark, and you'll be able to track our progress to it.
That's really helpful. The follow-up is for Sue.
Sue, you talked about the cash flow margin as a good indicator. Any idea where that could be by the end of this transition, or even if it's an absolute cash flow number, however you think about it, because presumably with the more recurring business, the cash flow should benefit disproportionately.
Yeah. It certainly does, and as we do our modeling, we see strong increases in cash flow throughout the transition and certainly in the end period. We're not ready to give a different metric on that. I think you should think of it as something that's going to increase over time.
Is that the absolute cash flow number or the cash flow margin that-
It's your margin.
Margin. Got it. Thanks.
Steve Koenig with Wedbush. Just stole the microphone, thanks.
Obviously the modeling of the optics is an enormous challenge for the analyst. The optics don't matter too much, except it's the headline for our model. I do want to ask you some things that impinge upon your trajectory towards the long-term targets. I know you're not ready to give guidance for perpetual license next year, but how do we think about that? In particular, one question that might bear on that, did you say when the LT perpetuals will become available? Is that the end of Q4 after you do your experiments? Then I've got one related follow-up question.
I didn't quite understand the question.
The LT perpetual licenses, when exactly do they phase out?
Yeah, we haven't given a date yet.
Okay.
What you did hear from Steve is that we're already in the process of shifting the margins so that they're neutral between the perpetual and the Matter of fact, it's actually better. Within neutral, it's better for the partner.
Desktop
a desktop subscription for the long-term health of their business. We're already making some pretty dramatic moves there. In terms of how we end selling perpetual new seats, that will be staged, and we haven't revealed a date exactly when that process has finished.
One of the Dynamics that I think you will see play out is as we announce phasing out of perpetual licenses in a particular geo or by product, you will see a rush to buy.
Absolutely
more perpetual licenses.
Yep.
Particularly customers who already have a fairly large installation, but even some new customers who say, "That is my preferred method of buying. This is my last opportunity. I want to buy more perpetual licenses." I think there'll be add-ons to existing. That's the dynamic that you'll see. A little hard to model because we haven't revealed the plans of which country, which product quite yet.
Okay. One clarification question on the LT margins that you'll be paying to the resellers. I think you said that the resellers would be made whole in terms of absolute margin dollars between a perpetual and a subscription. Is that over the course of three years, or is that in the first year that they're going to receive the same margin dollars on either sale? Then you said renewals as well. That sounds like a very expensive proposition. How exactly does that work?
I'll address that. What we're doing is we're changing the margin so that the absolute margin dollars are the same in selling a new seat of LT perpetual or selling a new contract for desktop subscription for LT. My comment about the renewal is oftentimes, when you sell a perpetual seat of LT, that's the last sell you get for quite some time. Our rate of attach of maintenance subscription there is lowest of all of our offerings. Desktop subscription should go over and over again. Our partners should have the opportunity to continue to go and drive those renewals. I wasn't talking about the renewal absolute dollars. I was talking about the new seat sales. There's a benefit to our partners to continue to renew those contracts compared to our current state of selling a perpetual LT license.
Steve Ashley, Robert Baird. I just want to ask a quick follow-up to that. Is the payout going to be the same on the renewal as it is on the initial sale?
The current payout is not the same. I don't expect the updated renewal rate dollars to be the same, they will be better than they are in the today state. By the way, I'll add, we think it's really important for every customer to get that first renewal. We find that once a customer acquires a license and renews one time, they become really sticky and recurring and things. It's very important to us that we not just get that first sale, but we get the first renewal. Typically, when that happens, our renewal rates are very high. We want to make sure that we get not just that sale, but the first renewal.
Hi, Gaurav Kapadia from Soroban Capital. First of all, congratulations. Seems like a tremendous amount of progress has been made since last year, that's, I think, very exciting. Just one question on the 20%. Now, as we think about it, given a lot of the initiatives you discussed, including a higher direct mix shift, essentially a similar commission dollar, but a lower percent commission as you shift people into subscriptions, shouldn't that mean that while it's a 20% uplift in value from a consumer, that drop-down to Autodesk is going to be higher than 20%?
It definitely could play out that way. We're not prepared to forecast what that would be, certainly the bias on at least those two facts and several others are that the bottom-line effect is larger than the top-line effect.
Exactly. Given your relatively low margin profile today, that should be very powerful on the EBIT margin line, I would think.
Yeah, it compounds nicely as the billings go up. You got the bias right on that.
Hi, thank you. Gregg Moskowitz from Cowen. Just a couple of questions. On the margin dollar profile changes made to LT effective Q4, I just wanted to confirm, is that the only product for which you are changing the margin dollar profile that quarter?
Yes. Right now, again, we want to try piloting things and see how it works. Ironically, though, we're piloting on our volume product. It's a big change, even though it's just one product.
All right, great. Secondly, Steve, on getting back to consumption, you talked about a goal to effectively instill this within all of your named accounts. I think earlier, it might've been Andrew who was talking about in fiscal 2016 and 2017, moving consumption down market. I would think as you start to get to smaller target customers, as well as ones that frankly aren't using or utilizing as many apps within Autodesk product portfolio, that that might be a little bit of a tougher opportunity. It'll be helpful to get a sense of really how broad you think consumption could be across your customer base. Thanks.
Sure. I'm excited about what we can do in the SMB space in the area of consumption, but we got to figure things out. We definitely see an opportunity to drive deeper account penetration and also get a stronger competitive position in that market space. It's a very large space. It may not be the exact same offering we have for the named accounts. The requirements for our largest accounts are specific to how they're running their enterprises globally. They have lots of different needs that may vary or differ from the SMB space.
Yeah.
This is the reason why this is a beyond FY 2016 thing. You should expect that we're going to pilot and test a lot of different configurations and learn from specific customer feedback what are the right options or offerings that that base of customers are looking for. That's why we're talking about it now. We're telling you we're working on it, but it's out there a little bit.
Yeah. Go ahead, Amar.
I was just going to say, One of the tests that we've been running is some of our cloud consumption services like simulation, rendering. That's a great place where that middle tier has been
Using services on a consumption basis, we've seen a lot of success with that. In fact, with visualization or rendering, we just crossed 10 million renders. Those are kinds of offerings where our customers, in addition to what they're already using, are flexing into the consumption model. That's the kind of test I think we'll continue to extend over time.
All I was going to say is just generally speaking, hopefully one of the things you take away from today is that we're very willing to spend time experimenting, design pilots, as scientific experiments, so we can get good data back from it, then we change our behavior based on that. We've been doing that with almost all the success you've seen today. There was an early pilot, two or three, in order to try to get it right. One of the things, once we get the whole machine working, we know what it can accomplish. It's critical to make sure that we're pointing it in the right direction and doing these small-scale experiments have been very effective at doing that. They're getting hungry.
Hi, can I ask you a nuanced question about the $310 for AutoCAD LT versus $240 number that you put out there? Does the $310 for the subscription include, I guess, a assumed renewal rate of 90%?
Yes.
It also includes the channel margin, I guess, discount that you-
It's the value back to Autodesk.
Okay. That includes all the initiatives.
Those are apples-to-apples compare as value back.
Based on historical.
Okay. Based off historical. Okay.
Yeah. It's not like projecting forward new ones. This is just taking the historicals and then basing it on that. You could add to that if you were to change some of those assumptions that we would do better in the future-
Yep
on some of those things. That 300 number could go higher if we do better, for example, on renewal rates. What we used to make an apples-to-apples is we took the historical metrics on the one side, and then we applied it to the other side.
Okay, thanks.
No, they're not. The data on the left is the actual, this is what-
Yeah
the value is today.
Yeah.
All right? The data on the right has an assumed renewal rate.
Yeah.
It's on par with what we're seeing. Those things are all steady state.
Right.
Remember, I'm saying it's value back to Autodesk.
Yeah, I'm sorry, it's Carl. Just to connect from Albert. Per Carl's point, though, that is a layer of conservatism that's probably built in because you have a little bit of a channel shift, which should be higher margin, as well as potentially higher renewal rates. Per what Carl just said.
Yeah
Those are the mechanics as to why the 310 could be higher.
Yeah. Got it.
If those were higher, the number on the.
If those were higher, the value shift is different.
Yes.
Yeah.
Sterling Auty with JPMorgan. Last year when you introduced the rental program, you talked about how it would probably expand the user base. I'm curious, I didn't really see a lot of metrics around the rental piece, but specifically.
I gave you a 35% number.
Well-
Desktop subscription is rental.
Yeah, desktop subscription is rental.
Specifically, when I'm thinking about the monthly versus the quarterly, what timeframe is more popular?
It followed very similar dynamics to what Adobe saw. Initially, a lot of monthly purchases, and as time's gone, it's shifted to annual. We're seeing that the bias is now much more towards annual than monthly.
Why don't you just repeat the 35%?
Yeah.
Just so we're clear.
Right out of the gate, 35% of the desktop subscriptions coming into the company right now are net new buyers. Buyers we didn't have before.
Specifically, when you look at the customer base, is there any industries or any type of user profile that are attracted to that?
Well, okay, they tend to be smaller companies. Out of the gate, they tend to be smaller companies buyers, which is great because that's where we get a lot of our new customer acquisition as it is. They do tend to be smaller companies. The rest of the demographic shifts across industry is actually pretty evenly distributed. Initially, when we rolled out the desktop subscription model or rentals, there was a huge influx of buyers from the media and entertainment business. We've seen that level out. We see proportional representation from all the industries now.
Yeah, remember, there are a number of reasons contributing. It's companies who are managing their cash flow closely. That's one group of people. Another are the ones who are project-based. Media and entertainment and AEC tend to be more that way than manufacturing.
Yeah.
There's people who deal with peak demand. That would be another reason to do it. I think for the last group, it's just a preference of how they want to buy.
Yeah.
It's a combination of all those reasons.
Right.
Just a quick clarification. You took the TAM up by several billion. Where was the incremental jump in the TAM from last year when you gave us that number?
That's a good question.
Well-
For example, the construction industry is growing.
Yeah.
The overall economic output is growing. They're spending more on software. That's one of the reasons, at least in AEC, that TAM is growing.
Yeah. No, the reason why I was a little bit avoiding it is, most of our TAM numbers are coming from third-party sources.
Yeah.
We're sourcing that data from external ones to validate it. Both, certainly the overall size of the manufacturing construction industries are growing, but also the software component. As Amar talked about, software spend within the construction industry is going up at a much faster rate than the overall growth of the industry.
For Buzz and Amar, it's hard to get market share figures where you're at in terms of share of your respective end markets, but is there a sense of where you're at today and Maybe you could just comment a little bit. Buzz, we've talked about this in the past about how you've typically lagged some of the customers in terms of some of the requirements. Now it feels like some of the products you're bringing to market are now well ahead of them even understanding what they are. You're helping.
Sure
drag them into some of these new offerings, which are.
Right.
Can you just walk through that?
Yeah. Amar, if you could first. I think we've led the market in terms of unit volume and number of seats out there that are being used by our software, particularly if you consider the whole portfolio of the many things that people do. I think we've shifted from being the fourth player in the market to the third player. I think we're outgrowing our competitors in that sense. You're right on the last point. I think we have shifted from being the volume player, the 80/20 player, to be the innovator. I think we're coming out with more news than our competitors are and doing things that are attracting really new types of customers. Yeah, I think there's been a big shift in the last three years.
Yeah, the other thing is, certainly in Buzz's arena, when we critically hold ourselves to number three, remember that a large proportion of the revenue that our competitors have is in a part of the business, the PLM part of the business, that historically we didn't participate in.
Right.
The numbers change dramatically when you look at the CAD portion as opposed to the PLM portion. Now that we have our cloud-based PLM product, it's fair to consider it. We have a couple-year-old PLM product compared to a, in some cases, 30-year-old PLM product.
Yeah. It's also a new generation product that just charges a different price point per user.
On the AEC side, we've been the leading vendor, especially in the buildings part of the AEC industry. In the engineering, which is a little more competitive with other folks, we've been gaining share and doing better with the examples that I told you about. In construction, we're certainly out in front with the software component of the construction industry. I would say the process and plant industry is one, but kind of like Buzz's world in the dynamics, we are the challenger, we are the disruptor. We are, of course, up against a couple of giant companies there. I think across the board, we've really established a leadership position in AEC. Nothing new. They're a little bit like the PLM example that Buzz said in manufacturing.
They've been relying more and more on that ProjectWise business of theirs, and they feel more similar to a Dassault or Siemens-type company now in AEC. Nothing new, really dramatic.
Yeah, thanks. Follow-up for Buzz and for Amar. Buzz, in simulation, you would appear to have less than 5% share of the simulation revenues. If we're right that it's about a $2.5 billion, maybe $3 billion market, what do you think is a realistic share objective for you in simulation? Then for Amar, when you last reported infrastructure as a division, and perhaps someday you'll do it again, it was just over $200 million a year.
How would you describe it today or where, again, for that business, you think it could grow?
Right. Why don't you, Buzz?
Well, I think simulation market, I think as it shifts more toward the designer, we will do relatively better and away from the heavy-duty analyst. Okay? I think we do a lot to focus our business there. That's a long-term shift, as you know. That's taken generations, I think, to start to change that. I think there's movement in the market. I think if when you start to see some of the way Fusion, for example, is doing simulation, it just has a very different perspective. We don't even call it simulation. You notice in my slides, I refer to it as task.
Which I think is really what customers want to do. I think that can change our shift. We are, at our center, a design and product company. Okay? I think there'll always be specialists. By the way, partners. Ansys is a partner of ours, as well as someone that we compete with sometimes, that I think customers will continue to use those products. Will use them in conjunction with our products. I do think we can see lots of expansion. Okay? I certainly see we could get to well over 10% of the market, maybe more towards 20% of the market. Yeah.
On the infrastructure side, it's substantially larger than the number that we had previously reported. Certainly Bentley was a company that we mentioned that we're taking share from. Some of the wins that I shared with you, we used to have a lot of story in just road, but now with rail and aviation and water, there's a lot of spend happening in the world of infrastructure, and there's more software spend. That business is set to grow nicely. With something like InfraWorks, we're even capturing what people would have spent on GIS systems with Esri and those kinds of companies, or urban planning, all that upfront, there's a potential to shift some of that. I think the opportunity is very large for us with infrastructure.
Thanks. Dave is allowing me one more follow-up for Steve. As you pull the perpetuals out by product and by region, would you, in conjunction with those changes, implement an agency model to supersede the VAR model, again, region by region, so that eventually, you will have a mostly, if not entirely, agency model?
By the way, Dave, the rule is if Dave wants to ask another question, next year he has to commit to bringing bagels for everybody. We're going to fill his luggage with bagels on the trip from New York so we don't have to eat that stuff.
Better quality bagels.
White fish.
We're going to take a look at it product by product and region by region. We are seeing places where it makes more sense to use the agency model. Again, we call it service fees. The agency is the same thing. We're already using that with our cloud-based offering. We have ideas about other places we'll use it as well. We really want to evaluate it. There'll be a lot of piloting going on, and we'll figure out what makes the most sense based upon how we keep our partners engaged, but also really how we serve the needs of our customers.
Hi. Steve Ashley, Robert W. Baird. I have a quick follow-up on the slide you put up on the name-account penetration with the consumption pricing. You have 10, just 10 accounts, very small pool, but that the average number of subscriptions went up 2.4 times. Does that suggest the amount of revenue under a new licensing model would be maybe twice what it was under the old model, if that were to hold that kind of usage level?
First, I want to make sure I'm clear. Those 10 were the ones that moved from our older model to the newer model. We actually have customers that have come in that didn't have that older model that are now in the consumption model. I wouldn't want anyone leaving thinking we just have 10 customers in the consumption-based model. It's hard to specifically draw a line in saying by having more users, you get a linear increase in the billings and revenue overall. It's going to vary account by account, situation by situation. It's a safe bet to assume we are getting more because we're delivering a lot more to our customers. We're giving them a lot more flexibility. They have more access to the software. As a result, they're applying it on more of their projects.
Again, a reminder, we also in our current, the new version of EBA, we have other items that are also bringing in billings and revenue, enterprise priority support, dedicated customer success managers, a lot more work with our consulting group overall. Very safe to assume they're driving deeper penetration and extracting more value out of that as well. I just wouldn't necessarily draw that straight line between numbers.
maybe I would get a little mathematically, I might draw the straight line, the slope might not be one.
Right.
There is a linear relationship. I mean, the relationship is linear. It's not exponential, and it's not logarithmic. There is a linear relationship. The slope of the line may not be exactly one.
A follow-up question. Your 2.9 million active users seems to be a great target for your subscription offering, lower hanging fruit. I'm wondering, is there a way to quantify the inactive users on your platform? Maybe as a follow-up, is there a way to think about the desktop subscription product? Will it have the ability to phone home to potentially eliminate the piracy issues, and maybe talk about maybe how that could expand the TAM?
The 2.9 number is the number of users that are on 1 to 5 releases back. There's other users that are 6 or more releases back. Generally, that pool isn't as stable as the 1 to 5 release back pool. There's lots of people that drop off. They become inactive. The companies go out of business. Is there a long tail of more inactive? Absolutely. There's even a bigger tail of pirate users out there that are using the software and had never engaged in a purchase relationship with us at all. Yeah, there's a larger number out there. It's just that when you talk about customers that are actually already in a relationship with us, it's just kind of spread out over a few years. That 2.9 million really creates a reliable pool to go after.
That's why it's such an interesting number.
Yeah, one of our worst cases, I went in, I saw this customer recently, it's a construction company in Germany. They run their whole automated factory using our software. It was running on AutoCAD 2002 or AutoCAD R14, which came out in the last century. They were trying to tell me what great customers they were and how happy they were, and I was just trying to tell them what lousy customers they were, and the conversation went downhill from there. Seriously, there are people, and there is beyond the 2.9 million people. It just becomes a lot less approachable for us. It's not something we can bank on.
Exactly.
Hopefully, we convert some of those people over time, but that doesn't really figure into it.
Exactly.
The second point you made is most of the things that we're doing around digital delivery of desktop software or cloud-based software will involve a different mechanism for licensing that helps eliminate some of the piracy. Just generally people, as they move towards using online services or online delivery, those become much more difficult to pirate. To be totally, none of those are in the 2.9 million.
Yep.
That's a different number that we can't count easily.
There's been a view that the transition could actually be negative from an NPV per sub basis. It was good to see the 30% metric you guys talked about for LT. Could you just talk about a comparable metric for some of the other products like Revit or anything else like that, just so we can get a better understanding of the change in per seat, per sub NPV economics as we go through the transition.
I think at this point, we're not prepared to do it around other products. I think you could expect things in the same ballpark. We said, a lot of what we do is this experimentation. We'll see how this goes. We'll update you on how we do. Our best thinking at this point is that the things we're doing around LT are not specific to LT. As we learn from that, we may adjust.
If we should think about it in the same ballpark, if only the move to subscription results in a 30% increase in per seat NPV value, yet you guys are adding so much more value to the customer with cloud, et cetera. Maybe this goes back to Albert's initial question. I'm struggling to map 30% plus to the 20% metric that you guys have been guiding to.
You started at negative, so we've now gotten you from negative to greater than 30%.
I didn't say I was negative. There's been a view out there that it's actually negative.
Exactly. We're somewhere in the middle of that too, and that's where we got to 20. We don't believe it could be negative either.
Okay.
Dave?
Sorry, if you have time for one quick one. On that topic of the 2.9 million customers on the last one to five releases, roughly 50% of it is LT. What's the other 50%? How do you incentivize an existing LT customer to pay that extra 30% rather than going to a competitor?
Let me talk to one, and Andrew can talk to two or more. The first thing is, one of the interesting dynamics about the 2.9 million customers is that the distribution of products within that 2.9 million is very similar to our distribution amongst subscribers or amongst total products sold. There's not a dramatic difference. What I was telling people is if you were going to look for first-order effects, I'd look at places like geography rather than product, if you wanted to start teasing it apart. There are differential rates of attach and renew in different parts of the world that are more dramatic differentiator than the product is for discerning between people who subscribe and people who don't. Is that a fair way?
Yeah. Fair way. Yep. When you get to the 30% value, you got to remember, the initial purchase they're putting down for this LT is dramatically lower than what they would have ever paid for the product up front. If you're one of these occasional LT buyers and you were looking at your five-year or three-year buy a new seat cycle, you were going to pay a lot more than that initial fee we deliver. If you feel the product's valuable, you keep paying that fee year over year, and the lifetime value starts to come out in three, four years out. You're already putting the customer in a situation where they're getting access to a professional-grade, well-respected product at a significantly lower price point than they've ever seen before. Are competitors going to try to come out and attack that price? Guess what?
They already are. Okay? ZWCAD recently came out with a campaign to try to match our upfront price. They're going to see as little traction with that as they've seen with some of the other initiatives, because people want the real deal. They want the real DWG-based application. They want to be part of the ecosystem that's on that. They're not going to see much traction. I'd love to see them matching that price point because it's going to undermine their little business even more. Remember, the customer's upfront experience with this price point is just saying, "Wow, really?" You should see what our partners are doing in some places. They say, "Get AutoCAD LT for as little as €360 a year." It's a big deal to be able to say that. Just two things to add to what Andrew said.
Our customers want industrial-grade software, right? They are in the business of making important things. They don't want to take a chance with some rinky-dink clone thing that falls apart. They value industrial-grade software, and I think that's the first thing. The second thing is our customers actually want to be on the latest technology, and we need to make it, and with this digital delivery, we're going to make it a heck of a lot easier to stay current and to be able to use everything, whether it's cloud-based tools or new capabilities in the core software. That has not been easy to do in the perpetual experience. That's going to become a lot easier with the digital delivery and the continuous update-type capabilities we're putting in.
I think a combination of attractive price offer, being able to stay up to date at all times, being able to collaborate with people, and have a reliable piece of technology, that's going to keep this move very compelling for our customers.
Yep.
All right. Well, I think that concludes our formal presentation and Q&A. If you have time, please stick around. We're going to have a buffet lunch right out in the gallery. Thank you for attending