Good day, ladies and gentlemen, and welcome to the fiscal Q2 2019 Autodesk, Inc. earnings conference call. At this time, all participants are in a listen-only mode. Later, we will conduct a question and answer session, and instructions will follow at that time. If anyone should require assistance during the conference, you may press the star then zero on your touch-tone telephone to speak with an operator. As a reminder, this conference call is being recorded. I would now like to introduce your host for today's call, Mr. Abhey Lamba. Sir, you may begin.
Thanks, operator, and good afternoon. Thank you for joining our call to discuss the results of our second quarter of fiscal 2019. On the line is Andrew Anagnost, our CEO, and Scott Herren, our CFO. Today's conference call is being broadcast live via webcast. In addition, a replay of the call will be available at autodesk.com/investor. As noted in our press release, we have published our prepared remarks on our website in advance of this call. Those remarks are intended to serve in place of extended formal comments, and we will not repeat them on this call.
During the course of this conference call, we will make forward-looking statements regarding future events and the anticipated future performance of the company, such as our guidance for the third quarter and full year fiscal 2019, our long-term financial model guidance, our cash flow expectations, the factors we use to estimate our guidance, including assumptions regarding ASC 606, our maintenance to subscription transition, ARPS, customer value, cost structure, our market opportunities and strategies, and trends for various products, geographies, and industries. We caution you that such statements reflect our best judgment based on factors currently known to us and that actual events or results could differ materially.
Please refer to the documents we file from time to time with the SEC, specifically our Form 10-K for fiscal year 2018, our Form 10-Q for the period ending April 30th, 2018, and our current reports on Form 8-K, including the Form 8-K filed with today's press release and prepared remarks. Those documents contain and identify important risks and other factors that may cause our actual results to differ from those contained in our forward-looking statements. Forward-looking statements made during the call are being made as of today. If this call is replayed or reviewed after today, the information presented during the call may not contain current or accurate information. Autodesk disclaims any obligation to update or revise any forward-looking statements. We'll provide guidance on today's call, but will not provide any further guidance or updates on our performance during the quarter unless we do so in a public forum.
During the call, we will also discuss non-GAAP financial measures. These non-GAAP measures are not prepared in accordance with generally accepted accounting principles. A reconciliation of our GAAP and non-GAAP results is provided in today's press release, prepared remarks, and on the investor relations section of our website. We report a number of numeric or growth changes as we discuss in our financial performance, unless otherwise noted, each such difference represents a year-on-year comparison under ASC 606. Now, I would like to turn the call over to Andrew.
Thanks, Abhey. Our strong Q2 results, led by 28% growth in total annualized recurring revenue, not only reflect a healthy demand environment, but also continued execution as we move further through our transition to subscription. We also achieved strong growth in ARPS, billings, revenue, and earnings. There are several key areas that I want to highlight. We had record growth in total ARR and total ARPS on both a year-over-year and sequential basis. The same goes for core ARR, which grew 29%, and core ARPS. Recurring revenue advanced to 96%. We continue to see steady migration of maintenance customers to subscription with the Maintenance to Subscription program, or M2S. We bolstered our presence in the construction market with the acquisition of Assemble Systems. First, let's dive into ARR. I'll repeat that we view ARR as the best proxy for measuring our progress and the overall health of our business.
I noted last quarter that I expected ARR growth to build as we move through the year, and we're absolutely seeing that in the second quarter. The strength in total ARR was once again broad-based with all three major geographies showing growth, with APAC showing the strongest growth. Subscription plan ARR more than doubled for the sixth time in the past seven quarters, driven by growth in all subscription plan types, led by product subscription. We continue to drive impressive growth in product subscription ARR on both a year-over-year and sequential basis. Looking at our core business, which represents the combination of maintenance, product subscription, and EBA subscriptions, it's not surprising that core ARR grew in line with total ARR, as our core business drives the overwhelming majority of our revenue, ARR, and billings growth. Simultaneously, customers continue to engage with our solutions for reimagining construction and manufacturing.
That progress is reflected in cloud ARR, which grew more than 20% year-over-year and 10% sequentially. You have heard me say many times that construction is our key initial focus for becoming a design and make company. The market is ready, and we have compelling technology and products. The biggest contributor to our cloud ARR is our BIM 360 family of products, our project delivery and construction management software that connects design and construction. A great example of how BIM 360 is being utilized in the field comes from Webcor, a large U.S.-based general contractor. They are utilizing a range of advanced technology to support an innovative and efficient approach to construction. The company is using Autodesk BIM 360 on the UC Merced 2020 project, which will expand the University of California's Merced campus by over 1.2 million square feet across 14 structures.
BIM 360 is enabling Webcor to gain greater efficiency in delivering documentation and coordinated models to the field. They are also using the Forge platform to integrate and connect project data with BIM data, resulting in a single source of information and enhanced project delivery and performance. We continue to aggressively pursue the $10 billion construction opportunity, and last month, we expanded our product offerings with the acquisition of Assemble Systems, a cloud-based solution offering combined 2D and 3D quantity takeoff for the construction market. Assemble is doing something very unique and was one of the first companies to recognize the power of BIM for pre-construction planning, and developed ways to efficiently get information out of the model and into the planning process. They extract the information and make it useful and accessible for contractors to help them with bid management, estimating, project management, scheduling, site management, and finance.
They were also an early adopter of our Forge platform. Assemble has a team of about 50 people, fairly evenly split between sales and engineering. Their products are used by over 200 companies, including a quarter of the ENR 400, and have been used on more than 7,000 construction projects. Assemble will enhance our market positioning and contribute to making BIM 360 a more comprehensive solution spanning across the design and construction phase. We believe this will be a great fit, both from a product and people perspective. We know them all well, as Autodesk, through our Forge Fund, was a lead investor in their Series A round last year. We're also seeing nice progress with the adoption of Fusion in the manufacturing market. An exciting Fusion customer is Fabric, makers of award-winning cycling products. Fabric's mission is to bring innovation to its industry by leveraging unique manufacturing methods.
The Fabric team was initially introduced to Fusion by another customer and started using it for the complex industrial design requirements of bicycle saddles, and for its ability to link the designs directly to manufacturing and 3D printing. Then they expanded their uses of Fusion into design, simulation, and validation. Recently, they have begun using generative design in Fusion to explore hundreds of practical design alternatives that deliver superior performance and can be manufactured efficiently. They're a great company, and they illustrate how manufacturers are innovating with the use of advanced tools like Fusion. Now I'll turn it over to Scott for more details on subscriptions, ARPS, and other financial metrics.
Thanks, Andrew. Before getting into the Q2 numbers, I want to comment on the terrific progress we've made since we started down this path to transform our model to subscription and cloud back in calendar 2014. We've added over 1.7 million net subscriptions over that time period, and it's now been two years since we sold our last perpetual license. The mix of our business has also dramatically shifted, as maintenance now makes up less than 30% of the subs and ARR basis, and the new model has allowed us to add a significant number of new customers to Autodesk. We've also migrated over 600,000 maintenance customers to subscription. From a financial standpoint, we're now firmly back in the growth stage.
We're back to non-GAAP profitability, positive cash flow, and we've transformed our business from less than 40% recurring revenue prior to the start of the transition to a highly predictable 96% recurring revenue in Q2. We still have some ways to go to achieve the fiscal 20 targets we've laid out and believe the best is yet to come, but it's worth noting the meaningful progress we've made to date. Looking at Q2 now, I'll start with a closer look at subscriptions. Subscription plan subs grew by 290,000 in Q2, with growth coming in all three categories: cloud, enterprise, and product subscriptions. Core subscription additions were 88,000 and increased 6% sequentially. We added 31,000 net cloud subscriptions, which is a nice step up from the 18,000 we added in Q1.
It's important to note that net subscription additions continue to be impacted by product consolidation from the adoption of Industry Collections. The good news is that many of these customers are increasing their total spend with Autodesk, contributing to solid increases in ARR and ARPS. The adoption of Industry Collections is happening through the regular run rate of new business, the renewal process, the legacy promo, and the M2S program. New subscriptions for Industry Collections increased 60% year-over-year and represented 40% of the net product subscription additions in Q2. Industry Collections now make up 27% of the total base of product subscriptions, up from 14% in Q2 last year. That's great progress.
This is important because Industry Collections generate higher ARPS and have a much higher renewal rate compared to standalone products, and it also signals a deeper relationship with the customer as they are able to utilize more of our solutions. Speaking of the M2S program, we continue to make solid progress in migrating our maintenance customers to subscription. In Q2, customers migrated 117,000 maintenance subs to product subs. While that's less than what we've converted over the past two quarters, remember that the pool of maintenance subscription gets smaller each passing quarter. The conversion rate remains strong, with over one-third of all maintenance renewal opportunities during Q2 migrating to product subscription. We saw that migrate once again, about 30% of eligible subscriptions upgraded from an individual product to an Industry Collection. The renewal rate for maintenance declined slightly sequentially as expected.
This is consistent with our long-term model, where we projected a decrease in maintenance renewal rates as we progress further into the M2S program. The renewal rate for product subscription experienced another sequential increase, and we expect it to continue to rise as the product mix shifts toward higher value products. Helping bolster that renewal rate are the M2S-related subscriptions, which have a very high renewal rate, as expected, because the program was designed to be very sticky. To go a little deeper on this topic, when we first announced the M2S offer, we provided a three-year price outlook to our maintenance customers who converted. As time went on, we heard customer feedback on extending this pricing outlook so they can plan for their long-term business needs and investment in Autodesk solutions.
In June, we announced that we're extending our price commitment to 2028 for customers who continue to renew after they switch to subscription. For these converted customers, the renewal SRP will increase by no more than 5% every other year, starting in calendar 2021, subject to currency movements, of course. This has nothing to do with the adoption of M2S, but rather putting customers at ease regarding their number one concern, that Autodesk will significantly increase pricing once they move to subscription. This action simply puts into writing what we've been verbalizing since we launched the program. The cost-of-living-type adjustments would be implemented after the initial three-year price freeze and is consistent with our long-term financial goals. Each quarter, the vast majority of the new subscription plan subs are added through traditional means. However, we continue to make progress in converting legacy users into subscribers.
In Q2, the legacy promo added 17,000 product subscriptions, and 35% of those were Industry Collections, which is the highest percentage we've ever achieved for the legacy promotion. Once again, the average age of the licenses that are turned in with the promo is about seven years behind the current release, indicating there's still a very long tail of legacy customers to convert. There continues to be about 2 million of these legacy users that are actively using an old perpetual license without a maintenance plan. We believe that over time, we'll convert a large number of them, utilizing our inside sales team to concentrate on converting this important cohort. A consistent attribute of the transition is that new customers continue to make up a meaningful proportion of the product subscription additions and represented over 25% of the mix for the quarter.
These new customers come from a mix of market expansion, growth in emerging markets, converting unlicensed users, and people who have been using an alternate design tool. Let's talk a little bit more about annualized revenue per subscription or ARPS. ARPS continued to inflect up in Q2. Various pricing changes we've made in the past two quarters had the greatest influence on Q2 ARPS. These have been relatively small adjustments, such as the price increase associated with the M2S program, lower channel discount on AutoCAD LT, and an increase for multi-user subscriptions.
Long-term ARPS drivers will continue to be the growing renewal base, which comes at a higher net price to Autodesk, the increase in digital sales also at a higher net price to Autodesk, the product mix shift to Industry Collections, the maintenance price increase for those customers who don't take advantage of the M2S program, and less discounting and promotional activity. We expect ARPS to continue to inflect up for all the reasons I've just discussed as we progress through the transition. Our eStore, which is playing a bigger part of our digital sales, grew over 75% in Q2. For the past four quarters, our eStore has generated over 20% of the product subscriptions, and Q2 also marked the seventh consecutive quarter of greater than 30% growth in our direct-to-enterprise business.
What's interesting is that while the growth of our total direct business accelerated to its highest rate in over two years, growth in the indirect business grew even faster, leading the mix of indirect business to tick up a point to 72% of total revenue. We continue to believe that over time, the mix of direct business will outpace the growth of indirect, leading to a more even split between direct and indirect revenue. Moving to spend management, our total non-GAAP spend came in at $556 million for the quarter, which is slightly higher than expected. However, if we normalize for ASC 340 and foreign exchange rates, the year-over-year growth in total spend would have been less than 2%.
Sequential increase in spend was related to the continued hiring ramp that we've been calling out for the past few quarters as we near the completion of the resource rebalancing we announced in Q4 of last year. Our intent for fiscal 2019 remains to keep non-GAAP spend roughly flat at constant currency relative to our fiscal 2018 budget at about $2.2 billion. Looking at the balance sheet, total deferred revenue grew 20% as reported and 24% under ASC 605. Unbilled deferred revenue decreased by $6 million sequentially to $406 million. It's important to note the impact of ASC 606 here because 606 requires early renewals to be captured in unbilled deferred revenue. Early renewals in Q2 were $20 million lower than in Q1 when some maintenance customers renewed early ahead of the M2S price increase.
While traditional unbilled deferred revenue related to moving our large EBA customers to annual billings increased by $14 million sequentially, it was more than offset by fewer early renewals. Looking at cash flow, we returned to positive cash flow as expected in Q2. Operating cash flow was $43 million, which benefited from growth in billings and better-than-expected billings linearity in cash collections. We used $147 million in the quarter to buy back roughly 1.1 million shares at an average price of $131.52. We continue to be committed to managing dilution and reducing shares outstanding over time. I'll turn the discussion to our outlook, and I'll start by saying that our view of the global economic conditions remains mostly unchanged from the last few quarters, but we're monitoring the potential macroeconomic impact from various trade and tariff disputes.
There have been some FX volatility, but our hedging program has succeeded in smoothing out the bigger swings. Overall, we're really proud of the results we achieved in Q2 in the first half of the year. As we look at our outlook for Q3 in the second half, we expect to see continuing sequential increases in most metrics, including ARR, ARPS, billings, revenue, spend, earnings, and subscription additions. We expect our hiring ramp to continue in the second half as we finish the rebalancing of resources to the most strategic projects, and as such, we expect our spend to increase sequentially and will likely be at the high end of our guidance range for the full year. Keep in mind that the adoption of ASC 340 capitalizes commissions, so we won't have as big of a step-up in spend in Q4 compared to historical trends.
I'll also point out that we're modeling cash flow to decrease moderately sequentially in Q3 related to the shift in billings linearity that I mentioned earlier, which resulted in more of our Q2 billings being collected in-quarter. We continue to expect a sizable uptick in cash flow in Q4, and it will be cash flow positive for the year. The acquisition of Assemble will not have a material impact on our overall results this year. We are confident in achieving our fiscal 20 goals. That said, subscription additions for this year likely will be at the low end of our guidance range, primarily related to the success we're having with the adoption of Industry Collections and the consolidation we're experiencing with the M2S program.
One side note with regard to subscriptions, we will continue to report out on subs and ARPS for the remainder of this year, but are not planning on reporting those metrics on a quarterly basis starting in Q1 of fiscal 20. Of course, we'll use events like our annual Investor Day to report out on important metrics that will help you build out your long-term models. Now I'll turn the call back over to Andrew for a quick closing comment.
Thanks, Scott. I wanted to note that I've recently completed my first year as Autodesk CEO. It's been an exciting four quarters, and we've made a tremendous amount of progress. The three primary focus areas that I outlined at this time last year have not changed and all are on track. Allow me to remind you what the organization is focused on for driving success. The first is completing the subscription transition. We remain focused not only on the financial results, but on enhancing the subscriber experience and delivering more value to our customers. Customers are acknowledging that we are absolutely providing greater value with subscription. Second is digitizing the company.
By that, I mean that we are investing in our own digital infrastructure to create opportunities for our customers to transact and engage directly with us, increase their level of self-service for a wide range of customer needs, and increase our ability to instantly and reliably understand how successful our customers are being with our products. We've made progress in this area, and we have expanded the self-service capabilities in Autodesk Account, allowing customers to easily manage users, add seats, align billings, and take advantage of the flexibility inherent in our subscription model. We've also rolled out a broad set of capabilities that help our inside sales and support teams understand the status of the customers they interact with every day. A third is reimagining construction, manufacturing, and production.
You should have no doubt that we are absolutely committed to winning the construction space and winning in the new world of digital manufacturing. We've done well to establish an early leadership position in construction, and we're not going to slow down. Operator, we'd now like to open the call up for questions.
Ladies and gentlemen, if you have a question at this time, please press star and then one on your touchtone telephone. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. To prevent any background noise, we do ask that you please place your line on mute once your question has been stated. Our first question comes from the line of Jay Vleeschhouwer with Griffin Securities. Your line is now open.
Thank you. Good evening. Andrew, let me start with you and ask a question regarding the alignment of your sales and distribution model to your product and market strategy in two respects. With respect to the growing intersection of manufacturing and AEC for your strategy, can you talk about how you're aligning, particularly the indirect channel, to do that? You've typically separated the AEC from the manufacturing specialties in indirect. Would it make sense for you to increasingly combine or converge the channel specialties in that way to be more directly aligned with your product and market strategy, the focus areas you just spoke of. Secondly, with regard to the self-serve model, when would you expect that the e-store could be at least half of the direct business?
Lastly, for Scott, your AEC and manufacturing segment revenues were largely in line with our models, and the upside was, as it turned out, in AutoCAD or the repackaged AutoCAD. Would you expect for the remainder of the year that the relatively better growth or upside might occur in the repackaged AutoCAD as compared with the AEC and manufacturing segments?
Jay, let me start answering your question by just kind of summarizing how we structured the sales force, because I think it actually talks directly to what you're talking about. At the beginning of this year, we realigned the sales force along a series of new models. First, what we did is we, of course, maintained and invested a little bit more in our major account team. As you know, the major account team is basically account-based. Account-based is kind of an industry-type model, but it focuses on what's most important to the account. Which, by the way, is the most valuable way to go to market with anything, because accounts blend many different things. You know some of our biggest accounts are both AEC customers and manufacturing customers because they have buildings and maybe they design products.
The other thing we did is we built out what we call a mid-market team, which is one level below our major account team. This is still a team that fulfills indirectly, but it's structured like a named accounts team in that the reps have a series of accounts assigned to them that they then work with designated channel partners to satisfy. Again, this is the best way to try to scale a business when you see some of our businesses coming together because the accounts are already identified either as manufacturing or construction or architecture accounts or engineering accounts, or as holistic AEC accounts. That model actually allows us to expand the account paradigm deeper into our channel network and actually start to align a channel on an account-based methodology and thinking. The next tier down below from that is what we call territory sales.
The territory sales is essentially split into two pieces. This is kind of the more generic horizontal part of our business, where it's more picked up by ambient demand. There's not specific accounts called out, but we do market along specific segments. What we've done in that area is we've got the bit that's completely driven by channels. The channels are in the lead there. Then we've got the digital piece, the piece that encompasses our inside sales teams and our eStore, which can be highly somewhat verticalized, just like what we've done with named account sales, where a particular inside sales rep might call on just M&E accounts or just manufacturing counts, just AEC accounts.
We believe that structure, combined with some intermediate steps around accelerating opportunities, say in construction and manufacturing, is the right structure going forward and aligns with some of the concepts you're saying where a lot of things that our customers do are getting closer and closer together. That answers your first question. The second question I'll answer this way. Our long-term target for our business is half direct, half indirect. We anticipate that that direct piece is going to be half eStore, half major account direct. Our eStore grew over 70% this quarter, year-over-year. We're posting impressive growth numbers down there. More and more of the LT business is being captured in that channel. I can't give you a specific number on when that mix becomes 50/50. You can see what the investment is doing.
It's driving a lot more growth in that channel, we expect that growth to continue to be robust, especially as it starts to get more and more robust in, say, Europe, then APAC starts to pick up. There's still lots of room to grow down in that channel. Scott, I'll turn the last question over to you.
Sure. Your question on AutoCAD upside. You know, Jay, we repackaged all the AutoCAD verticals and AutoCAD together into One AutoCAD and launched that earlier this year. The whole goal of that was to make it easier to sell, easier to buy, easier to consume AutoCAD. Some of the benefit we're seeing is from that, for sure. I think we'll see some of that benefit continue through the year. I think the second is, as I think most of you know, we made changes in our channel margin structure for AutoCAD LT, which also rolls into the AutoCAD family, where we reduced the channel margins on AutoCAD LT with the goal of having the channels put more of their focus on selling collections, which we see is actually having quite a nice outcome.
When you layer in both the acceleration from One AutoCAD and the change in the channel margin on LT, those are the two things that are fueling the growth in the AutoCAD family.
Okay. Thanks very much.
Thanks, Jay.
Thank you.
Thank you. Our next question comes from the line of Philip Winslow with Wells Fargo. Your line is now open.
Hey, thanks, guys. Thanks for taking my question, and congrats on a great quarter. Just have one question for Scott and then a follow-up for Andrew. Scott, obviously another strong quarter in ARPS here, and you laid out a lot of the drivers that we should see over the next multiple years that continue to move that higher. I was hoping you'd give some color as you think in sort of the next few quarters here, how we think about sort of the incremental drivers in ARPS versus what we've seen in the first half of this year. To your comment about sort of collections being strong, obviously a positive to ARPS. Could you also maybe provide some color on just the impact to the subscription unit number? Like I said, just one quick follow-up for Andrew.
Sure, Phil. On the drivers of ARPS, I mean, long-term drivers of ARPS are going to be the same ones that we've talked about at Investor Day and that I highlighted in the opening commentary. The growing renewal base, which is higher net to us. The digital sales continues to grow. That's a higher net to us. M2S, and M2S both as a long-term driver of growth and also had a significant impact on Q2. As you know, the beginning of Q2, we got into year two of maintenance to subscription, which meant the maintenance price indexed up another 10%, and the conversion price indexed up 5%. That's a long-term driver. That's also a specific Q2 driver for us on ARPS. Less discounting and promotional activities, so greater net yield to us across the board.
The LT channel discount change that we made earlier in the year, we're seeing that now flow through. Obviously, it goes first into deferred revenue, from deferred revenue into the P&L and affects ARPS. Was a benefit to ARPS in Q2. Industry Collections continues to have great success. 40% of the net product subscriptions added in the quarter were Industry Collections, which is a high water line for us. That's driving ARPS. It is having an offset. To the second part of your question, Phil, it does have an offset to subs, obviously, but I think in most cases, the customers end up spending more money with us when they make that change from standalone products to Industry Collections. We drive more ARR and of course, higher ARPS, and it trade-off faster subscriptions growth. I think that's a good trade-off.
Finally, we made a slight price adjustment on our multi-user products. We also saw the benefit of that for the first time, really flowing through deferred revenue and into reported results during Q2. It was a lot of things that all moved the right direction on ARPS in Q2. To probably get ahead of your follow-up on this, when you look at our guidance for the full year, you can see we do expect to see continued ARPS growth in both Q3 and Q4 throughout the year.
Got it. Thanks, Scott. Then, Andrew, obviously, you highlighted the acquisition of Assemble. If I think about just your construction management portfolio as it broadened, obviously, you started on the design side, but then you had BIM and then Assemble being able to connect that BIM data into planning systems. You've got planning, construction management, TAM, you think about the sort of life cycle of CMS. How do you think about how sort of deep Autodesk wants to go onto the CMS side? Because it seems to me you're inching further along kind of that life cycle, but still being grounded in the design side, kind of like CAD and PLM.
Yeah. Long-term, our goal is to capture every aspect of that cycle. From all the way from design through pre-construction to site execution and to operation. We're very interested in that whole process, and the reason I've said this multiple times is, it's because it's moving in a direction that the manufacturing cycle moved in the past. The model is going to become the currency that moves from one aspect of the design process to the next aspect of the make process, to ultimately what you do to retire or manage the asset and make the next asset. We think the model is going to be a critical piece of this process moving forward, and we intend to manage the flow of that information across the entire process. You're correct in what you said around Assemble Systems.
That was a classic move for us to double down on the area of pre-construction, where we're trying to help people take BIM data and turn it into actionable information in the pre-construction planning cycle. That's exactly what Assemble does, and it helps us basically drive open the opportunity in an area that we feel we're highly competitively advantaged in the space.
Got it. CLM, not just PLM. All right, thanks, guys.
Thanks, Phil.
Thank you. Our next question comes from the line of Saket Kalia with Barclays Capital. Your line is now open.
Hi, guys. Thanks for taking my questions here. First, maybe for you, Scott. Nice job on the billings this quarter. I think clearly the acceleration in ARR is the main driver, you touched on some other items as well. Just wondering if you could talk about some other factors there, like duration and linearity and how you thought about some of the drivers of billing strength in the quarter.
Yeah. Thanks, Saket. The duration really didn't have an effect during the quarter on billings. If you look at our weighted average term length of our deferred revenue, in fact, on the balance sheet, you see our long-term deferred actually declined again sequentially. We're not seeing a change in duration that's driving billings. It does get back to a lot of the things I just talked about in response to Phil's question about the drivers of ARPS. ARPS, the layering on of greater growth of Industry Collections, some of the changes that we made on channel margins, the growth through the channel, frankly. We had a really strong quarter in direct, as you heard in the opening commentary. We actually ticked down one percentage point in the percent of direct because the channel grew really strongly during the quarter.
It was a combination of those things, much more so than any kind of one-time benefit that hit billings during the quarter.
Got it. That's very helpful. Andrew, maybe for my follow-up for you. I know we've talked about renewal rates a little bit, but I'd like to zoom in a little bit on subscription renewal rates in particular now that we've gone through a couple price increases. I guess the question is, how have those trended? I think we've touched on it a little bit, but can you recap for us how those have trended? More importantly, how have you thought about those renewal rates for subscription in the long-term model?
All right, Saket. Let me answer that question from a couple of angles. First, one of the things that we watch very closely here as we watch the transition is the maintenance renewal rates in particular. All right? We know that the prices were going to be going up. We modeled in certain changes in maintenance renewal rates over time. We've got a long-term model on what we expect to happen to maintenance renewal rates as the prices ratchet up on maintenance and as the M2S program progresses. What I can tell you right now is those renewal rates are smack dab in line with what we've been modeling and what our goals were with regards to that. Now, do the modeled numbers create the outcome, or are we just brilliant modelers in predicting customer behavior?
We can debate that, what we're seeing is behavior that's right in line with what we were modeling. The same goes for what we're seeing in the product subscription area, the pure product subscription area. We're seeing Nice increases in renewal rates, what we expected, what we have built into our model. I think the important headline here is we're seeing things that are consistent with our expectations. One of the things that was also very gratifying, and Scott mentioned it earlier in the opening commentary, is the M2S program was designed to be very sticky. It's designed to incent our maintenance customers to come over and feel great coming over. What we're seeing is really high, our highest renewal rates for customers that took on the M2S program.
Which, by the way, is an important fact pattern in terms of taking care of our best customers, but also making sure we retain that base into the future.
Very helpful. Thanks, guys.
Thanks, Saket.
Thank you. Our next question comes from the line of Richard Davis with Canaccord. Your line is now open.
Hey, thanks. BIM's been around for, I've been hearing people talk about it for at least a decade, but it's doing well. The question I have is, so when I talk to some of your customers, some bought it because of the digital transformation. Others were more, I guess, cold-blooded and saw hard dollar ROI. So when I think about how markets evolve, is this a market, and you guys have always much more visibility in this than I will. So is this a market that kind of starts with like a hard dollar ROI sale, and then as you get momentum and people say, "This is really important," you'll start to get these kind of bigger deals because you'll hear this digital transformation discussion? This is exactly what we saw with Salesforce, is what we're seeing with ServiceNow and other industries, inspectors.
I wonder if that's how this market's going to play out. Thanks.
Yeah. Look, here's how these markets generally play out. There's always a first mover, and when you're talking about BIM, I wanted to distinguish whether or not you're talking about BIM as in the whole, the discipline of building information modeling or BIM 360 in terms of the construction solution we have.
Yeah, no, that's true.
Let's just talk specifically about BIM 360, which what you're seeing right now in the space is that there is a pressing problem to digitize the site. The mainstream problem, the problem that just about every construction company sees, is this idea of, well, I need to digitize my site. I need to get model information to the construction site. I need to get up-to-date drawing information to the construction site. There's a lot of flow in that segment of the market right now because the technology's ready, the customer's ready, they want it, and they need it. That's getting that kind of classic adoption curve that we typically see in any kind of technology adoption.
More and more people are buying. Now they're starting to ask the questions about, "Okay, what's the next step?" If I go and I look at some of our more forward-looking customers, especially some of our largest construction and general contractors, they're already looking at, "Okay, I'm able to digitize my site, but I want to digitize my entire construction process, just like those guys do over in Boeing or Airbus, and make the same processes that they have for how I build a building, a road, or a bridge." They're experimenting early with this, pushing the building information model further and further down into the process. What you're going to see is this battle for competitive advantage.
It's now becoming who can be more digital first so that they can increase their fidelity on project bids, cut a few hundred million dollars off their bid for a project because they have much more precise information about what the margin is. It's going to become a digital arms race. That's exactly what starts to happen in classic technology adoption. There's all sorts of words for it, the tornado, and you've heard many different kinds of words. It's been called different things over the ages. We're starting to see the early signs of basically a digitization arms race between various people in the construction space. It's going to play out that way.
Yeah, that's super helpful. Thank you very much.
Okay. Thanks, Richard.
Thank you. Our next question comes from the line of Heather Bellini with Goldman Sachs. Your line is now open.
Great. Thank you. I just wanted to talk a little bit, you guys are obviously well on your way to $6, which was the first target you guys laid out years ago. As people start to shift the focus to $11, and I know you've touched on this a little bit here, but can you share with us how you're thinking about the contribution from cloud subs? How important are they to getting to that number? How do we think about broader adoption in the construction market, given they've typically been really slow to embrace technology? I guess I'm wondering, what do you think the industry needs to see to be more open to leveraging technology? Thank you.
Yeah. Let me approach this from a couple of directions. First off, we're definitely committed to our long-term capital targets that we put out there in various forms, especially at Investor Day. I want to remind you of something we said at Investor Day, because I think it's very important for how we look at the company moving forward. We look less at the ratios that we've been classically talking about and much more at the sum of revenue growth and free cash flow margin, and trying to optimize that number to provide the right long-term outcome for the company and the right kind of value creation.
Just remember, as we look forward out beyond FY 2020, we're really looking at the sum of those two numbers, revenue growth and free cash flow margin growth, as kind of the ways to look at where we're going moving forward. Though we remain completely committed to the free cash flow targets we put out there. Now, when you look at the construction market, so I want to challenge something you said. There already is a large increase in technology spending going on in the construction space, right? It's not just because of all the VC money that's been invested in this space. There's actually hundreds of millions of dollars in revenue being generated right now on the site execution problem.
What's changed fundamentally is, it's getting harder and harder for these construction companies, especially on large projects, to win without having some kind of digitization in their process. They're simply not able to get the margins and hit the scheduling requirements they have without having tighter control over the data flow. They've been historically reluctant because the industry as a whole has been historically sloppy. Because the industry as a whole is sloppy, there was no competitive pressure to actually drive technology adoption. What you're starting to see is types of excellence inside the construction industry, where people are adopting, some of our best customers are becoming very digital, and they're highly visionary in terms of how they see the construction industry evolving. That vision and that move towards being highly digital is changing the way people have to compete in this space.
It's been slow in the past because sloppy has been tolerated, and everybody was sloppy, and they were all equally sloppy. People are getting a lot tighter, and that's what's going to change things, Heather, as we move forward. It's basically back to this fundamental thing about as a segment of the market gets more digital, the rest of the segment is dragged into digital by necessity.
Thank you very much.
You're welcome.
Thanks, Heather.
Our next question comes from the line of Sterling Auty with J.P. Morgan. Your line is now open.
Yeah. Hi, guys. Thanks. A lot of discussion around BIM, not only on this call, but a lot of investor questions leading up to the call. I just want to ask it simply, has the financial results out of your BIM initiatives lived up to your expectations?
Well, Sterling, we just had 21% growth in cloud ARR this quarter, which is mostly made up of BIM 360. We added 31,000 subs in the BIM 360 space. We're absolutely seeing results that we want to see in this space. You also saw us move to make an acquisition around Assemble to try to double down on, frankly, what we believe is a highly differentiated part of the process for us around pre-construction. It's an area where our customers are actually saying, "Well, we're glad you're doing this site digitization stuff, but can you do more on pre-construction for us?" We're actually seeing a lot of the things we expected to see, and we're getting increased demand from some of our best customers around areas they want to see us spend more effort on.
I do feel pretty good about where we're going, and I feel very good about our prospects. One of the things I want to iterate here too, because I've said it many times, is it's our goal to be number one in this construction space. We're not looking for number two. We would never be satisfied with number three. We're going to be looking to be number one, and we're heavily focused on doing everything it takes to become number one in this space. I think you just want to make sure that you hear our commitment to that and frankly, our passion for that.
One follow-up, again, on the space. If you look at the big customers, like the ENR 300, et cetera, how are they adopting? Are they going with a single vendor and trying to build, or are they using a couple of different vendors and a couple of different projects to test the waters and then making their decisions and developing from there?
Not to be facetious, but what you see is the biggest in the ENR 100 and the top of the GC chain, what they do is they have a combination of vendors they bring in, and also they bring custom-developed solutions on top of what they're doing to fill the gaps between what they see that no vendor's providing. The great news about the fact that they're engaged in some of these custom solution executions is that they're becoming big adopters of Forge, because Forge lets them stitch together some of our data into some of their flows in ways that we haven't done the work in some of our off-the-shelf products to do that. Everything you just described is exactly what's happening.
I will tell you that there's more and more pressure on us to consolidate more and more of the process around the building information model, and we're trying to respond to all of those requests. The way you described it, as multiple vendors involved and custom software coming together at the big GCs, that's exactly what happens.
Got it. Thank you.
You're welcome.
Thank you. Our next question comes from the line of Keith Weiss with Morgan Stanley. Your line is now open.
Hi, this is Hamza Fodderwala in for Keith Weiss. It seems like the overall ARR growth was strong, but the cloud ARR did come in a bit below that of the core business. Is that largely just a result of a tougher year on your comparers? Do you see any indication of a re-acceleration in the second half from the revamped BIM 360 product?
Hamza, the cloud business actually accelerated sequentially, right? We talked about the ARR being at 21%, and the number of subs going from 18,000 net adds, right? It grew in aggregate, 18,000 net adds in Q1 to 31,000 net adds in Q2. Cloud business is actually performing quite well. You remember, this is the last quarter where our year-on-year compares in cloud compare back to a quarter where we were using a lot of seeding strategy, pushing out a lot of high volume, low price BIM 360 Team subs out into the marketplace. On a year-on-year basis, maybe is where you're drawing your conclusion. If you look at sequentially how it's moving, it's actually moving quite nicely. Fueled by, as Andrew just said, fueled mostly by the growth of BIM 360.
Got it. On the direct business, I understood the dynamic between the indirect business growing a bit faster, but when could we see a more material pickup there given the sales restructuring that you talked about? Are there any sort of exclusive discounts that you're offering through the e-store to incentivize customers through that channel?
No, we're not, Hamza. I don't think it makes sense to do a lot of that. It's a high-class problem to have when you've got the fastest growth rates in our e-store and in our enterprise sales that we've had in the last two years, yet the indirect channel's growing faster. It's not a case that the direct channel is not growing. It's growing really quickly. The indirect channel is just keeping pace at this point. Longer term, we still believe it settles in closer to a 50/50 blend. Frankly, as long as we continue to grow that direct touch space, if the indirect business keeps pace, that's a good problem to have.
Okay, thank you.
Thank you. Our next question comes from the line of Gal Munda with Baird and Baird Capital Markets. Your line is now open.
Hi, thanks. This is actually Alex on for Gal. I just wanted to dig a little deeper in the collection upselling. I was wondering which collections do you see selling the best, and I was wondering if you could give a quick breakdown on roughly how they're split. Secondly, just on a 2 million user base that are not currently subscribers, what is the aging profile of that pool, and are you thinking about any more promos to get them to move over? How are you trying to attract them to become paying subscribers? Thanks.
Alex, on your first question, we don't actually provide that level of granularity. As you'd expect, AEC, which is the biggest piece of our overall business, you would expect the AEC collections to be the bigger piece of our total collection business as well. We don't provide that level of granularity down to the vintage of the collections. The overall good news is collections are growing rapidly.
Okay.
Your second question on legacy, the 2 million legacy customers, we're making good strides on that, right? The statistic that we gave earlier is 17,000 took advantage of the legacy promotion that we ran during Q2. Andrew also touched on this, though, in his opening commentary. We will continue to go after that space, continue to migrate the legacy customers. We continue to see the average of the scattergram of the licenses they turn in, the legacy customers turn in, to take advantage of the promo. We continue to see the midpoint of that bell curve to be seven years old, seven years behind the current version, which means there continues to be a pretty long tail of legacy customers out there. I think what you'll see us do is, price is only one reason they move.
Andrew, you can jump in a little bit more on this as well. The older their licenses, their perpetual licenses, and remember, we sold the last perpetual licenses for standalone products more than two years ago. The older it gets, the more it ages out, the harder it is for them to work effectively within their ecosystem. Price is only one lever to get those guys. Just as those licenses age, they will come back and need to update the licenses. We'll continue to farm that base, probably more so with an inside sales touch through our direct sales hubs, as opposed to trying to just get there with price discounts.
Perfect. Thank you.
Thank you. Our next question comes from the line of Gregg Moskowitz with Cowen and Company. Your line is now open.
Okay, thank you very much. The core ARR was strong, and it was good to see some improvement in cloud subscriptions as well. Scott, you did mention that you expect net new subs to come in at the lower end of the range for the year. Is that solely because of the headwind on net adds from the M2S consolidation in the back half of the year? If so, when do you expect that dynamic will significantly subside?
It is mostly related to success with collections, is what I'd say. Obviously, when maintenance customers consolidate, that turns into a gain in ARR, as we've talked about, but a decline in the total number of active subs. We'll continue to feel that headwind through the second half of the year. The other things that will fuel, though, the other part of your question, which is what's going to fuel the subs growth in the second half of the year, there's a couple of things inside there. One is we'll continue to see increased productivity from the mid-market channel that Andrew talked about. This is a field sales team that we put out there that's driving direct touch with customers and that tier of customers below named accounts and selling side by side with a partner. We just started that process at the beginning of this fiscal year.
They got their accounts assigned February 1st, first day of our fiscal year. They've been meeting those accounts. They've been driving opportunities. They've been filling the pipeline. That business will uptick through the second half of the year. The productivity of that channel, now the longer they're in position, will drive that. The second is just, historically, new product subs are higher in the second half of the year than they are in the first half of the year, is what will drive that growth in the second half.
Gregg.
Perfect.
With respect to the consolidation on the M2S side, I said last earnings call that we expect that to work its way out of the system through the second half of this year, and the reason for that's very simple. The accounts that are most likely to consolidate are our largest accounts. They're the ones that actually make the move from M2S earlier in the cycle. For instance, in the U.S., the conversion rate for M2S is much higher than the average. I mean, they came out of the gate strong, and the largest accounts are already really moved in terms of M2S. You're starting to see that phenomena work its way through the European market. You're seeing the large accounts start to take the M2S offering and move.
Just like with any pattern of anything we roll out in the company, I don't care what program it is, APAC comes up to speed third in that process. It will start to see, as we move towards the end of the year, its largest accounts start to move. What you'll see is these large accounts that move, that are most likely to consolidate, will start to flush out as we move to the second half of the year. Watch for that to happen. As we move through that, we'll see the headwinds start to slow down.
That's very helpful. Thanks, guys.
Thanks, Gregg.
Thank you. Our next question comes from the line of Rob Oliver with R.W. Baird. Your line is now open.
Just one for Andrew to start. Andrew, I know you won't be held to a number on piracy in terms of conversions there, but you did float last quarter a little bit about some of the initiatives you guys had. I know, Scott, you just touched on them as well with the mid-market channel, but new sales motions, some end-product messaging, things like that. I was just wondering if you could talk about the progress of some of those initiatives.
Yeah. It's good. I'm glad you're allowing me to refresh what I said last quarter, that there will never be this announcement of some massive non-paying user event that fundamentally changes our trajectory. You're right. We did roll out in English-speaking countries a new program that allows us to get more information and more direct engagement with people basically using invalid licenses of the software. That program has been integrated with our license compliance teams and with our inside sales teams. We're actually getting leads through that program and getting into discussions with customers. It's beginning to have an impact on the run rates. It will continue to grow. Like with all these programs, they roll out, we start to get traction with them, we start to understand them more, and they start to add more and more contribution.
I think the great thing about the whole non-paying user paradigm here is that it's going to be the gift that keeps giving over multiple years. These users aren't going away. They continue to use the software. They continue to procure more invalid licenses as time goes on. We're going to be continuing to reach out with them more and more as we move forward. Look for it just to continue to support the ramp-ups in volume that we have built into our models over the next few years.
Great. Thanks, Andrew. Scott, I just wanted to make sure, I think you touched on it pretty clearly, but just on that, the change to the subscription contracts, if you could just repeat that. I think what you said was people get the price freeze for 3 years and then are subject to sort of a cost of living adjustment under that, but just wanted to make sure I had it right. Thank you guys very much.
Yeah, you got it right, Rob. When we first rolled out maintenance subscription, if you remember, we gave 3 years' worth of visibility to what the prices would be for maintenance and what the increases would look like, and then what it would be to convert. For those that converted, we offered a 3-year. They didn't have to pay up front, but they got grandfathered at their conversion price for up to 3 years. What we've added to that now, one of the feedbacks that we got from the channel is, "Okay, I want to know what's going to happen after that. I get that you're going to grandfather me for 3 years, but I'm giving up a perpetual license that I'd spent a lot of money on, and I did exactly what you asked me to do.
I kept it current on maintenance for the entire time that I had it. I'm giving that up, so I want to understand some sense of what pricing's going to be longer term." You've heard us say several times that there would be cost-of-living price adjustments that go into that. What we've done is just formalized that and put out to the channel that there will be an every other year 5% price increase, subject to currency changes, obviously, but 5% price increase, which averages out to a 2.5% annual, about a cost-of-living price increase every year for up to 10 years. That's what we've done on M2S, just to put in writing what we've been saying verbally, and of course, what's been built into our modeling for the entire time.
Operator, we have time for one more. Yes.
Thank you. Our next question comes from the line of Brad Zelnick with Credit Suisse. Your line is now open.
Hi, it's Kevin Ma on for Brad. Thanks for taking the question. I just wanted to ask, unpack the 10-year pricing plan a little bit more. Has that extra assurance helped incremental M2S conversions at all? How does that pricing compare to that of non-promo subscriptions beyond 2021?
Yeah. We didn't provide this clarity because of any issues we're having with M2S and the program. As you can see, we have very, very robust conversion rates. They're actually ahead of our original model. We're feeling good about this. We did this communication more to ensure that our sales resources didn't have to have protracted, long back-and-forth conversations about the implications of moving to M2S. It's an efficiency move for us, and it's a visibility move and trust move with regards to our customer base. It's purely something that we're trying to do to look ahead to the last year and a half of the M2S program, make sure it's highly efficient, and make sure our customers basically have no questions. All right?
Remember, as we looked at the 3-year period, there already was a published price increase true-up at the end of that 3-year period. All we did was give them visibility to how the cost of living will play out in years after that.
Beyond that terminal price.
Beyond that. Now, your question about promos, are you referring to the legacy promos?
Were you asking what is the relative price of this cohort versus the people that come in to a net new sub? If you recall, when we laid that out, it starts at about a 50% discount.
Yeah
For this converted group versus what the price of a new product sub is. With these cost-of-living price increases that probably take place on both lines, they will continue to have an advantage price.
Yes
for some period of time.
Just remember, as time goes on, this cohort, this M2S cohort that moved from maintenance subscription, it starts to become a fairly small percentage of our total subscriber base.
With an extremely high renewal rate.
With an extremely high renewal rate.
Got it. That makes sense. Thank you.
Thanks, Kevin.
Thank you.
Go ahead, operator.
That does conclude today's Q&A session, I'd like to return the call to Mr. Abhey Lamba for any closing remarks.
Thanks everybody for joining us. This concludes our call. Please feel free to call us at 415-547-3502 if you have any questions. Thank you.
Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a great day.