Autodesk's Q2 revenue up 15%, comes out swinging on AEC

Autodesk has lost none of its swagger, yesterday reporting that total revenue was up 15%, with results across the metrics that investors look at ahead of consensus estimates. Even so, the company's guidance for its fiscal third disappointed, leading Autodesk's share price to be down 3% after hours. First, the details, then quotes and comments:

• Total revenue was $913 million, up 15% as reported and up 16% on a constant currency basis (cc)
• Design revenue was $821 million, up 15% (up 16% cc). Autodesk defines this bucket as the maintenance and product subscriptions related to the design products -- so including AutoCAD, AutoCAD LT, Industry Collections, Revit, Inventor, Maya and 3ds Max. For reasons that I don't quite understand, this category also includes the CAM solutions that incorporate both design and make functionality; and all EBAs
• Make revenue was $71 million, up 37% (up 38% cc). Make includes cloud products such as Assemble, BIM 360, BuildingConnected, PlanGrid, Fusion 360, and Shotgun -- in the case of AEC, clearly used to execute ("make") AEC assets. It's more confusing in the case of Fusion 360, which is lumped into this category even though it includes significant design capabilities
• We also got the more traditional breakdown: Revenue from the AEC products was $397 million, up 19%
• Manufacturing product revenue was $186 million, up 6%
• Media & Entertainment revenue was $53 million, up 5%
• Revenue from AutoCAD and AutoCAD LT was $272 million, up 18%
• Finally, in the catchall category Other, revenue was $5 million, down 8%
• Subscription plan revenue was $841 million, up 27% (up 28% cc)
• Maintenance plan revenue was $51 million, down 51% (down 49% cc)
• By geo, revenue from the Americas was $372 million, up 14% (up 14% cc)
• From EMEA, $355 million, up 12% (up 16% cc)
• From APAC, $187 million, up 21% (up 21% cc)

CEO Andrew Anagnost started the call by talking about COVID, and what Autodesk saw as the world slowly reopened during fiscal Q2: "We closely monitored the usage patterns of our products across the globe. In China, Korea, and Japan, we are seeing usage above pre-COVID levels. In some areas of Europe, we continue to see a recovery as well. In the Americas, we experienced a slight uptick in usage for most key products in July. We see a positive correlation between usage trends and new business performance, which gives us confidence that the green shoots we see in usage will translate to improved new business performance in subsequent quarters." CFO Scott Herren added that "business is recovering in the markets that were impacted by the pandemic earlier on. Some of our major markets like the US and UK have stabilized, but are yet to show meaningful improvement ... Second quarter new business activity was more impacted [by COVID-related issues] than Q1, with new business declining in the mid-teens percent. We think the second quarter will be the most impacted by the pandemic."

Autodesk said it continues to see success in bringing non-compliant (aka pirated) and legacy (ie lapsed/very old version) users into the fold. The company says it igned 3 license compliance deals worth over $1M in APAC.

Autodesk reports revenue, with all of the accounting treatments of subscription revenue, as well as billings -- the sum of revenue and the net change in deferred revenue from the beginning to the end of the period. In other words, revenue actually received and bills sent out but not yet paid by customers. And that's where investors were disappointed: in FQ2, billings were down 12% from a year ago to $787 million, and the company forecast billings for the year to be down by as much as 3%. Since invoices sent out this quarter turn into revenue in some future quarter, when billings decline, that's a cause for concern. Why did Autodesk say its billings would go down? Because it saw a dip in the contribution from multiyear contracts when compared to prior quarters, a trend that Mr. Herren was beginning to reverse itself towards the end of fiscal Q2. My take on two possible reasons: first, customers have less confidence in Autodesk's ability to deliver value to its subscriptions, meaning a growing "show me" attitude even in the face of discounts for longer periods. Second, less confidence in customers' need for the software in the far-of future -- they don't need subscriptions for workers they aren't sure they'll still have. We'll have to tune into this metric in FQ3 to see what develops.

Mr. Anagnost also commented on where billings are coming from -- online versus indirect sales versus direct sales. He said that "we saw strong double-digit billings growth through the online channel during the [fiscal second] quarter. Our online sales are helping attract new customers to the Autodesk family, as nearly three out of four new customers in the quarter came in through e-commerce". In general, he said, "We're still trying to get that direct online business up to 25% of our total business." That's been the stated goal for quite a while -- and since Autodesk is taking more of the high0end / large account business direct, this squeezes the reseller channel.

Autodesk didn't release channel performance data for FQ2, but Mr. Herren said that Autodesk saw a strong quarter among smaller accounts, which matches what other PLMish companies told us -- smaller decision teams, faster cycles, are possible at smaller companies. The mid-market was "tepid" as it waited for Autodesk's multi-user to named-user deals to kick in at the start of FQ3, and as they paused to assess their prospects for the rest of the year. The named accounts (biggest prospects) "didn't have a big Q2 [but] that seems to be heavier in the second half of the year. We've got a very full pipeline of large transactions, large accounts, EBA [Enterprise Business Agreements-- token-based access to a pool of products over a defined period] renewals that are coming up." [UPDATE: My bad. Autodesk did release that 30% of revenue was direct and 70% came from indirect sources, on par with prior quarters.] 

Mr. Herren told investors that large deals are still hard to close and that the (re)opening activities around the world create a complex mix of business climates. He sees "varying degrees of demand in the Americas, which includes our largest end market. At the upper end of our guidance range, we are modeling meaningful recovery in the region in the third quarter, with continued improvement in the fourth quarter. At the low end of the range, we anticipate a slower recovery in the third quarter and improvement in Q4" That translates to a forecast of FQ3 revenue between $930 million and $945 million. For fiscal 2021 (ending January 31, 2021) Autodesk sees revenue of $3,715 million to $3,765 million, up 13.5% to 15.%. That's a slight increase in the midpoint of the guidance, and a decrease in the range -- as usually happens as we get closer to the yer-end.

And now, the elephant in the room: Mr. Anagnost listed a lot of AEC wins in his opening remarks, and spoke further about the unhappy architect customers who are causing such a kerfuffle by answering an investor question this way:

"[These customers] have legitimate concerns about the functionality in Revit and we take those incredibly seriously. And the fact is, is that from an architectural standpoint, Revit hasn't gotten a lot of incremental investment. A lot of [our] AEC investments have gone to construction, to revenue enhancements targeting the engineering component and workflows– structural workflows, in particular. So, there are some real, legitimate concerns there.

"The other concern they have is the move from multiuser to named users. These are large multi-user clients and they've seen multi-user prices drift up. They really want a pay-per-use model. We want them to have a pay-per-use model, which they would prefer to a cloud licensing. We're all on the same page.

"But that said, these customers come from a highly privileged, roughly 20% of our subscription base, that moved from maintenance to subscription and have pretty deep price protections relative to the rest of the base. And if you look at their expenditures over a five-year period, frankly, even moving out another five years, as they add seats, they are actually paying less to Autodesk than they would have under the old perpetual model. And that was a deliberate part of the transition, even as multiuser prices go up in everything. If you add up what they would have paid us for adding users over time, they actually end up paying less over a five-year period and, frankly, as they add users over a 10-year period.

"We're not concerned about that. We said very early on that we were going to take care of these maintenance customers ... We did that. Lots of debates with all of you [investors] about the maintenance subscription program and 10-year price lock. It wasn't exactly something that all of you were behind. But we think it was right. And yes, it has resulted in this.

"We're never going to be on the same page with this audience [meaning, investors] about that particular part of the equation. But remember, this is a shrinking bit of our subscription base, the protected 20% now. There'll be less than that later. But, over time, they pay less than they used to in the old perpetual model."

That started out so well, got lost a bit in the middle, and perhaps recovered towards the end. But it didn't address the fundamental question: when will Autodesk provide more value to these disaffected customers? Admittedly, Mr. Anagnost is in a tough spot in speaking to investors who want one thing (more revenue and profit) about customers who want lower prices (meaning, less revenue to Autodesk). But it's completely of his own making: framing the transition to subs as a way to raise revenue per customer was never going to have any other outcome than this unless Autodesk over-delivered on product-related promises.

We did get a glimpse into Autodesk's thought process on R&D. Later in the call, answering a question about roadmaps, Mr. Anagnost added,

"[The question] is, where we put new dollars. So, for instance, at the beginning of this year, this whole concern around architecture and architects, is something we saw coming because this has been a five-plus year kind of tension. We actually increased investment in AutoCAD Architecture at the beginning of this year. We used incremental R&D dollars to increase investment in that space.

"Moving forward we will deliberately choose where we add incremental investment, and we've been very forthright with the construction space in terms of incremental investment. We're not going to shift money away from that. But as we add incremental investment into next year and year after that, we'll probably add more incremental investment into other places over time.

"We're in the enviable position to be able to [add incremental nivestment], we're spending more in R&D than we ever had in our history. And we have still room to invest more, we're just going to choose deliberately to add incremental investment in certain spaces, like we did at the beginning of this year for architecture."

Notice it's not Revit. And if someone saw it coming, why let it get to this point?

Topic switch to manufacturing, where revenue grew 6% in FQ2. Mr. Anagnost told investors that "we're growing faster than our biggest competitor in the space. We had good strong growth coming into the year so we're comparing 6% to a good year last year -- we're actually happy with the performance we're seeing right now. And it's only going to continue to get better".

Last thing: AutoCAD and AutoCAD LT. We don't hear much about it, but it's likely the most used CAD product on the planet. Mr. Anagnost said that "We used to talk about that as the canary in the coal mine for market dislocation at the low end but subscription changes everything. The subscription price point for LT is very attractive and most of the customers that are buying it are small to medium businesses. It does what they need." Autodesk used to talk about AutoCAD and LT as entry points into the Autodesk product family; it clearly has value on its own, too.

Well. Verticals and geos. Subs and maintenance. Low- to high-end. A very wide-ranging call with investors, a solid FQ2 and decent outlook.

Note: the quotes are from my notes, checked against the recording of the earnings call, which you can get to here. Listen for yourself!