Autodesk Inc (ADSK) 2018 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Q4 Fiscal 2018 Autodesk, Inc.

Earnings Conference Call.

(Operator Instructions) As a reminder, this conference call may be recorded.

It is now my pleasure to hand the conference over to Mr. Dave Gennarelli, Investor Relations.

Sir, you may begin.

Thanks, operator.

Good afternoon.

Thank you for joining our conference call to discuss the results of our fourth quarter and full year fiscal 2018.

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 first quarter and full year fiscal 2019; our long-term financial model guidance; the factors we use to estimate our guidance, including assumptions regarding ASC 606 and tax reform; our maintenance-to-subscription transition; ARPS; customer value; cost structure and market opportunities and strategies; and transfer 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 the fiscal year 2017, our Form 10-Q for the periods ending April 30, July 31 and October 31, 2017, 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 will 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 will quote a number of numeric or growth changes as we discuss our financial performance, and unless otherwise noted, each such reference represents a year-on-year comparison.

And now I'd like to turn the call over to Andrew.

Thanks, Dave.

Q4 was another milestone quarter for our subscription transition.

Key to the quarter was the strong growth in both ARR and ARPS, even as subscription additions fell below expectations.

We finished the year with better-than-expected performance on many of the traditional financial metrics such as revenue, deferred revenue, EPS and cash flow, all of which will be coming more relevant again as we pass the inflection point on our business model transition.

Overall, these results bolster our belief in our ability to achieve our fiscal '20 goals around ARR and free cash flow.

There are several key outcomes of Q4 that I want to highlight.

Total annualized recurring revenue or ARR grew 25%, and subscription plan ARR more than doubled again.

Both ARR and subscriptions for subscription plan are now greater than the ARR and subscription base for maintenance, which is a significant milestone and in line with our projections when we began the transition.

Annualized revenue per subscription or ARPS inflected up in Q4, also in line with our projections.

Recurring revenue has increased to 93% of total revenue.

And we continue to see faster-than-expected migration of maintenance customers to subscription with the maintenance-to-subscription program or M2S.

Beyond that, we remain enthusiastic about our long-term market expansion initiatives in both manufacturing and construction as we continue to introduce new technology that brings design and make closer together and drives the convergence of manufacturing and construction.

Now I know you want to hear what's going on with sub adds, and I will get to that, but let's first start by talking about ARR.

The continued positive trends we're seeing in ARR are clear signals that the transition is working.

I'll stress once again that ARR is the most important metric when evaluating the health of the business at this stage of the transition.

The strength in total ARR was broad-based, with all 3 major geographies growing ARR at 20% or more.

Subscription plan ARR more than doubled, driven by growth in all subscription plan types but led by product subscription.

We continue to drive tremendous growth in product subscription ARR on both a year-over-year and sequential basis.

Now let's talk about subscription additions.

To provide you more insight into the subscription dynamics, we need to break out core and cloud subscriptions.

To be more explicit, the core business represents a combination of maintenance, product subscription and EBA subscriptions, while the cloud business represents all the results generated by standalone cloud offerings.

When you break these 2 views out, our core business, which drives the overwhelming majority of our revenue, ARR and billings growth, is performing quite well.

Our cloud business performed up to our reset expectations for the quarter.

We added 45,000 cloud subscriptions in Q4, which represented a nearly 50% decrease against the tough compare to Q4 of last year when we ran a ceding program for a component of BIM 360.

However, from a billings perspective, cloud had its biggest quarter ever, fueled by several large wins, including 6 with top rank construction companies.

We see continued momentum in terms of customers moving to higher value products like BIM 360 Docs and Field.

We remain very enthusiastic about the opportunity with our cloud products, but keep in mind that cloud is still a small contributor to our overall business.

ARR for standalone cloud grew 23% in Q4 but contributed less than $100 million of our total $2.05 billion in ARR.

So while cloud will not be a major driver of our FY '20 performance, we remain confident it will be a major component of our business in the years beyond FY '20.

This brings us to the question, why did the net sub additions fall short for the quarter?

The answer is that we experienced greater-than-expected subscription consolidation as customers are reducing their total subscriptions in favor of collections.

We're seeing this reflected in the general adoption of collections where the mix of collections within the product subscription base more than doubled year-over-year and now represents over 20% of the product subscription base.

The good news is that most of these customers are increasing their total spend with Autodesk, contributing to solid increases in ARPS and ARR.

So our core strategy of driving upsell to industry collections is working better than we anticipated.

We're also seeing the impact of collections upselling with both regular renewals and with customers participating in the M2S program.

And the upsell effect related to M2S is happening across all geographies.

To give you an idea of how this works, I'll give you a real example of a Canadian engineering firm and their M2S transaction from Q4.

They had 42 maintenance subscriptions up for renewals: 20 AutoCADs, 21 Navisworks and 1 premium suite.

They migrated to 19 subscriptions of the AEC collection and 1 AutoCAD subscription, a net reduction of over half their seats.

However, the account value for this customer increased by over 10%.

This happens enough times and you get a depressive impact on our net sub adds as a result of collections upsell but a pronounced increase in overall ARR.

So that should help you understand what's happening.

This is a positive outcome for ARR and ARPS, but it negatively impacts our net sub adds.

Beyond that, we believe this issue will work itself out over the course of the year as most of our largest customers complete their maintenance-to-subscription migration.

Despite the impact from collections upsell, sub adds in our core business increased 14% year-over-year and accelerated from the prior 3 quarters, led by a record number of product subscription additions.

Even when normalizing for M2S, the base of product subscriptions nearly doubled year-over-year, and EBA sub additions increased over 30%, fueled by the strong Q3 EBA deal activity.

The core business drove more than $1.9 billion of our total ARR in Q4 and grew more than 25%.

Another consistent attribute of the transition is that new customers continue to make up a meaningful portion of product subscription additions and represented close to 30% 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 alternative design tool.

We'll go into much greater sub results and modeling details at our Investor Day in a few weeks, but there wasn't anything in these numbers that alters our conviction in our ability to drive ARR and cash flow.

Now I'll turn it over to Scott for a few more details on the M2S program, ARPS and other financials.

Scott?

Thanks, Andrew.

Subscription plan subs grew by a record 371,000 during Q4, with growth in net subscriptions coming in all 3 categories: cloud, enterprise and product subscriptions.

Partially offsetting the growth in subscription plan subs was the expected decline in maintenance plan subs, primarily related to the M2S program.

The M2S program continues to progress faster than expected, especially in the Americas.

In Q4, customers migrated 168,000 maintenance subs to product subs.

Similar to last quarter, approximately 1/3 of all maintenance renewal opportunities during Q4 migrated to product subscription.

[Of] those that migrated, over 1/3 of eligible subscriptions upgraded from an individual product to an industry collection, which is the highest upgrade rate we've seen yet and relate to the collection's upsell effect that Andrew just spoke about.

The renewal rate for maintenance customers held steady in Q4.

However, remember that Q4 has the biggest pool of maintenance plan renewal opportunities.

Consequently, the decline in maintenance subs is always greatest in Q4.

We're very pleased with the M2S program to date, and we'll continue to encourage maintenance customers to move sooner rather than later.

We expect fiscal '19 to be the biggest year for M2S migrations.

It makes more economic sense for our customers as the cost of staying on maintenance will be higher than the cost to migrate.

And product subscription provides them the greatest value with increased flexibility, support and access to our cloud products.

Now I'll talk a little bit more about annualized revenue per subscription or ARPS.

This is the anticipated quarter where we saw ARPS inflect up for all the reasons we've been calling out, including improvements to the product mix and the geo mix and the base of our product subs, the price increase for the M2S program, less discounting and promotional activity and selling more direct to our customers through our eStore.

Collections upsell is also having a positive impact on ARPS.

Our total ARPS grew 5% year-over-year and 4% sequentially.

Breaking it down, maintenance plan ARPS continues to grow as expected, driven by mix and the annual price increases we rolled out as part of the M2S program.

Product subscription plan ARPS showed a 6% sequential growth.

If we exclude the effect of M2S, product subscription ARPS grew 8% sequentially, had its fifth consecutive quarter of sequential growth and grew 20% year-on-year.

That's meaningful growth in ARPS for the largest component of our core business.

Further, if we isolate on our core business, which again is maintenance plus product subs plus EBA subs, core ARPS grew 10% year-on-year and 5% sequentially.

These are the ARPS trends we've been predicting since the start of the transition and I know have been a source of question for many of you.

Looking at our business mix, once again, total direct was 30% of the Q4 mix.

One of the key investment areas for Autodesk has been our digital infrastructure but the goal of making it easier for our customers who choose to do business directly with Autodesk.

Our eStore is a big part of that, and we're very pleased that we've already grown that channel to nearly $100 million in fiscal 2018 revenue.

In addition, our eStore generated approximately 20% of the product sub sales in Q4, and close to 50% of LT subs in the Americas came through the eStore in Q4.

That's tremendous progress in a short amount of time, and we expect to see this continue to grow.

The biggest component of our direct mix is still the business we do with large enterprise customers.

Q4 is always our biggest quarter for large deal activity, and we signed a record number of $1 million-plus deals in Q4, over 70 of them, including 14 contracts valued at $5 million or more.

Most of these large deals were EBAs.

And on average, the contract value per EBA renewals increased over 40% compared to the original EBA contract value.

For those of you who might not be as familiar with the history of Autodesk, these large deal stats are quite remarkable, even compared to just 5 or 10 years ago.

Through our product innovation and forward thinking, Autodesk has evolved to become a trusted partner and thought leader with our customers.

Many are now coming to us seeking our guidance on how to prepare for the confluence of design and make, which is already happening in certain industries.

Moving to spend management.

We continue to be able to execute well while keeping spend flat on a constant currency basis for both Q4 and fiscal 2018.

The restructuring action we initiated last quarter is allowing us to reallocate our spend, to increase investment in areas that drive long-term value while reducing spend and making targeted divestments in other areas.

We also remain committed to keeping fiscal 2019 non-GAAP spend flat at a constant currency relative to our fiscal 2018 budget at about $2.2 billion.

Looking at the balance sheet, reported deferred revenue grew 9%.

At the same time, unbilled deferred revenue increased by $178 million sequentially, bringing total unbilled deferred revenue to $326 million.

As a reminder, this completes the first full year of moving our enterprise customers to annual billing terms.

If we consider total deferred revenue as reported deferred plus unbilled deferred, which is a fair comparison for last year, deferred revenue grew more than 25%.

Since most of our enterprise customers are on 3-year contracts, an entirely new group of enterprise customers will come up for renewal this year and next year.

So the amount of unbilled deferred revenue will continue to grow meaningfully.

Q4 cash flow was stronger than expected, driven by good billings linearity in the quarter.

The strength of the Q4 cash flow allowed us to finish the year just in the black, which is also better than expected.

When it comes to capital allocation, our stock repurchase program continues to be primary use of cash, and we opportunistically accelerated that program in Q4, buying back roughly 2.4 million shares.

At our last Investor Day, I indicated that we would use the majority of the $1.7 billion cash balance we had available at that time for stock repurchases.

Since then, we spent over $900 million on share buybacks.

In fact, over the past 2 years, we've reduced our absolute share count by close to 3%, and we remain committed to managing dilution and reducing shares outstanding over time.

Before getting into our outlook, I want to touch on 2 high-profile items that are impacting every company: tax reform and ASC 606.

With the start of the new fiscal year, we have adopted the new revenue accounting standard, ASC 606, and we'll be applying the modified retrospective transition method.

The new standard will not result in a change in timing and amount of the recognition of revenue for the majority of our product subscription offerings and enterprise agreements.

In fiscal '19, the estimated impact will be a net reduction to revenue and EPS of approximately $40 million and $0.15, respectively, compared to what would have been recognized under ASC 605 and a reduction of approximately $20 million in ARR.

We will be required to capitalize and amortize sales commissions under the new standard.

While we do not expect a significant impact on reported expenses for the full year, the timing of when we recognize the deferred commissions by quarter will vary compared to historic seasonality.

606 impacts are greatest in Q1 and then dampen as we move through the year and become nominal by fiscal '20.

And of course, none of the 606 impacts affect cash flow.

Regarding the impact from tax reform, saying it's complex may be an understatement, and clarifications from the IRS seem to come out daily.

But we have enough information to provide the following: all in all, U.S. tax reform is good for Autodesk, whereby the lower U.S. tax rate and the ability to access foreign cash in the future will increase our profitability and help us manage capital more efficiently.

We will utilize our deferred tax assets to offset the cash costs of the onetime transition tax.

We're still analyzing the full impact of tax reform, but we currently estimate our fiscal '19 non-GAAP effective tax rate at 19%.

For fiscal '20 and beyond, we estimate our non-GAAP effective tax rate to be between 17% and 18%.

Now I'll turn the discussion to our outlook.

And I'll start by saying that our view of the global economic conditions remains unchanged from the last few quarters, with most of the mature markets performing relatively well and little change in the emerging markets.

We're providing guidance this quarter under both ASC 605 for comparison to our historic financials and 606 and would expect The Street to model us using the 606 numbers.

Recognize that as we introduce guidance for fiscal 2019, you will be able to fill in the blanks for several fiscal '20 metrics based on our stated fiscal '20 goals.

As Andrew said at the top of the call, we are confident in our ability to achieve our important goals around ARR and free cash flow.

As we head into the growth phase of the model transition, we're bringing back annual guidance on billings defined as reported revenues plus the change in deferred revenues, which should be helpful for modeling our cash flow.

I'll note that while we expect billings to increase by approximately 26% at the midpoint for the full year, billings growth in Q1 will be much more modest due to a tough compare against strong billings in Q1 last year.

Another thing to keep in mind as we model out free cash flow is that there are a couple of onetime impacts to fiscal '19 cash flow that total about $130 million.

These pertain to charges for the restructuring and the exit tax from moving our European operations center from Switzerland to Dublin, Ireland.

These onetime items, together with the strength of our Q4 outflows, will have an impact on the strength of cash flow for Q1, which is likely to be negative.

As we emerge from the inflection in our business model transition, cash flow ramps up quickly through fiscal '20.

Significant part of the ramp is driven by the increase in billings, primarily from what will be a much larger renewal base of product subs and multiyear subscriptions returning to the historic levels we used to see with maintenance.

In addition, in fiscal '20, we'll then have 2 years' worth of unbilled deferred revenue flowing into billings, following our transition to annual billings for enterprise customers.

The underlying positive trends in our business give us confidence in accelerating ARR growth to approximately 30% for fiscal '19.

This growth will be driven by fewer subscriptions at higher ARPS, which reflects the trends we're seeing with both our cloud products and collections upsell.

As such, we've revised our outlook for subscription additions for the year.

As Andrew mentioned, our next Investor Day event is just about 3 weeks from today.

We'll use that opportunity to do another deep dive on the model and provide more details on the recent numbers and the path ahead.

We'll also go into greater detail on subs and ARPS model that get us to the fiscal '20 goals, and we'll revisit our 5-year model.

Now I'll turn the call back over to Andrew.

Thanks, Scott.

I'll reiterate what Scott just said and assure you that we are confident in our ability to accelerate ARR growth to achieve our fiscal '20 ARR and free cash flow goals, and we have a realistic plan in place to achieve those goals.

Given the changes we've seen at the end of this year, it shouldn't come as a surprise that we'll be reducing the subs CAGR and increasing the ARPS CAGR, consistent with the move to fewer higher-value subs in both the core business and in our cloud business.

The transition is on track, and these model adjustments are happening for the right reasons.

We'll go into all the details at our Investor Day on March 28.

Now if we look back at the year, we've taken significant action to realign our investments and position Autodesk to meet our long-term goals.

We are investing and building and expanding the digital infrastructure of the company, increasing go-to-market and development spend for the construction opportunity and maintaining development of our core products.

I'll finish by repeating the 3 strategic priorities that will drive long-term success at Autodesk: completing the subscription transition; digitizing the company; and reimagining manufacturing, construction and production.

We have already made significant progress in addressing the tremendous new market opportunity in the construction market, and we will continue to pursue that market aggressively.

In addition, some of you may have noticed that we are increasing our efforts in the manufacturing space as we just opened a new advanced manufacturing facility in Birmingham to highlight the work we're doing to move product design and manufacturing companies to a new hybrid, additive and subtractive future.

You'll see more of that over the next couple of years, and you'll also begin to see the first results of our efforts to reimagine production.

To wrap things up, I want to thank our customers, but especially recognize our employees and partners who have worked so hard to make last year a success.

We're excited to be now in the growth stage of the transition, to see accelerated growth in ARR and to be another step further along the journey of our transition.

We're confident in our long-term plans and ability to execute while providing our customers with greater flexibility, more compelling products and a better user experience.

Operator, we now like to open the call up for questions.

(Operator Instructions) And our first question will come from the line of Philip Winslow with Wells Fargo.

Obviously, we've been focused on ARPS since mid-2015, so it's really exciting to see the inflection here in that number in Q4.

So congratulations on that.

And my question is on ARPS.

Obviously, you said you're going to give us more details in terms of just the framework for the long-term guidance at Analyst Day.

But wonder if you could talk about sort of '19 and '20, the puts and takes that you see, because there's a lot of things going on.

Obviously, the price increases on maintenance, the less discounting on the M2S as well as just discounting overall.

Just help us kind of frame out the puts and takes here of ARPS as you think about it for 2020.

Yes.

So let me give you a little bit of perspective on what's driving this.

First, kind of let me frame the problem a little bit.

One of the things -- I want to make sure you're seeing the same thing that we saw.

So the core is doing incredibly well.

Subs base grew 14% year-over-year last year.

We're going to see that number continue moving forward, at least at that level, which is in line with historical levels.

So the core is doing really good.

And when you look forward, you guys have probably noticed where we're at with our guide.

Most of that guide is due to our reset expectations in cloud.

So you're zeroing in on exactly the right thing right now on ARPS.

So what's driving the ARPS, specifically?

So what we're seeing is an acceleration of some of the strategies that we already put in place and we've talked about in the past.

The first ones being the upsell to collections.

Like I said in the prepared remarks, we're seeing an acceleration.

More people are -- as they move from M2S, are taking us up on the move to collections.

More people in the run rate of our core business are moving to collections.

You're going to see those phenomena play out into next year and into beyond.

The other thing that's getting accelerated right now is our price realization efforts, and this comes from basically 3 components.

One, better execution on the eStore, which, by the way, very high price realization channel.

Our renewal base changes in terms of the channel costs to renew product subscriptions.

We're making some changes there.

We've made some changes in the low end margins of our business, particularly around LTs.

That is accelerating our price realization.

And we're doing much better in terms of managing the promo discounts across the products.

All of those things and the fact that they are accelerating are going to drive up ARPS as we move into next year and beyond.

And our next question will come from the line of Saket Kalia with Barclays.

First, maybe for you, Andrew.

Thanks a bunch for the example on the Canadian engineering firm to kind of show that net add and ARR dynamic, but just to make sure that we understand.

The question is, are customers buying for fewer engineers?

I guess, why is it that a collection is able to handle fewer seats if, presumably, seats are a good proxy for employees?

And I guess, I think we all appreciate the ARR lift that you're getting but talk us through why you think about this -- I guess, how you think about this dynamic when you think seat share within your customers, if that makes sense.

Yes.

Saket, I'm really glad you asked this question, all right?

So this is the dynamic of subscriptions versus users, all right?

And this is an anomaly of how we're counting our business.

So what you saw in that example is each user actually had 2 subscriptions.

So the number of users in this company did not change at all, it just so happened that each user had a seat of AutoCAD and a seat of Navisworks Manage sitting on their desktop.

So what happens is, is the customer said, "Well, I can now take this M2S offer, move all of those seats to collections." Not only do they get AutoCAD and Navisworks Manage, but they get Revit as well.

And that's what the customer did.

So the number of engineers sitting there, exactly the same.

The number of subscriptions sitting on each engineer's desktop, 1 from 2. Make sense?

That does.

That makes a lot of sense.

Maybe...

We're seeing that broadly.

Got it.

Maybe from my follow-up, just a little bit of extension of Phil's question.

Could either you or Scott just maybe talk a little bit about the shape of M2S in fiscal '19?

I guess, it feels like we saw a nice reaction to some of the pricing strategies here in fiscal '18.

I believe the next action will build on that.

Can you just maybe walk us through how you're thinking about the maintenance space by the end of fiscal '19?

And tying it back to ARPS it back to ARPS, how should we think about the impact that, that could have on ARPS?

Yes, Saket, it's a great question.

We are expecting to see an acceleration of people, maintenance subscribers, that as they come up to the point of renewal, seeing them move to product subscription.

It's already moved faster than we expected.

As you know, in fiscal '18, 110,000 migrations in Q3, 168,000 in Q4.

Both of those numbers were higher than expected.

As we look at next year, the economics change again, and the price to renew goes up 10%, the price to convert only goes up 5%.

So the economics actually swing in favor of conversion even more strongly.

And of course, the product sub has higher value to our customers in terms of access and ease of management.

So I think we'll see it accelerate.

Now what we haven't done is gone into trying to forecast each of the subcomponents to give you a sense of that, but I think our expectation all along was that fiscal '19 was the biggest year for M2S conversions.

And I think seeing how quickly it's already moved in fiscal '18, I think we will see that acceleration happen again in fiscal '19.

In terms of the effect on ARPS, I don't think it has a massive effect either way.

The price to convert or the price to renew is still relatively close, a little bit more expensive to renew maintenance.

But I don't think it will have a big -- I don't think that by itself will have a big effect.

What is likely to have a bigger effect is people who, at the point of migration, from maintenance to subscription, to people who do like the example we used in the opening commentary and actually move up from individual licenses up to the collection.

And our next question will come from the line of Sterling Auty with JPMorgan.

Based on the positive comments that you made around the number of users eligible to move to subscription that did and some of the other elements, is there a read-through on churn?

So in other words, has churn actually improved over the last couple of quarters?

So in fact, Sterling, what we've seen is there's absolutely no change in the churn rate of the maintenance base.

So the maintenance base quarter-over-quarter, year-over-year, is holding solidly at the same kind of renewal rates we've seen historically.

What you just see is more people are deciding to take the maintenance-to-subscription offer, and of those people, even more, an acceleration over the previous quarter, are choosing to go to collections.

But the actual renewal rate of that base has remained solidly intact on a quarter-over-quarter and year-over-year basis and even trend-wise.

And to your point, Sterling, that's an aggregator renewal rate.

We do it on seat count today.

So what it implies is because you've got consolidation going on, the remainder is actually seeing a bit of an uptick in renewal rate, right?

Because some of them are dropping off -- like the example we pointed to, are dropping of certain subscribers as they consolidate up to collections.

So I think it's -- we're very pleased with the renewal rates we're seeing in aggregate.

All right.

Great.

And then one follow-up.

On the EBAs, I think you measure the subscriptions in arrears, and I think that's an element that can cause some noise within the net subscriber count, if you will.

I wonder if you can give us some insight into maybe some previous cohorts.

So an EBA from 3, 4 quarters ago, how does that number of subscriptions' contribution into the total count fluctuate?

Yes.

The biggest impact we see is on the first measurement of MAUs, right?

So a customer that buys an EBA, in every case, has had the large estate of perpetual licenses that are on maintenance that then convert those perpetual over to an EBA.

So the first effect you see is maintenance subs go down, enterprise subs go up.

And then the first thing we -- and because anyone in the company now has access to the license, we count MAUs instead of people who have access to it when they move to an EBA.

After 90 days of measuring the MAUs, that's what gets added to the subscriber count.

And what we typically see is a pretty significant uptick at that first read.

Beyond that, we continue to see growth, and it's that volume growth in usage that drives one of the stats that we talked about in the opening commentary, which is we've had several renewals of EBAs in Q4, and on average, the renewal is 40% higher than the original EBA.

That's driven by usage.

That's not -- it's not as much price-driven as it is by usage.

So big uptick on the first measurement and then some slight upticks out through the remainder of the 3 years.

And our next question will come from the line of Zane Chrane with Bernstein Research.

I was wondering, how should we think about the timing and magnitude of the EBA billing cycle change to deferred revenue for fiscal '19 and '20?

And then the second question, what's been the feedback from customers on the new features and functionality with a new model subscription towards legacy licenses?

Is it the features and new functions that's driving the migration of license-only users?

Or is it more of a financial decision given the upfront cost?

Okay.

Zane, I'll take the front-end of that and let Andrew talk about the value prop of moving to subs.

This was the first year where we migrated the majority -- the overwhelming majority of our EBAs over to annual billings.

So it's still a 3-year commitment but billed annually.

So if you notice, one of the stats that we're providing now is unbilled deferred.

What that represents -- almost entirely represents is year 2 and year 3 of EBAs that we signed this year.

A few other little small cats and dogs in there, but almost all of that $326 million is the second and third year of the EBAs we signed this year.

So you can tell, half of that $326 million will come out in fiscal '19.

The other half will come out in fiscal '20.

In fiscal '19, we'll sign another cohort of EBAs, all right?

And in each case, those will also be on annual billing.

So I expect to see that $326 million balance grow this year as we add 2 more years of unbilled deferred to the balance but only bill 1 year out of that $326 million, right?

So the $326 million will be something higher at the end of fiscal '19.

I think it's pretty straightforward.

And we'll provide you those stats each quarter.

It's pretty straightforward for you to set up a quick waterfall model and understand how that's going to accrete into future billings.

Yes.

And Zane, this is Andrew.

With regards to the other question about what's driving their motion right now, if you're looking at a standalone product, right now, the financial incentive is probably the core thing that's driving the momentum in terms of maintenance-to-subscription moves, because they're essentially seeing the same product with a cloud wrap around it.

We are getting a lot of feedback that the inclusion of some of the new cloud wrappers we put in with the products around collaboration and sharing, in conjunction with the support offering we've entangled with subscription, is adding a lot of value.

When you look at collections and the increase in people choosing collections, that's more than just the financials incentive.

The collections have actually been beefed up progressively.

So for instance, in the product design and manufacturing collection, we've added machining capabilities into the collection, and people are noticing that and choosing the options.

Same with the AEC collection, we've added some core capabilities that are making it more attractive for people to decide to take the collections option as they move.

So it's kind of bifurcated in terms of standalone products and what we're seeing with collections.

Everybody likes the [support].

Got it.

Sorry, what was the last part?

I said everybody likes the support, the added support.

And our next question will come from the line of Michael Nemeroff with Crédit Suisse.

We've been talking with a lot of your channel partners throughout the quarter, and there's obviously a shift occurring in the channel's involvement and your new product sales.

How -- Andrew, how do you view the channel's role in new product sales over the next couple of years?

And what are you expecting from that?

Yes.

So I think I've stated pretty emphatically over time that the channel, one, is still an important part of our business.

As we move out to a steady state, we expect our business to be 50-50 split between direct and indirect.

With everybody seeing a larger business, we'll have a larger direct business, they'll have a larger indirect business.

So the channel is a strategic component of our execution moving forward.

In the new businesses, when it comes to booting up new businesses, a lot of the hard work has to be done by Autodesk first.

Channel partners really, really want to get engaged in new businesses early on, and we do engage them at various points in new businesses.

But it's better if we do some of the early market seeding and market development efforts and then start to migrate the business moreover into the channel.

We've been doing that with BIM 360 and some of the new applications, but we kind of try to take a measured approach.

One area where we've really engaged partners in the future is the Forge platform and building out customizations on top of our cloud platform.

That's an area we're encouraging partners to step up and get engaged in really early because that allows them to stitch together our solutions for their customers and really drive their services business, which we think is not only important for us long term with the fact that we're going to be seeing new types of custom solutions built on Forge, it's important for them.

And our next question will come from the line of Heather Bellini with Goldman Sachs.

I was just wondering, Andrew, how do you feel about cloud as you look further out as being a TAM expander?

There's been a lot of question talking you through kind of how you get here to $6 in free cash flow.

And by the way, I'm assuming you guys didn't explicitly state that, but, I guess, people are asking if that's still your target.

And then when you look out to your further target, the fiscal '23 target, how important is cloud as a TAM expander to get to that level?

And what do you see as being a driver for that to drive adoption?

All right.

Thanks, Heather.

So first off, let's be very clear, we are affirming the $6 and we're committed to that.

So let me be super clear on that one.

So now when we look at the cloud, so let's be super clear about a few facts.

So first off, in Q4, we added 45,000 cloud subscriptions in the quarter, okay?

A lot of people would love to have that number.

So that's the net adds.

Yes, it's the net adds.

So there's a robust add, we have record billings, robust billings growth.

The cloud is right where we expect it to be at this point.

What we've decided to do strategically, and I think it was appropriate, is we've deemphasized and moved away from these super low-end cloud subscriptions we had.

They came under the guise of BIM 360 team and Fusion 360 team and moved much more to our portfolio strategy around things like Docs and Field and the higher-value Fusion offerings.

That was a very deliberate choice, and that's why you see some of these subscription guidances that are being heavily impacted by the change in the way we're executing on cloud.

But now as you look out forward in terms of our business development efforts in the cloud, we're exactly where we should be in the cycle as we move to FY '23.

So one more fact, just to comment on FY '23 before I talk about the cloud, is the business model transformation, the move to subscription of the core business doesn't just suddenly run out of steam as we move to FY '23.

It provides a solid growing foundation that we build upon.

But to get to those FY '23 targets, you're definitely going to see us add a lot more cloud business as we get to FY '23.

It's not necessary for FY '20.

It's absolutely part of the FY '23 plan.

And you're going to see us grow that both organically and inorganically, like we've done with any new business in the past.

And our next questions will come from the line of Jay Vleeschhouwer with Griffin Securities.

First question, Andrew, maybe you touched on this, is the role of the eStore.

I think you said about $100 million from fiscal '18.

And at some point, you expected it to be at least half of the direct business, which would mean 1/4 of the total business.

So if by, let's say, fiscal '22, '23, you're a $4 billion-plus company, you're looking at potentially roughly a $1 billion eStore business.

So if the math is wrong, tell me, but could you talk about the infrastructure scaling requirements or mix expectations that you have to get to that kind of multiplier of your eStore business from fiscal '18?

And then last question has to do with your product road map or your product development execution.

You did that very well starting 15 years ago with the maintenance program moving to an annual release, and that sustained your flywheel for recurring revenue.

Given all the changes in the company, you talked about AU, in terms of core products and changes in R&D, R&D management and all the rest, could you talk about your execution plans for development not just in cloud over the next year or 2 to make sure that you keep that flywheel going for recurring revenue?

Yes.

Okay.

So let me start with the infrastructure question.

So you're right, we'd be moving up close to $1 billion eStore business as we move out into those outlying years.

And infrastructure investment is required to make sure that we managed that base in the appropriate way.

So I talked about this whole effort about digitizing the company and investing in digitizing the company.

We already have the transactional infrastructure to get to that number.

So we could take the orders, we can manage the entitlements, but there's that whole wrapper of infrastructure around $1 billion base.

I mean, just think how many subscribers that's going to be.

It's in the millions, right?

And you want to take care of that base, you want to engage with that base, you want to have intelligence on that base so that you can easily see what's going on with those customers.

That's the infrastructure we're building out.

The inward-facing infrastructure that allows us to understand the customer that came to us electronically but also the infrastructure that faces the customer that allows them to, on their own, without any intervention from an Autodesk person, manage their relationship with Autodesk.

So you'll see us building out both of those pieces of infrastructure over time.

Right now, obviously, we've already got the core transactional infrastructure to drive the results here.

Now when we talk about the R&D cadence, so one of the things I did, and I think you astutely notice this, is that we split the development efforts up into essentially what is the core business design and creation products, Inventor, AutoCAD, Revit, Max, Maya; and the new business products, essentially around the cloud; construction-based products; the new kind of advanced manufacturing products.

One of the key things that's going on inside the design and creation organization is spinning that flywheel on development, what is a rapid cadence of small, incremental additions and upgrade to the product that keep the customer seeing progress.

Now they're going to see progress not only in the product, but for those people who manage large installations over the next 18 months or so, they're going to see progress on how easy it is to manage their installations with Autodesk.

So over a 12- to 18-month period, people are going to see a lot of functionality to show up.

The experience we aspire to, Jay, is kind of like what you're seeing with Office 365 right now where these little kind of updates come in and it gives you a quick little dialogue that says, "Here's the things that I just -- we just updated for your Office experience." That's where we're heading.

We're actually making very good progress in that direction.

And that's how customers are going to be able to digest a continuous stream of upgrades without this kind of big thud that we used to do.

And our next question will come from the line of Ken Talanian with Evercore ISI.

So I was wondering if you could give us a sense of what kind of uptake you've seen from nonsubscribers during the year.

And then any plans to address that base going into fiscal '19?

Yes.

So look, we've -- historically, the nonsubscribers are kind of absorbed into our run rate.

We run promos.

We've been running promos on a cadence of about twice a year.

So you saw that we did fairly successful in the promos this year.

I can't remember exactly the number of nonsubscribers we brought in through promos last year.

It was -- we did 2 promos, one in Q1 and one in Q3.

What we're doing this year as we move forward, one, on the nonpaying, nonsubscriber side, this is the year we're kind of instrumenting our license compliance efforts.

So we're actually putting out more intelligence and capability that actually tell the customer in the product that they are working with a noncompliant copy and maybe they want to investigate how they got there and also allowing them to contact us quickly and buy.

If we look at what we did with the legacy -- what we're doing with the legacy base, the nonsubscribers that have fallen off maintenance and they are sitting on all perpetual licenses, we're actually rolling out a new set of promos at lower discounts but actually had some insurance policies built into it where if the customer moves and falls off 2 years from now, they can default back to the old perpetual release they had at the time they moved.

They don't get the latest, they just get their old release.

So it's like an insurance policy to provide even more attraction to that base to come and move forward.

We're going to see how that works as we move into this year.

This past year, just -- I got the number in front of me, we moved 50,000 users into our paying base from promotions like this.

We actually expect this new promotion to do as well, if not better, than some of those past promotions.

Yes.

And Ken, promotions are one way that we go after the legacy base.

As you know, in many cases, they end up coming back to us as the product they are using, if they're not on maintenance, the product they're using ages out.

And as it gets older and older, their ability to communicate with other partners around them, whether they're in manufacturing and part of a supply chain or it's in their contractor on AEC, as their version becomes more and more down level, their ability to exchange files becomes harder and harder.

File types change through time.

So promos is one way, kind of a financial lever to pull.

The other is they -- as the -- since all of these are network-type used products, in other words, they're used in conjunction with other companies, they end up needing to get to the latest release just to stay compliant.

Great.

And I guess, just a bigger picture question.

Could you rank order what factors might cause you to exceed your current ARR growth targets for fiscal '19?

Yes.

I mean, that's a little bit of a loaded question.

What I would say is that the -- what we've seen so far is faster progression in ARPS growth than we had previously anticipated, a lot of it based on the factors that, again, Andrew talked about earlier.

Some of it is coming from consolidation of subscriptions where people are moving up to the collections at a faster pace than we saw.

The uptick of maintenance-to-subscription has been a bit of a catalyst for that.

eStore has been bigger than we expected, and we'll continue to drive that faster.

Our renewal base -- remember, when we sell a new product through the channel, there's a higher margin for the channel partner than when we sell a renewal through a partner.

And as our renewal base grows out through time, the lower channel cost, in effect, accretes to us as well.

That's also driving ARPS.

So there's a whole set of things that are driving ARPS growth even faster than we had expected.

I'd say that's probably been the bigger upside.

And you've seen subs in particular around cloud, subs are coming in at a slightly lower level but at a higher price point.

And our next question will come from the line of Keith Weiss with Morgan Stanley.

So not to be too bearish, but we've seen 2 quarters in a row where the net subscriber adds have come in sort of below people's expectations, and you guys have taken your targets down.

And if I'm not mistaken, there's 2 different explanations kind of for what -- why it happened.

Last quarter, it was more about sort of heavily promoted cloud subscriptions not being #1 on a going-forward basis.

And this quarter, it's more about the difference between what a subscriber is and what a subscription is.

And subscriptions -- more subscriptions occurring to [onces] or you're getting -- collecting into 1 collection.

How do we garner confidence that this is the last kind of takedown in terms of our net subscription add estimates on a going-forward basis?

Is there any kind of visibility you could give us into sort of any sort of excess that might be in that base on a going-forward basis that you feel comfortable that this number is not coming down again?

First, let me acknowledge your frustration that you might be having with the way this has played out over the last 2 quarters.

I just have to acknowledge that.

I mean, it doesn't help you with your modeling, it doesn't help you with some of your core efforts.

So I absolutely want to acknowledge that.

I want to just go back to some fundamentals here so we get kind of level set, and then I'll answer your question as specifically as I can.

So first off, remember that we broke out the core and the cloud deliberately now so that you can start seeing dynamics.

The core grew 14% for the year.

So the subs base in the core grew 14% for the year.

That's in line with historical behavior.

We're going to see that same number or greater moving forward.

So the core shows some nice strength and stability, all right?

What we did see this quarter that was not anticipated at the rates was this acceleration in the collections activity associated not only with the run rate with M2S.

And it was an acceleration that was off the trend of the previous 2 quarters in terms of how fast people were moving.

And yes, that did result in this consolidation resulting from upsell but for the right reasons.

We saw the appreciation of ARR associated with that.

So when you look forward to what we've done with the guide, like I said earlier, the vast majority or the -- most of that guide is due to our cloud reset expectation.

And the majority of that cloud reset is due to simply not driving these super low-value cloud subscriptions.

When you look at the core, our CAGR for FY '20 is essentially unchanged for the core business, right?

So what can garner your confidence here is we understand a lot more what's going on with the cloud.

We're now prepared for more acceleration around consolidation as we've seen it move forward.

We know it will play out through FY '19.

And we're holding to the CAGR for the core business that we expect to have in FY '20.

And that's where we're at.

You want to add anything else, Scott?

Yes.

Keith, the only other thing I'd say is even in those 2 quarters, as you point out, with subs coming in, perhaps -- certainly lighter than everyone externally expected in Q3 and lighter than our own expectations in Q4.

In each quarter, ARR performed nicely, and that really has been -- we've said all along, that's our top goal, is to drive ARR and to drive cash flow out of that ARR.

If you remember Q3, ARR grew 24%, so faster than Q2 before that.

Actually, it was the seventh consecutive quarter of increasing growth rate in ARR, and it grew again 25% in the quarter we just closed.

So ARR continues to perform nicely even as subs have been a little bit lower than our expectations since in the last quarter.

I definitely understand that ARR is the key number that you're looking at.

But also want to have confidence in the components that's building out sort of that ARR.

Sort of there's a trend line there going on in line with my model (inaudible).

I get that.

I totally get that.

Now I want you to go back to that core strength that we're talking about here because that's the important driver.

You layer on top of that, the factors we're talking about around ARPS appreciation that were associated with upsell to collections, that are price realization and the reduced promo activity, and by the way, don't underestimate how the acceleration of price realization has been -- is going to affect the year moving forward.

Those are the things you want to pay attention to, the core.

That's why we're trying to help by breaking out the 2.

Longer-term, I mean, do you think there's any chance that you'd have a view on subscribers versus subscriptions?

It's a conversation we have a lot.

Part of the investment -- Andrew has talked about 3 big priorities that we're investing in.

One of them is digitizing the company.

Part of the effort there in digitizing the company is to be able to count users as opposed to subscriptions.

Historically, we sold shrinkwrap product that had a license number with it.

So you can track license numbers.

What you can't always track on a one-to-one basis is who the actual user of that license.

Part of the investment that we're making in digitizing the company will give us much better insight into account-level metrics and then within that usage message.

So yes, we may get to that point, but we're not there today.

And our next question will come from the line of Monika Garg with KeyBanc.

Sorry to go back to the ARPS point.

If you look at your fiscal 2019 guidance, you're guiding ARR to about 30%, but your net sub adds growth is about 14%.

That means you're guiding ARPS growth of some 16-ish percent.

So could you walk through the math and factor kind of how we achieved this kind of 30% increase?

Sure, Monika.

And it sounds like you're looking for 605 guide.

The midpoint of the guide to ARR under 606, which is really what I'd like everyone to adjust their models to, is 29%.

But you math is still directionally right.

And I think the ARPS growth drivers will be the components that we talked about.

It's continuing upsell to industry collections.

And particularly, as maintenance-to-subscription accelerates, what we're seeing is, as people make that choice, of those who are eligible to upgrade, more than 1/3 are taking that, right, which is -- so that's going to push more people to industry collections that obviously drives ARPS up.

eStore will continue to grow.

The renewal base is a big part of the ARPS growth.

It's a significantly higher yield to Autodesk when a channel partner sells a renew versus selling a new product.

And so as renewal becomes a bigger part of our overall revenue, obviously, the cost of selling through the channel comes down.

That drives up our price realization.

And we've made some very specific changes on partner margins around the low end, specifically AutoCAD LT and what those margins look like.

So you add those factors together, along with kind of reduced promotional activity and discounting, that's what drives -- that's what's going to drive that ARPS growth rate.

And if anything, what we've seen is it moved up faster than we expected.

Got it.

Then just as a follow-up.

Your first q revenue growth guidance is somewhere 14%, 14.5%, but your yearly guidance is somewhere 21%, 22%, which would mean steep kind of ramp acceleration in revenue growth toward second half.

Maybe walk us through the factors that relate to the accelerate.

Is it like ARPS growth is more second half-weighted or any of the [other] factors?

Yes, it's exactly what it is.

And if you recall -- so it's ARPS growth, mostly.

We do see some growth in subs throughout the year, but it's mostly ARPS growth.

But remember, too, the maintenance-to-subscription program, the annual change in price points on that, come into effect in Q2.

So that also -- that not only affects Q2, but, of course, when we sell that, it drops into deferred revenue, it comes back out of deferred revenue as the year goes on.

So it has a lesser effect earlier in the year and a bigger effect later in the year.

And our next question will come from the line of Steve Koenig with Wedbush Securities.

Thanks for your explanations on the mechanics of how this is going.

I wanted to ask one question on maintenance subs and then just a follow-up question on ARR and then converting to billings growth.

So on the maintenance subs, Scott, I think you said the expected decline in maintenance subs was primarily related to the M2S program.

So it kind of leads me to the question, if that was the primary factor, what are the secondary factors in the decline in maintenance subs?

Yes, there's always a certain amount of non-renew, right.

We don't have 100% renewal rates, Steve.

So as that -- as the maintenance base is bigger in Q4, that drives the other piece of why you saw the decline in maintenance subs.

So 168,000 of the 244,000 that maintenance came down, 168,000 migrated.

The remainder are some form of non-renew.

Remember the example we gave though, right?

That customer, those were 42 licenses on maintenance that converted over to 20 subs.

So some of that non-renew is actually good news.

It's individual licenses that are converting up to collections.

Got it.

Got it.

Okay.

And then on the follow-up.

Is there anything that could surprise you guys when it comes to eventually converting that ARR to billings growth?

And then I want to sneak in maybe a related question, if I could define it that way.

Thinking about post fiscal '20, how should we think about how ARPS and free cash flow, how that trajectory flows after fiscal '20?

Should the ARPS growth peak in fiscal '20, then decelerate fairly gradually because you're still layering on subscriptions faster than you're [treating] the base?

So how should we think about that longer-term trajectory as well?

Okay.

That's quite a suite of questions there, Steve.

Let me take the billings one first.

The -- I think the other thing to think about as we talk about ARPS and how that's going to grow and how that's going to drive billings out through fiscal '20.

Billings -- so the 2 big drivers of cash flow, which I think is the source of your question, are net income, you have a good sense of that from the targets we've given; and billings, which drives deferred revenue growth.

The only other effect besides the ARPS growth that will drive the billings side of that, that you need to bear in mind is we've always had a steady proportion of customers who buy multiyear.

So if you go back to when -- before we started the transition, we were selling maintenance agreements.

Somewhere between 20% and 25% of our customers who bought a maintenance agreement bought a multiyear.

That's been pretty steady.

Of course, when we launched maintenance-to-subscription, we cut that off.

We cut off multiyear maintenance subs.

What we see, interestingly, in products subs.

off a much smaller base, of course, is many of them are also buying.

In fact, it's almost the exact same proportion of product subs customers are buying multiyear as well.

So as you're thinking about your billings model out through fiscal '20, remember that we've had a decline in fiscal '18 in billings driven by no multiyear maintenance.

The same proportion of customers are also buying multiyear product subs.

And as product subs becomes a bigger part of the base, that's going to drive billings as well.

The other factor to layer in -- 2 other factors to layer in on that, the bleed back of unbilled deferred, which we talked about earlier with Zane's question, make sure you're modeling that in, right?

Do we have a year of unbilled deferred that's built up to $326 million?

That will build again in fiscal '19.

And then by fiscal '20, we'll actually have 2 years of unbilled deferred leading back into the billings stream.

So add that in.

And then finally, just from a cash flow, not from a billing standpoint, bear in mind that fiscal '19 has these 2 onetime events that total about $130 million of cash outflows, partly driven by the cash payments for the restructuring, partly the exit tax from moving our operational center in Europe from Neuchâtel, Switzerland to Dublin, Ireland.

And the only point I'll add to that, Steve, just to reinforce what you said about multiyear, the multiyear is not going to happen for no reason at all.

As we enter into fiscal '20, we're actually entering the first group of M2S customers that are seeing their M2S period end.

So they're going to want to lock in their price.

That's been a historic behavior.

What happens is customers say, "Okay, I don't want to see -- I want some price stability out there." So they'll buy multiyear.

And the fact that we had such a large tranche of initial M2S customers, you'll see that just naturally migrate into the business in FY '20.

And then, guys, if you could answer the question I snuck in there, that would be great, which is how should we think about the trajectory after fiscal '20 in terms of growth rates as we layer on the subscription revenue and more of that until you get to more of a steady state?

How should that progress?

Steve, I think it's -- there's no big change to the path that we had laid out on that at our last Investor Day.

Probably the better thing to do rather than try to tackle the moving parts here is to say we have our Investor Day 3 weeks from tomorrow here.

And at that point, we'll not only give you a little more granularity on the path to fiscal '20, but we'll also give you a little more granularity on the 5-year plan.

So maybe the best thing to do is to hold off on that until March 28.

Ladies and gentlemen, this is all the time we have for questions today.

So now I'd like to hand the conference back over to Mr. Dave Gennarelli, Investor Relations, for closing comments or remarks.

Great.

Well, thanks, everybody.

As Scott just mentioned, we have our Investor Day coming up on March 28 here in our office in San Francisco.

We'll also be at the Bernstein conference on April 10 in Boston.

We're going to follow that up with an NDR in Boston on April 11 and in New York on April 12.

If you have any questions in the meantime, you can reach me at (415) 507-6033.

Thanks.

Ladies and gentlemen, thank you for your participation on today's conference.

This does conclude the program, and you may all disconnect.

Everybody, have a wonderful day.