It’s been a while since Autodesk reported results for the fiscal fourth quarter and year ended January 31, and I encourage you to look at the details here, if you’re interested. Given that Autodesk is moving its users from perpetual to subscriptions, the results have lots and lots of metrics that look bad –Q4 revenue down 26% as reported (and down 25% in constant currencies, cc), which means all categories and geos are down, too– and factoids that are supposed to help investors understand the transition and see the company as a valuable asset in their portfolio –like “New model ARR”. I don’t find any of that news particularly troubling since (math alert) a subscription means less upfront revenue; and transitional metrics like ARR are simply confusing.

What I want to know, and I think you do, too, is this: does Autodesk’s financial position allow it to serve its customers? Is there long-term viability in the technology solutions Autodesk brings to market, or will the company run out of money/interest/time? And, bottom line: are customers moving to subscriptions or are they leaving Autodesk in search of a more comfortable business model?

I ranted about this last week, and want to thank everyone who got in touch via email, comments and LinkedIn. Many in that bunch are angry — they recently bought a perpetual and now see their maintenance fees go up. Some are sticking with Autodesk because of legacy data, work process, training issues and so on. Others are looking at alternatives. Once again, there’s no “best” solution that works for everyone and I urge you to check out the comments here and here for ways to approach the problem.

Given the tone of the customer comments and emails, what is Autodesk seeing? Are people jumping ship? Is revenue leveling out soon, and are Autodesk’s costs aligned to whatever that level is/will be?

Autodesk shared the following in its Q4 announcement:

• New model subscriptions were up 227,000 sequentially to 1.09 million. “New model” means product subscriptions, enterprise flexible licenses and cloud subscriptions
• Maintenance subscriptions were down 73,000 from Q3 to 2.02 million
• Total subscriptions were 3.11 million, a sequential net increase of 154,000. For comparison, 121,000 net subscribers were added in FQ3 (not counting those from the Solid Angle acquisition).

The way I read this is is positive: 154,000 new subscribers plus 73,000 who switched from perpetual maintenance to subs. Of course, “new” is a squishy term here. They could be truly new customers, never using an Autodesk product before. Or they could be legacy customers, finally paying for a new version of a product. Or they could have been off-license users, getting legal. The important factoid is that 154,000 users signed up, meaning that there’s a lot more activity here than just maintenance users switching to subs. I’m not sure Autodesk will ever tell us (even if they could) how many users they’ve lost as a result of the subscription business model change, but it seems that many are sticking with Autodesk.

Of concern though might be this: FQ4 revenue was $479 million, down 26% because math, while the cost of running the business (R&D, selling, marketing and so on) was $646 million, down 2% from a year ago but still bigger than incoming revenue. If Autodesk revenue were to continue to decline faster than expenses, this would pose a problem. Spending more than comes in isn’t a winning strategy for businesses or individuals. But the company continues to generate positive cashflow, and the situation won’t last much longer. Autodesk expects revenue to start trending upwards by fiscal 2019 (really, next calendar year) while it maintains tight spending controls (see below).

For the immediate future, though, forecasts didn’t quite meet Wall Street expectations. The company sees FQ1 revenue of between $460 million and $480 million leading to fiscal 2018 revenue of $2,000 billion to $2,050 billion, roughly flatish for the year. The 8% FQ1 year/year decline is, again, math: no perpetuals to sell.

A couple of other things of special note in the FQ4 earnings:

• CFO Scott Herren said that total direct revenue in FQ4 “increased once again and represented 32% of total revenues — that’s up from 23% in Q4 last year and just 19% two years ago”. Depending on where you stand, that could be good or bad. Autodesk built its business on a strong VAR channel, which is losing ground as more people buy subscriptions via online stores and Autodesk boosts its (direct) large account activity. To some extent, this is inevitable: big companies want to deal directly with the vendor and not with an intermediary. As Autodesk technology became more specialized and strategic to its customers, this was bound to happen. But Autodesk still has a huge number of small customers, and I hope their Autodesk channel partners stay strong and relevant.
• Mr. Herren also reconfirmed Autodesk’s earlier guidance on spending, saying that the company “remain[s] committed to keeping spend flat through fiscal 2019 and believe that we can do so without compromising the long-term health of the company.” That affects many things, including hiring, new offices and acquisitions but the company announced no new cost reduction programs.
• Cloud matters. Co-CEO Amar Hanspal told investors that FQ4 was a “record quarter in the cloud with three times the number of subscriptions of any prior quarter”. Some of that may be due to promotions and bundling cloud credits into desktop offerings, but a lot of the concerns people had about cloud (the Amazon AWS outage notwithstanding) seem to be disappearing. This is the strongest metric from the PLM universe to date, even if it is on top of what is likely a smallish number.
• Regarding the move to subs, Mr. Hanspal said that Autodesk’s biggest customers are “responding very well to the flexibility of subscriptions — it suits their way of doing business. At the [other] end we are seeing people respond to the lower cash outlay the subscriptions demand from them. And the channel is really energized and is excited …. So I think right now it feels like the right set of things that are under way.” I don’t see Autodesk rolling this back.

Finally, you know that Carl Bass is no longer CEO of Autodesk. Co-CEOs Amar Hanspal, SVP Products, and Andrew Anagnost, SVP Marketing and the architect of the move to subscriptions, are now in charge. An analyst on the call asked them how they were dividing responsibilities; here’s what Mr. Hanspal said:

We are sharing responsibilities rather than dividing them. Andrew and I have been the co-architects of the plan that you’re seeing unfold right now. We also continue to work very closely together on the technology transition as well as the business model transition. He and I have been focusing on driving greater focus and urgency on the execution of the plan, and are making decisions jointly through this period while Andrew continue to do his marketing role and I continue to do my product role. The executive decisions we are making, we’re making jointly.

Mr. Anagnost added

We started this mode well before the announcement of Carl’s departure. We’ve work together for a long time so we have a pretty good cadence set up already. There was no disruption as we move from the pre- to the post-Carl era.

Can’t wait to see what happens next in the post-Carl era.