Autodesk: the last perpetual license?
Download case PDF (6 pages)Request teacher’s pack
Case overview
- Audience / course level: MBA and executive strategy
- Company / industry / geography: Autodesk, Inc.; design software; global
- Decision owner and decision date: Autodesk executive leadership; simulated decision as of 23 March 2016
- Information cutoff: 23 March 2016, close of day
- Decision scope: Whether to proceed with the announced 1 August 2016 end of new perpetual Design & Creation Suite licenses or extend a limited perpetual option; not a decision about existing license rights or Autodesk's entire cloud strategy
The case
Opening decision
On 23 March 2016, Autodesk filed its annual report for the fiscal year ended 31 January. Most new perpetual licenses for individual software products had stopped selling on 1 February. A second milestone was four months away: on 1 August, Autodesk planned to stop selling new perpetual licenses for its Design & Creation Suites and affected products.
The company had announced this sequence six months earlier. A customer purchasing an eligible perpetual license before the deadline would keep the right to use that software; customers with Maintenance could continue renewing it and receiving its benefits. The question for this exercise is whether Autodesk's executive leadership should now keep the second phase on schedule or temporarily preserve a new perpetual choice while the subscription model matures. This is a simulated management reconsideration, not a claim that Autodesk disclosed an internal debate or asked shareholders to vote.
Decision question: Should Autodesk complete the move to subscription-only sales of new suite licenses on 1 August 2016, or extend a limited perpetual option for one year while continuing to offer subscriptions?
Assignment: Prepare a two-page recommendation to Autodesk's executive leadership. Choose one alternative, identify the assumptions that matter most, specify three measures to monitor over the next year, and name evidence that would cause you to reverse your recommendation. No forecast is required; do not treat ARR as revenue or cash flow.
A business built around design workflows
Autodesk sells software and services used to design buildings and infrastructure, manufacture products, and create digital media and entertainment. Its portfolio includes widely used desktop products as well as cloud, mobile and flexible licensing offerings. Many customer workflows involve specialized products, collaborators, and files that must be exchanged or revisited over time. A license choice therefore affects more than the payment schedule: it can shape who can enter, when users update, and how customers coordinate around compatible tools.
Historically, Autodesk sold perpetual licenses and offered Maintenance attached to those licenses. Maintenance customers could receive support and product updates while renewing the contract; customers without active Maintenance could continue using the version already licensed. Autodesk had stopped selling upgrades in March 2015. By February 2016, the company had also ended most new individual-product perpetual sales. Its announced next step was to end new perpetual suite sales on 1 August.
The new offers were not all hosted software. Autodesk grouped desktop subscriptions, enterprise flexible licenses and cloud offerings in its “new model.” Management described lower initial cost and flexible access as ways to reach project-based users and small businesses, while emphasizing ongoing product investment and a shift toward cloud and mobile workflows. A perpetual suite customer could buy before the deadline and retain usage rights; active Maintenance could continue to be renewed.
A transition with two clocks
The company was measuring a shift in customer commitments and a change in financial reporting at the same time. At 31 January 2016, total subscriptions had grown 15% year over year to 2.58 million. New-model subscriptions had nearly doubled, but their average annualized revenue per subscription had fallen. Autodesk said the lower-priced offers drove seat growth faster than recurring revenue growth.
The company also warned investors that more revenue would be recognized over subscription terms rather than up front. That changes the timing of reported revenue and can affect billings, margins, earnings, deferred revenue and operating cash flow during the transition. At the same time, a shift in demand, customer renewals, pricing or costs would be an economic change, not just an accounting delay. Autodesk expected Maintenance subscriptions to peak as perpetual sales ended and then decline gradually. It warned that renewals could fall faster than expected or new-model growth could disappoint.
Fiscal 2016 had produced $2.504 billion in net revenue, approximately flat with the prior year, and only $1.3 million in GAAP operating income. It also included a large U.S. deferred-tax valuation allowance after recent U.S. losses. Autodesk attributed operating-income pressure to higher spending and cost of revenue, lower license and other revenue, and higher cloud/hosting costs, partly offset by subscription revenue growth. These results coincided with the transition, foreign-exchange pressure, restructuring and other spending; they do not isolate the transition's causal effect.
A market still choosing its license path
Autodesk's competitors showed that enterprise software firms could make different choices. PTC had begun offering subscriptions for most solutions in 2015 while retaining perpetual licenses. That menu preserved customer choice, though its filing also anticipated near-term revenue and earnings effects as the mix shifted. Adobe's Creative Cloud was subscription-based and bundled frequent updates with cloud storage, file synchronization, collaboration, mobile services and Adobe Stock. That offer was more than a new payment term, but its creative workflow and customer base differed from Autodesk's engineering and design markets.
Autodesk faced its own product and customer mix. A project-based design team may value low entry cost and temporary access. An enterprise with established workflows may prefer shared licenses or predictable access across teams. A smaller customer that uses a suite intermittently may compare cumulative subscription payments with a perpetual license plus optional Maintenance. Autodesk disclosed no customer-level data in this annual report showing which groups were converting, adding seats, renewing, or generating profitable incremental use.
The choice
Autodesk could complete the plan: end new perpetual suite sales on schedule and concentrate future investment, channels and customer migration on subscription and flexible offers. This would make the public transition more consistent and could support broader access and recurring customer relationships. It would also end future up-front suite license sales and opportunities to attach new Maintenance contracts.
Or Autodesk could extend a limited choice: continue selling new perpetual suite licenses for one additional year, through 31 July 2017, alongside the subscription offers, and then reassess. This one-year extension is an author-created alternative for analysis, not a policy Autodesk announced. It could preserve customer choice and reduce the risk of forcing customers into a model that does not fit their usage. It might also prolong two selling systems, slow conversion to the new model, and make future product economics harder to interpret.
Both alternatives leave existing perpetual licenses and renewals of active Maintenance rights intact. Neither gives management a clean experiment: users self-select into offers, product needs differ, channel behavior matters, and financial statements combine the transition with other changes. Management must decide what to learn from subscription additions and ARR, how to protect customers' ability to use their existing software, and which signals would justify continuing or changing course.
Exhibits
Exhibit 1. Licensing transition timeline [F]
| Date | Event | Source and locator |
|---|---|---|
| 6 Mar 2015 | Autodesk discontinued licensing upgrades. | C01, Item 1A / transition discussion, printed pp. 16–17. |
| 3 Sep 2015 | Company announced that new perpetual Design & Creation Suite licenses would end after 31 Jul 2016. | C02, announcement opening paragraphs. |
| 1 Feb 2016 | End of most new individual-product perpetual license sales. | C01, Item 1 Business, printed p. 8; Item 7 transition discussion, p. 37. |
| 23 Mar 2016 | FY2016 Form 10-K filed; information cutoff for this case. | C01 SEC filing index, accession 0000769397-16-000067. |
| 1 Aug 2016 | Planned end of new perpetual suite sales. | C01, Item 1 Business, printed p. 8; Item 1A, p. 17. |
Exhibit 2. Autodesk transition indicators [F]
| Autodesk metric | 31 Jan 2015 | 31 Jan 2016 | Change | Definition and scope |
|---|---|---|---|---|
| Maintenance subscriptions (thousands) | 2,013.3 | 2,151.0 | +7% | Maintenance attached to perpetual licenses; C01 printed p. 45. |
| New-model subscriptions (thousands) | 220.4 | 427.2 | +94% | Desktop, enterprise flexible, cloud and Shotgun offers; C01 glossary p. 14 and p. 45. |
| Total subscriptions (thousands) | 2,233.7 | 2,578.2 | +15% | Company-defined subscription measure; C01 p. 45. |
| Maintenance ARR ($ millions) | 1,103.8 | 1,121.4 | +2% | Three-month average recurring revenue annualized; C01 glossary p. 14 and p. 45. |
| New-model ARR ($ millions) | 151.6 | 255.0 | +68% | Same annualized definition; C01 glossary p. 14 and p. 45. |
| Total ARR ($ millions) | 1,255.4 | 1,376.4 | +10% | Do not add to GAAP revenue or cash flow; C01 p. 45. |
| New-model ARPS ($) | 688 | 597 | -13% | ARR divided by subscription count; blended mix, not a same-customer price change; C01 p. 45. |
Exhibit 3. Selected results and transition context [F]
| Metric | FY2015 | FY2016 | Source / interpretation |
|---|---|---|---|
| Net revenue ($ millions) | 2,512.2 | 2,504.1 | C01 printed p. 43. FY2016 was approximately flat reported; the filing says constant-currency net revenue would have risen 4% (p. 45). |
| GAAP gross margin | 86% | 85% | C01 printed p. 57. |
| GAAP operating income ($ millions) | 120.7 | 1.3 | C01 printed pp. 43 and 57. |
| GAAP net income (loss) ($ millions) | 81.8 | (330.5) | C01 printed p. 57. FY2016 tax expense included a $230.8m U.S. deferred-tax valuation allowance; see C01 p. 55. |
| Recurring revenue / net revenue | 46% | 54% | C01 printed p. 44. A company statistical measure, not a substitute for net revenue; specified offerings are excluded. |
| R&D expense ($ millions) | 725.2 | 790.0 | C01 printed p. 8; R&D was 29% and 32% of revenue, respectively. Correlation with transition is not causation. |
Autodesk's 10-K warns that discontinuing perpetual sales would end future up-front license receipts and future opportunities to attach Maintenance; it also says new-model revenue may not grow as quickly as expected and near-term cash flow may decline while more revenue is recognized ratably (C01, Item 1A, printed p. 17). The same filing identifies employee and cloud/hosting costs among drivers of higher costs in FY2016 (C01, MD&A, printed p. 44).
Exhibit 4. Competitor license choices [F]
| Company | Pre-cutoff disclosure | What it does and does not establish |
|---|---|---|
| PTC | Its FY2015 10-K says it began offering most solutions under subscription in 2015 alongside perpetual licenses. | Shows a hybrid menu was operationally available to an industrial software peer. It does not show relative profit, adoption, or customer satisfaction. C04, Item 1A, SEC filing lines 315–318. |
| Adobe | Its FY2015 10-K describes Creative Cloud subscriptions with frequent updates, cloud storage/sync, collaboration, mobile services and Adobe Stock. | Shows how subscription can bundle continuing service value; product/workflow differences limit direct comparison. C05, Item 1, printed pp. 4–7, 264–278. |
Student questions
- Which customer segments might prefer each license model, and how could their different upgrade and usage patterns affect Autodesk's economics?
- Does the subscription growth in Exhibit 2 indicate that the transition is working? What additional evidence would distinguish new demand from migration, pricing/mix effects and renewal behavior?
- Which alternative do you recommend? What measures should management review after one year, and what results would cause you to change course?