Report contents
At a glance
- Business model: Autodesk sells annual and multi-year subscriptions for design, engineering, and construction software, mostly through resellers, with revenue recognized ratably. File-format lock-in and embedded workflows support renewal pricing, the clearest strength. The obstacle is volume: seat counts follow customer project pipelines and hiring, the construction-facing group is largest and cyclical, and the company does not disclose the split between price and volume.
- Financial quality: In FY2026, free cash flow reached $2.41 billion on $7.21 billion of revenue, 2.2 times net income, with capital spending near $40 million. Deferred revenue of $4.69 billion and cash of $2.67 billion at April 30, 2026 give capacity. Equity is thin at $3.19 billion and goodwill is $4.34 billion; stock-based compensation keeps GAAP earnings well below non-GAAP.
Business model
Business model: subscription seats carry a cyclical customer base
Autodesk sells subscription software that architects, engineers, contractors, and manufacturers use to design, model, and document buildings, infrastructure, and physical products. Named users pay annually or monthly, most of them through resellers, and revenue is recognized ratably over the subscription term. The buyer is usually not the user: design leads, IT, and finance approve multi-year agreements, and the seat count follows the customer's project pipeline and headcount. Autodesk is a mature software company selling per-seat access; growth now comes from adding seats, raising prices, and attaching higher-tier products.
Customers are project businesses, so their hiring sets the seat count
Three bases pay. Architecture and engineering practices, contractors, and infrastructure owners buy the design and documentation tools; manufacturers use them for product design and tooling; media and entertainment studios use them for visual effects and games. The construction-facing group is the largest and the most cyclical, because its own revenue depends on building and infrastructure budgets. Manufacturing seats track industrial capital spending and product launch schedules, and media demand follows film and game production slates that can be shelved quickly. Customers span one-person drafting shops and the largest engineering firms, and the biggest accounts hold a disproportionate share of revenue.
Because contracts are signed in advance, a slowdown reaches Autodesk through renewals. Customers shed seats when projects end, but annual and multi-year terms delay that by a quarter or more, and the same lag slows the rebound. Terms typically run a year or longer, with billing schedules that put cash ahead of recognized revenue, so large renewals can swing a quarter's billings without changing underlying demand. That is why a weak construction quarter shows up in billings and remaining performance obligations before it reaches revenue.
Distribution: direct at the top, rented in the long tail
Large accounts buy directly. Smaller firms are reached through a global reseller network that handles installation, training, local-language support, and credit, and those partners decide which products land on a small firm's short list. Autodesk shares the small-account relationship with partners whose economics have to keep working. A move to capture more of that revenue risks the coverage that makes the fragmented end of the market reachable at all. That structure also shapes acquisition cost: winning a firm through a partner is cheaper than building a direct sales team for every market, but it caps how much of the customer relationship Autodesk controls. A second layer is the developer ecosystem: third-party add-ins run on Autodesk's file formats and APIs, deepen customer workflows, and make the platform harder to leave. Reach is global, with the Americas the largest billing region and Europe, the Middle East and Africa, and Asia Pacific the other two.
Renewal is where pricing power shows
Most of the installed base renews each year, and a firm that has built its drawing library, custom scripts, and staff skills around Autodesk faces real cost in switching. That gives leverage on price at renewal, a cleaner lever than winning new seats in a soft construction market. The catch is attribution. Annualized recurring revenue, billings, and remaining performance obligations measure the size of the subscription base, not the split between price and volume, and a net revenue retention rate, where the company reports one, folds the two together. Flat seat counts can sit alongside respectable recurring-revenue growth for a while, which is why the mix is hard to read from outside. Published price-list changes are visible; enterprise discounts are not.
The profit pool sits in the design seat
Software is a small line in a construction project's budget next to labor, materials, and equipment, but the drawings it produces govern the work. A price increase that is modest against project cost is meaningful against Autodesk's own revenue, which is why renewal pricing has held even when customers are under margin pressure. Those same firms watch seat counts closely, because unused subscriptions are among the most visible costs they control. The limit on pricing is the customer's project pipeline.
Adjacent products test the model more than the core does
Autodesk has pushed beyond the core design seat into cloud products for construction management, data collaboration, and manufacturing workflows. These raise revenue per account but compete with specialist vendors and need a direct sales motion the reseller channel does not fully cover. Attach rates will decide whether the effort produces durable recurring revenue or mostly higher selling costs. Core fragility runs the other way: seat-based pricing in cyclical end markets makes the model only as steady as customer hiring, and the small-account end is where cheaper point tools can take hold.
Autodesk's renewal base supports durable pricing power, while the volume side depends on construction and industrial hiring, and price cannot carry recurring revenue indefinitely on a static seat base. What would overturn that: seat additions stalling while published price increases keep doing the work, or churn concentrating among smaller customers who use only part of the product. The disclosures do not separate price from volume, so the read has to come from recurring revenue growth set against pricing changes, partner terms, and product mix.
Questions worth following
- Can Autodesk grow seats without leaning on price increases?
Recurring revenue can rise on seat additions or on price. If price does most of the work, growth depends on customers accepting annual increases while their own project pipelines are flat, which is harder to sustain and invites cheaper alternatives into the account base. The split decides how durable the revenue is.
- Will weaker construction demand cut Autodesk's seat growth?
Construction and manufacturing customers add seats with project volume and headcount, so a downturn would slow net subscription additions and, with a lag, recognized revenue. The effect is cyclical, but price increases deferred during a weak stretch can outlast the downturn itself.
- Can Autodesk's reseller channel keep winning small-account seats?
Small and mid-sized firms supply much of the seat volume, and they buy through partners. If partner economics stop working, coverage thins at the low end, where cheaper point tools are good enough for single-discipline work. That would slow seat growth well before it shows up in large-account renewals.
Financial quality
Financial quality: strong cash conversion, thin reported earnings
Autodesk turns revenue into cash reliably, but its reported earnings are much smaller than the cash flow, and its equity base is too thin for return ratios to mean much. Free cash flow was positive in each of the last five fiscal years and reached $2.41 billion in fiscal 2026 on $7.21 billion of revenue, a 33% margin, with capital spending of roughly $40 million. The gap sits on the other side of the ledger: Autodesk reported GAAP operating margin of 22% against 38% non-GAAP for fiscal 2026 in its February 26, 2026 fourth-quarter release, and much of that 16-point spread is compensation paid in stock rather than cash.
| Metric | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 | Q1 FY2027¹ | Assessment |
|---|---|---|---|---|---|---|---|
| Revenue (USD bn) | 4.39 | 5.00 | 5.50 | 6.13 | 7.21 | 1.93 | Strong |
| Gross margin | 89.6% | 89.5% | 89.8% | 90.6% | 91.0% | 91.2% | Exceptional |
| Net income (USD bn) | 0.50 | 0.82 | 0.91 | 1.11 | 1.12 | 0.49 | Adequate |
| Net margin | 11.3% | 16.6% | 16.7% | 18.1% | 15.5% | 25.4% | Adequate |
| Operating cash flow (USD bn) | 1.53 | 2.07 | 1.31 | 1.61 | 2.45 | 0.89 | Strong |
| Free cash flow (USD bn) | 1.47 | 2.02 | 1.28 | 1.51 | 2.41 | 0.88 | Strong |
| FCF / net income (x) | 2.9 | 2.5 | 1.4 | 1.4 | 2.2 | 1.8 | Strong |
| ROE | 54.8% | 82.5% | 60.4% | 49.7% | ≈40% | n/a | Adequate |
| ROA | 6.3% | 9.1% | 9.4% | 10.7% | ≈9.6% | n/a | Adequate |
| Total assets (USD bn) | 8.61 | 9.44 | 9.91 | 10.83 | 12.47 | 11.93 | Adequate |
| Liabilities / assets | 90.1% | 87.9% | 81.3% | 75.8% | 75.5% | 73.3% | Adequate |
| Cash (USD bn) | 1.53 | 1.95 | 1.89 | 1.60 | 2.25 | 2.67 | Strong |
| Goodwill (USD bn) | 3.60 | 3.62 | 3.65 | 4.24 | 4.29 | 4.34 | Weak |
| Current ratio | 0.69 | 0.84 | 0.82 | 0.68 | Not disclosed | Not disclosed | Adequate |
¹Three months ended April 30, 2026. Fiscal years end January 31. *Calculated on average equity and average assets.
Falling ROE reflects a rebuilding equity base, not weaker economics
Net income more than doubled from $0.50 billion in fiscal 2022 to $1.12 billion in fiscal 2026, yet ROE fell from 54.8% to roughly 40%. Equity rose over the same stretch, from $0.85 billion to $3.04 billion, so the ratio compressed for arithmetic reasons. The more informative measure is the return on the asset base: operating income of $1.79 billion was about 14% of year-end assets, and roughly 22% once $4.29 billion of goodwill is excluded, a calculation rather than a disclosure. That is a good business earning solid returns, not a 50% machine. Fiscal 2026 net income was also flat against fiscal 2025 despite 18% revenue growth, because the tax provision rose to $0.48 billion from $0.27 billion, or about 30% of EBIT against 20% a year earlier. If that higher rate is structural rather than a one-year swing, reported earnings power is lower than the top-line trend implies.
Leverage in name only, with one disclosure gap
Liabilities of 75.5% of assets and a current ratio between 0.62 and 0.84 would normally read as strained. Here they largely reflect money customers pay before Autodesk delivers: deferred revenue was $4.69 billion at January 31, 2026, about half the liability base, with total remaining performance obligations of $8.3 billion, more than a year of revenue. That settles in software access, not cash. Cash stood at $2.25 billion at year-end and $2.67 billion at April 30, 2026, against $0.05 billion of short-term debt, and $1.36 billion of financing outflows in fiscal 2026 were comfortably covered by free cash flow. Long-term debt near $2.0 billion to $2.5 billion appeared on the balance sheet through fiscal 2025 and the second quarter of fiscal 2026 but not on later dates I can compare; the fiscal 2026 annual report's debt note is the place to confirm whether it was repaid or simply not broken out.
Quarterly pattern: one cash-heavy quarter and a guidance-beating start
| Metric | Q1 FY26 | Q2 FY26 | Q3 FY26 | Q4 FY26 | Q1 FY27 | Assessment | Trend |
|---|---|---|---|---|---|---|---|
| Period ended | Apr 30, 2025 | Jul 31, 2025 | Oct 31, 2025 | Jan 31, 2026 | Apr 30, 2026 | ||
| Revenue (USD bn) | 1.63 | 1.76 | 1.85 | 1.96 | 1.93 | Adequate | Growing |
| Gross margin | 90.2% | 92.3% | 91.4% | 91.3% | 91.2% | Exceptional | Stable |
| Operating income (USD bn) | 0.34 | 0.45 | 0.47 | 0.53 | 0.54 | Strong | Improving |
| Net income (USD bn) | 0.15 | 0.31 | 0.34 | 0.32 | 0.49 | Adequate | Improving |
| Net margin | 9.3% | 17.8% | 18.4% | 16.3% | 25.4% | Adequate | Improving |
| Operating cash flow (USD bn) | 0.56 | 0.46 | 0.44 | 0.99 | 0.89 | Strong | Growing |
| Free cash flow (USD bn) | 0.55 | 0.46 | 0.43 | 0.97 | 0.88 | Strong | Growing |
| Total assets (USD bn) | 10.58 | 10.86 | 11.20 | 12.47 | 11.93 | Adequate | Rising |
| Total liabilities (USD bn) | 7.97 | 8.14 | 8.30 | 9.42 | 8.74 | Weak | Mixed |
| Total equity (USD bn) | 2.62 | 2.71 | 2.89 | 3.04 | 3.19 | Weak | Rising |
| Cash (USD bn) | 1.82 | 2.00 | 1.99 | 2.25 | 2.67 | Strong | Rising |
| Goodwill (USD bn) | 4.28 | 4.28 | 4.28 | 4.29 | 4.34 | Weak | Stable |
| Current ratio | 0.65 | 0.76 | Not disclosed | Not disclosed | Not disclosed | Adequate | Stable |
Five standalone quarters are available; earlier periods were reported only on a full-year basis. Cash generation is heavily seasonal, with fiscal fourth quarter operating cash flow of $0.99 billion accounting for about 40% of the year as annual contracts bill. The April 2026 quarter produced $0.89 billion of operating cash flow against $0.56 billion a year earlier, and revenue of $1.93 billion came in above the $1.885 billion to $1.90 billion the company guided in February. Its 25.4% net margin is far above the 15.5% full-year figure and deserves a look at the 10-Q before anyone annualizes it.
Red flags worth checking
- Billings minus revenue. Billings grew 30% to $7.77 billion in fiscal 2026 against 18% revenue growth, which is why cash outran earnings. The company's own fiscal 2027 guidance assumes "temporary risk to billings and revenue" from its sales optimization plan, and the free cash flow guide of $2.7 billion to $2.8 billion implies growth of 12% to 16% versus 54% in fiscal 2026. Watch quarterly billings and current RPO, which grew 23% to $5.48 billion; if the backlog cools faster than revenue, cash follows first.
- Stock-based compensation. It is the largest add-back in operating cash flow, and free cash flow per share of roughly $11 sits near non-GAAP EPS of $10.43 rather than GAAP EPS of $5.23. Track compensation as a share of revenue and the direction of the diluted share count; GAAP EPS rising $0.11 on flat net income implies repurchases slightly outpaced issuance, but that balance can flip.
- Goodwill against thin equity. Goodwill of $4.29 billion exceeds total equity of $3.04 billion and is 34% of assets. Operating income rose from $0.99 billion in fiscal 2023 to $1.79 billion in fiscal 2026 while goodwill grew only about 6% a year, which supports the case that the acquired assets are contributing, but a write-down would land on an already small equity base.
- Tax rate durability. The step-up to roughly 30% of EBIT absorbed most of fiscal 2026's operating improvement. Whether it is a permanent change in the mix or a one-off item determines whether GAAP earnings grow with revenue from here.
Questions worth following
- Can Autodesk keep billings ahead of revenue through the fiscal 2027 sales reset?
Billings lead both revenue and cash at Autodesk, so a stalling billings line would hit free cash flow before it reaches reported revenue. Management's fiscal 2027 guidance already assumes temporary billings risk from the sales optimization plan.
- Can Autodesk's free cash flow hold up net of stock-based compensation?
GAAP operating margin of 22% compares with 38% non-GAAP, and free cash flow per share of about $11 tracks non-GAAP EPS rather than the $5.23 GAAP figure. If buybacks stop offsetting issuance, cash generation is smaller than the headline suggests.
- Can Autodesk earn a return on its $4.3 billion of goodwill?
Goodwill of $4.29 billion now exceeds total equity of $3.04 billion and is 34% of assets, so acquisitions are a large share of capital employed. A write-down would hit GAAP earnings and an already thin equity base.
Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.
Continue reading the full report
- Key risksWhat could break the investment case
- Valuation and margin of safetyWhat the current price assumes, and how much room for error it leaves
- SummaryOur overall judgement and what to watch next
Free to sign up · 20 research points · unlocking this report uses 10
20 points cover 2 full reports. No card needed.
Research reflects its generation date. Preview the first two modules.
More on Autodesk Inc
Autodesk Inc
Business, moat and key questions
Open company page