DigitalOcean's Q3'26 Earnings Conference Call - DIGITALOCEAN HOLDINGS INC

Financial Performance:

• Q3 2026 revenue reached $321 million, a 40% year-over-year (YoY) increase, exceeding the high end of guidance by over $14 million.
• Annual Recurring Revenue (ARR) grew by a record $160 million YoY (up 266%), bringing total ARR to $1.28 billion; AI-specific ARR reached $376 million, up 288% YoY.
• Top 100,000+ ARR accounts grew 133% YoY to represent 41% of total ARR; million-plus accounts grew 273% to represent 30%.
• Adjusted EBITDA was $144 million (45% margin), significantly above the 38-39% guidance; Adjusted Operating Income margin was 23%.
• Non-GAAP EPS was $0.45, beating the $0.28-$0.30 range; GAAP net loss of $43 million included a one-time $57 million induced conversion expense.
• Gross margin was approximately 50%, with a slight decline attributed to new data center ramp-up, though operating leverage remains strong.
• Trailing 12-month adjusted pre-tax cash flow was $220 million (20% margin).
• Balance sheet remains robust with $845 million in cash, equivalents, and marketable securities, plus $400 million available under a revolving facility.
• Net leverage stands at 1.2x; secured a $785 million equipment lease facility and a $1 billion at-the-market equity program for flexibility.
• Customer concentration is minimal, with the top 25 customers representing only 27% of revenue and no single customer exceeding high single digits.

Guidance and Future Outlook:

• Raised Q4 2026 revenue guidance to $348-$351 million (44-45% YoY growth).
• Raised full-year 2026 revenue guidance to $1.208-$1.211 billion (approx. 34% YoY growth), with adjusted EBITDA margin expected at 41-42%.
• Raised 2027 full-year revenue growth outlook from >50% to 60% or more.
• Confidence in 2027 and 2028 growth is driven by software levers, yield per megawatt improvements, and high attach rates rather than just capacity expansion.
• Management expects capacity utilization to ramp quickly, similar to the 2026 trajectory where new facilities became profitable within months.

Business Segments and Product Lines:

• AI-Native Cloud: Token volume grew 10x quarter-over-quarter; 90% of AI revenue comes from customers also consuming core cloud primitives.
• Inference & Agents: Shipped managed agents, agent learning services, and agent data services; over 70 platform releases in 65 days.
• Customer Behavior: 60% of revenue is now generated by customers using four or more products; agents act as both builders and direct buyers, with 37,000 unique customers deploying via agents this year.
• New Offerings: Launched "Agent Droplets" to simplify access to agentic primitives for non-AI customers; 37 model launches since April.
• Data Centers: Richmond, Kansas City, and Memphis facilities launched in 2026 ahead of schedule; secured an additional 20 megawatts for 2028.

Market and Competitive Landscape:

• Market shift identified from human-chatting models to autonomous "agentic" systems that require secure runtimes, memory, and data services.
• Differentiation strategy focuses on a complete platform (GPU, CPU, tokens, agents, data) rather than point solutions or bare metal, reducing reliance on single-feature competition.
• Demand for tier-one capacity significantly outstrips supply; competitors face challenges with gigawatt-scale deployments, while DigitalOcean scales in tens of megawatts.
• DigitalOcean is positioned to capture value from falling token prices and compute equilibrium by driving platform attachment and higher-margin workloads.
• Model routing capabilities are cited as a 3-4 month lead over competitors, customized for specific workloads like coding versus media generation.

Risks and Challenges:

• Supply Chain & Power: Risks associated with securing power and data center capacity remain, though mitigated by relationships with tier-one operators and on-time delivery track record.
• Margin Pressure: Short-term gross margin pressure expected due to depreciation and lease costs on new data centers before revenue ramps fully.
• Technology Cycles: Rapid shifts in model efficiency and hardware generations require continuous capital deployment to maintain yield per megawatt.
• Financing Environment: While the balance sheet is strong, reliance on equipment financing and potential equity dilution remain factors to manage.

Management Commentary and Tone:

• Management expressed high confidence ("never been more confident"), citing a "different by design" strategy with a diversified customer base and profitable core cloud business.
• Tone was aggressive yet disciplined, emphasizing durable, profitable growth over pure capacity expansion.
• Patty Srinivasan (CEO) highlighted the "agentic era" as a tailwind, with agents now autonomously selecting DigitalOcean as a platform.
• Matt Steinforth (CFO) emphasized the balance sheet as a competitive advantage in a tightening debt market, highlighting proactive deleveraging.

Analyst Questions and Answers:

• Differentiation: Management clarified that customer demand is driven by the full agentic platform (runtime, tools, memory), not just token throughput or inference engines.
• Core Cloud vs. AI: Distinguishing between AI and non-AI customers is becoming difficult as "core" workloads increasingly require intelligence; core cloud ARR grew 176% within the AI funnel.
• Supply/Demand Normalization: Management believes falling spot prices will be offset by exploding use cases (e.g., personal agents, physical intelligence) and higher platform attach rates.
• Capacity Cadence: 75MW expected in 2027 and 30MW in 2028; all new capacity will be deployed via tier-one partners, with utilization ramping quickly.
• Financing Strategy: The $1B ATM is an option for liquidity; the primary funding strategy remains equipment financing to match revenue timing, preserving flexibility.
• Margin Outlook: Long-term margins are expected to expand as data centers mature and depreciation costs stabilize; operating leverage will continue to drive profitability.
• CapEx Efficiency: CapEx per megawatt estimated in the mid-$30s to $40M range for new generations, with ROI maintained through higher token density and yield.
• Competition: DigitalOcean maintains a lead in model routing and agentic recursion (self-improvement), with the platform story outweighing single-feature competitors.

Other Key Points:

• 90% of AI customer revenue is derived from customers using the full platform, indicating deep stickiness and cross-selling success.
• "Agent Droplets" introduced to lower barriers for traditional cloud customers to adopt agentic workflows.
• Management noted that agents are increasingly direct buyers, bypassing human decision-making for infrastructure selection.
• The "yield per megawatt" metric is prioritized over raw capacity, with industry-leading ARR per active megawatt of $18 million.
• A one-time GAAP loss was driven entirely by the repurchase of convertible notes, not operational performance.