DigitalOcean Holdings, Inc. Q3 FY2026 earnings call

• Accelerating Growth: Q3 revenue grew 40% year-over-year to $321 million, beating guidance by over $14 million. This marks the fifth consecutive quarter of acceleration and the highest growth rate in a decade.
• AI-Native Cloud Adoption: Token volume grew roughly 10x quarter-over-quarter. The company launched managed agents, agent learning services, and agent data services, creating a complete platform for agentic AI. 90% of AI customer revenue comes from users of core cloud primitives, indicating deep platform adoption rather than single-product usage.
• Agent Economy: Agents are becoming direct customers; roughly half of unique visitors to DigitalOcean.com are now agents, with over 37,000 unique customers deploying via agents this year. Agents autonomously select DigitalOcean due to its ability to read, reason, and deploy without human intervention.
• Yield per Megawatt: Growth is driven by yield per megawatt ($18 million ARR per average active megawatt) rather than just capacity additions. Software differentiation allows DigitalOcean to capture higher prices and utilization across CPU and GPU fleets.
• Data Center Execution: All three planned 2026 data centers (Richmond, Kansas City, Memphis) launched ahead of schedule. The company secured an additional 20 megawatts for 2028, bringing total committed capacity to ~180 megawatts.
• Differentiated Profile: DigitalOcean maintains minimal customer concentration (top 25 customers = 27% of revenue) and low leverage (net leverage of 1.2x). The balance sheet includes $845 million in cash and a $785 million equipment lease facility.

The transcript does not provide a breakdown of financial performance by specific product segments (e.g., Core Cloud vs. AI Services) in terms of absolute revenue or percentage contribution for each distinct line item. However, it provides aggregate metrics: Total Revenue was $321 million. AI Customer ARR reached $376 million (up 288% YoY), while total ARR was approximately $1.28 billion. It is noted that 90% of AI customer revenue came from customers also consuming core cloud primitives, and customers on four or more products drove roughly 60% of total revenue.

• Q4 2026 Revenue: Raised to $348–$351 million, representing 44–45% year-over-year growth.
• Full Year 2026 Revenue: Raised to $1.208–$1.211 billion, representing ~34% year-over-year growth. Adjusted EBITDA margin expected at 41–42%.
• Full Year 2027 Revenue Growth: Raised significantly from >50% to 60% or more year-over-year growth, off a materially higher 2026 base.
• Q4 2026 Margins: Adjusted EBITDA margin expected at 42–43%; Non-GAAP diluted EPS expected at $0.30–$0.32.

• Supply Chain and Data Center Delays: While execution has been strong, reliance on Tier 1 data center operators introduces dependency on external partners for timely capacity delivery.
• Market Normalization: Potential normalization of spot prices as industry supply increases could impact margins if not offset by platform stickiness and yield improvements.
• Customer Concentration: Although currently minimal, the risk exists if large-scale buyers shift strategies or if funding constraints affect AI startups' ability to spend.
• Technological Competition: Rapid innovation in model routing and agentic platforms requires continuous investment to maintain the 'three to four months' lead over competitors.

Q: Analyst asked how DigitalOcean differentiates its inference engine and agentic platform in customer conversations, given many providers claim similar capabilities.

A: CEO Paddy Srinivasan explained that customers are moving beyond simple token consumption to production agentic workloads requiring secure runtimes, tools, and memory. DigitalOcean offers Agent Gateways with 20,000+ tools, context primitives, and high-performance storage. Use cases are proliferating into media production and personal agents, with Amplitude cited as a key example building self-improving products on the full stack, valuing the lower total cost of ownership compared to frontier models.

Q: Analyst questioned the disparity between 176% growth in Core Cloud attached to AI versus slower growth in non-AI core cloud, asking if traditional customers are adopting AI.

A: CFO Matt Steinfort noted it is becoming harder to distinguish AI from non-AI customers as traditional core cloud users deploy agentic frameworks like OpenClaw or Hermes. These customers use AI to drive business but may not declare themselves as 'AI customers.' Simultaneously, the core cloud business remains a stable, highly profitable $900 million ARR base generating significant cash flow, accelerating alongside new AI-oriented products.

Q: Analyst asked about the implications for DigitalOcean's model when industry supply normalizes and spot prices potentially decline.

A: CEO Srinivasan stated the company is favorably positioned because growth is driven by yield per megawatt and platform attach, not just capacity. As supply-demand equilibrates, the explosion of agentic use cases (e.g., physical intelligence companies) will accelerate demand. Since 90% of AI revenue comes from multi-product users and only a small fraction are GPU-only customers, the full-platform strategy protects against price compression on single features.

Q: Analyst asked for details on the $18 million ARR per megawatt metric, specifically the mix of CPU vs. GPU and future trajectory.

A: CFO Steinfort clarified this is a fleet-wide number including all products and technologies. He noted that newer GPU generations (GB300/Vera Rubin) will likely have higher CapEx per megawatt (~$35-40M) but deliver much higher yields. CEO Srinivasan added that as customers expand from tokens/GPUs to higher-margin agentic runtimes and data services, the overall yield per megawatt is expected to increase further.

Q: Analyst asked about financing needs for 2027 and why equity (ATM) was chosen over further debt leases.

A: CFO Steinfort emphasized the balance sheet is a competitive advantage, citing $850M+ in cash and flexible credit facilities. The $1B ATM provides option value rather than immediate necessity, allowing the company to fund growth opportunistically. The primary strategy remains matching equipment costs with revenue via asset-backed financing, but the equity line ensures no time-boxed constraints during a noisy market environment.