DigitalOcean Holdings, Inc. Q2 FY2025 earnings call

Product Innovation - Core Cloud: DigitalOcean officially announced the Atlanta data center, which is purpose-built for high-density GPU infrastructure optimized for AI inferencing. It also enabled NFS for GPUs and introduced advanced networking features like Bring Your Own IP address (BYOIP) and Network Address Translation gateways (NAT gateways). - AI/ML Platform: The Gradient AI Agentic Cloud consists of three components. The Gradient AI Infrastructure expanded its GPU Droplets lineup, collaborated with AMD, and introduced an inference-optimized GPU Droplet. The Gradient AI Platform reached general availability, offering an end-to-end agent development life cycle. The Gradient AI Agents include the first commercial agent, Cloudways Copilot. - Go-to-Market: There was strong progress in product-led growth with core cloud customers' first 12-month revenue outpacing prior years. Direct sales and ecosystem partnerships were driving more AI native customers, and there was an increase in the remaining performance obligation (RPO) from large multiyear deals.

In the second quarter, DigitalOcean achieved revenue of $219 million, marking a 14% year-over-year growth. The AI/ML business saw revenue growth exceeding 100% year-over-year. Revenue from Scalers+ customers (annual run rate > $100,000) grew 35% year-over-year and accounted for 24% of total revenue. The incremental ARR in the second quarter was $32 million, the highest since Q4 2022 and the highest organic incremental ARR in over 3 years. Gross margin for the second quarter was 60%, adjusted EBITDA was $89 million with a margin of 41%, and adjusted free cash flow was $57 million, which was 26% of revenue.

Third Quarter 2025: Revenue is expected to be in the range of $226 million to $227 million. Adjusted EBITDA margins are projected to be between 39% and 40%. Non-GAAP diluted EPS is expected to be $0.45 to $0.50. ### Full Year 2025: Revenue guidance has been raised to $888 million to $892 million. Adjusted EBITDA margin guide is 39% to 40%. Non-GAAP diluted EPS is expected to be $2.05 to $2.10. The adjusted free cash flow guide has been raised to 17% to 19% of revenue.

No specific risks were detailed in the transcript beyond general disclaimers about forward-looking statements and reference to SEC filings.

Q: Could you talk a bit more about the AI/ML revenue and the over 100% increase?

A: AI/ML revenue grew over 100% year-over-year. The Gradient AI stack, consisting of infrastructure, platform, and agents, is driving growth. Infrastructure is used by AI-native companies, the platform allows SaaS providers to integrate AI, and agents automate tasks.

Q: Can you explain what acted against the NDR being higher?

A: NDR was 99%, up from 97% last year. There was a mix of customers, with some being cautious and others accelerating. The lapping of the Cloudways price increase wasn't a major factor as market dynamics caused NDR to fluctuate.

Q: What is the industry's state in terms of capacity constraints?

A: Capacity constraints are a factor in AI scaling, but DigitalOcean differentiates with its twin stack cloud (core cloud and AI agentic cloud), which is a key advantage.

Q: Can you give a breakdown of the business between AI and non-AI?

A: AI is an increasingly material part of revenue but remains small. Core cloud is healthy and growing, with AI complementing core cloud and driving growth.

Q: How should we expect large deals in the future?

A: Large deals are a new area for DigitalOcean. They are initially lumpy and spiky as both the company and customers learn, but early successes are encouraging. Guidance is conservative in forecasting large deals until visibility improves.

Q: How should we think about gross margin for the back half of the year?

A: Gross margins are expected to be relatively consistent over the balance of the year. AI is a small part of the business, so no significant impact. Next year, a modest headwind is expected, but the core business remains high margin.

Q: Are there higher volatility or churn in the AI customer base?

A: AI customer acquisition is a different motion. Early-stage start-ups are evolving, but inferencing customers are staying, expanding, and leveraging different parts of the stack with less test-and-leave phenomenon.