DigitalOcean stock surges 12% on strong Q2 outlook

DigitalOcean shares rose 12% after it guided to record Q2 FY2026 results, with remaining performance obligations (RPO) exceeding $800M—up more than 10x YoY. The company expects ~29% Q2 revenue growth (vs. 14% a year ago) and projects RPO to rise by over $550M in Q2 2026, supported by expanding committed data center capacity to ~155MW (adding 20MW for late 2027/early 2028). Management also expects customer momentum to lift previously provided FY2026 exit revenue growth guidance, pointing to a stronger demand backdrop for inference and cloud products.

Analysis

DOCN is signaling that demand is outrunning its current footprint, which matters more than the headline growth rate. The key mechanism is operating leverage: if the new AI/inference commitments are real and durable, incremental revenue should fall through faster because the company is effectively pre-committing scarce capacity into 2027-28. That makes DOCN interesting not as a generic cloud name, but as a cost-sensitive inference platform that could win share as customers optimize away from premium GPU-heavy environments.

The second-order risk is that this is still a capacity-and-contract story, not yet a fully proven usage story. A longer weighted average life improves visibility, but it also means less near-term flexibility if demand proves lumpy; if utilization disappoints, fixed commitments can become a margin overhang rather than a growth lever. The next 1-3 months matter for conversion: billings-to-revenue, gross margin trajectory, and whether management raises exit-2026 guidance in a way that implies sustained demand rather than one-off deal wins.

On the other side, NVDA remains exposed to the market’s fear that custom silicon and inference optimization reduce the long-term mix of high-end GPU spend. That is likely a multiple headwind before it becomes an earnings headwind, because the first-order impact is sentiment around share capture, not an immediate collapse in demand. The contrarian read is that custom chip headlines often overstate near-term substitution; the real beneficiary may be the infrastructure layer that can sell cheaper inference capacity to price-sensitive customers, which is where DOCN has a plausible niche.

I would watch for two falsifiers: if DOCN’s backlog growth does not convert into sustained revenue acceleration by the next two quarters, or if NVDA’s enterprise/hyperscaler capex commentary re-accelerates enough to drown out custom-silicon chatter. Over 6-18 months, DOCN can rerate if management proves this is a durable AI workload franchise, but if growth normalizes before the added capacity comes online, the stock could quickly de-rate back to a low-growth cloud multiple.

Key Decisions for Investors

• Long DOCN on pullbacks, not strength: wait for the post-gap consolidation and use the next 1-2 weeks to build exposure only if management reiterates margin discipline. Upside is a re-rating toward higher-visibility cloud infrastructure names; stop if the next print shows backlog conversion slowing or EBITDA guidance coming down.
• Pair trade: long DOCN / short NVDA into any NVDA bounce on AI headline fatigue. Thesis is that cost-optimized inference demand can expand at the edge while custom-silicon headlines pressure NVDA multiple more than DOCN fundamentals; cover the short if NVDA capex commentary or Blackwell demand re-accelerates.
• Use DOCN calls only if you need convexity: a 3-6 month call spread is preferable to outright stock after a 12% move, because the event is now partially priced and the main risk is a slow conversion of backlog into revenue. Avoid naked calls unless the next earnings date is within the holding window.
• Set a watch item on DOCN’s next quarterly report: if revenue growth and margin both improve while RPO keeps extending, the stock can trend for multiple quarters; if revenue beats but margins weaken, treat it as a capacity-funded growth trap and fade the move.
• For NVDA, stay tactical rather than structural short: use the DeepSeek/custom-chip narrative as a trading overhang on rallies, but do not press a medium-term short unless there is evidence that cloud capex or inference spend is rolling over.