DigitalOcean: Inference Cloud Growth Isn't Fully Reflected in the Market

Hook & thesis

DigitalOcean is no longer just a low-cost developer cloud - it's become a focused inference cloud play for SMBs and developers. Q4 2025 results showed AI-specific ARR hitting $120M (150% year-over-year growth) and company ARR at $970M, while revenue trended to $242M in the quarter. Yet the stock is trading near its 52-week high at $90.55, and the market capitalization still leaves room for upside if DigitalOcean converts faster-growing AI ARR into recurring high-margin revenue.

My thesis: the market has not fully priced a scenario where DigitalOcean captures a disproportionate share of inference workloads for small and mid-sized businesses, captures Heroku migrations, and monetizes high-value customers at scale. That outcome would support materially higher revenue and justify a higher multiple than today's price implies. This is a tradeable long with a clear entry, stop and target and a defined 180 trading-day horizon.

What the company does and why it matters

DigitalOcean provides cloud infrastructure and platform services targeted at developers, startups and SMBs across multiple regions. Historically the company competed on simplicity and price; the recent pivot is toward inference-optimized infrastructure and developer-facing AI services where hyperscalers are heavy on training but less focused on the developer-to-application layer for smaller customers.

The market should care because inference workloads are sticky and higher-margin once customers standardize on an execution layer. DigitalOcean's growth in high-value customers - ARR from >$1M customers grew 123% with reported zero churn - signals early success at landing account expansion dynamics that scale revenue without proportional increases in sales intensity.

Concrete numbers backing the argument

Use the following datapoints to judge momentum:

• Q4 2025 revenue: $242M.
• AI-specific ARR: $120M, up 150% YoY.
• Total ARR: $970M.
• High-value customers (> $1M ARR): growth of 123% with zero churn reported.
• Current price: $90.55; 52-week range: $25.45 - $91.53.
• Market cap (snapshot): $8.28B; enterprise value reported at $10.53B.
• Reported free cash flow (trailing): $53.69M.
• Valuation metrics: P/S ~ 10.4, P/E ~ 36.

Valuation framing

At a market cap near $8.3B and EV of roughly $10.53B, DigitalOcean is priced like a company expecting continued high-growth, software-like monetization. The trailing P/S around 10.4 and P/E near 36 reflect that expectation. Third-party commentary in recent coverage referenced forward P/S estimates of roughly 7.3 for 2026 and 5.6 for 2027, implying the market has already baked in substantial revenue expansion - but perhaps not the full inference monetization runway.

Put simply: if AI ARR continues to grow at triple-digit rates and the company converts a growing share of that ARR into billable inference consumption and platform fees from higher-value customers, revenue could expand materially and compress forward multiples. Conversely, if the market’s concerns around capital intensity (the company is raising capital for capacity) or margin pressure materialize, the current multiple could look expensive.

Trade plan (actionable)

Direction: Long

Entry price: $90.55

Target price: $140.00

Stop loss: $78.00

Horizon: long term (180 trading days) - expect the trade to play out over the next 6 months as capacity buildouts, AI ARR expansion and customer migrations (including Heroku exodus) convert into visible top-line and ARR upgrades. The 180 trading-day window allows time for the company to convert announced capital plans into capacity and for multi-quarter ARR momentum to show through to revenue and margins.

Rationale for levels: Entry is at the current market price. The $140 target assumes the market assigns a higher multiple to materially larger revenue and ARR (think forward P/S falling toward mid-single digits as revenue expands), while $78 sits below the recent 10-day/21-day EMAs and provides a disciplined invalidation point if momentum or fundamentals deteriorate.

Key catalysts

• Ongoing quarterly results showing durable AI ARR growth and improving gross margins on inference workloads.
• Progress on the $800M infrastructure raise and visibility on added MW capacity - incremental capacity should reduce latency and enable new enterprise-sized deals.
• Customer migration flows from Heroku and other PaaS exits. Publicized migrations and case studies would be an adoption accelerator.
• Upsell to existing high-value customers (> $1M ARR) that can materially lift ARR per account and margins.

Risks and counterarguments

Below are the principal risks that could undermine the thesis.

• Capital intensity and margin pressure: The company is raising significant capital to add data center capacity. Heavy near-term capex can depress free cash flow and margins even as revenue grows.
• Competition from hyperscalers: AWS, Google Cloud and Microsoft remain dominant and can undercut pricing or bundle inference services as part of larger enterprise deals, limiting DigitalOcean's share gains.
• Execution risk on enterprise wins: High-value customer growth is promising, but scaling enterprise-grade SLAs, security, and sales motions is hard and expensive. Any slowdown or churn among those accounts would hurt forward ARR expansion.
• Valuation vulnerability: At a P/S above 10 and P/E near 36, the stock is exposed to sentiment and multiple compression if growth slows. A single quarter of disappointing AI growth could trigger a sharp repricing back toward the 52-week median.
• Short-interest and volatility: Short interest has been notable (roughly 9.6M shares reported in the last settlement), and short-volume data shows persistent activity. That raises the potential for volatility in either direction.

Counterargument

One solid counterargument is that DigitalOcean's market is structurally limited relative to hyperscalers; most large enterprise inference demand will route to cloud incumbents who control data, model training, and ecosystem integrations. If inference monetization remains targeted to SMBs and small developers, the TAM may be smaller and margins more constrained than some bulls expect. In that scenario, the current multiple would be hard to justify and downside risk would increase.

What would change my mind

I would downgrade this trade if any of the following occur:

• Quarterly AI ARR growth decelerates materially from the 150% YoY pace reported in Q4 2025.
• The company reports significant churn among > $1M customers or a contraction in ARR per account.
• Margins fail to show improvement after capacity investments complete, or capex plans escalate well above the currently reported raise.

Conclusion

DigitalOcean is a defensible, developer-first inference cloud with early signs of durable AI monetization. The combination of rapid AI ARR growth, expanding high-value customer counts, and an announced infrastructure buildout creates a credible path to materially larger revenue over the next several quarters. That path is not guaranteed - execution and capital intensity are real risks - but the trade offered here buys that upside with a disciplined stop and a 180 trading-day runway to let macro, capacity and ARR signals play out.

Trade idea summary: Long DOCN at $90.55, target $140.00, stop $78.00, horizon long term (180 trading days). Medium risk trade that bets on inference monetization and high-value customer expansion outpacing near-term capex drag.