Analyst: Saul Rosenthal | Date: 2026-04-06 Source: Scout brief (2026-03-25), Q4 FY25 earnings call (2026-02-24)
Let me be straight with you. DigitalOcean at 18% revenue growth, 101% NDR, 59% gross margins, and $1 billion in net debt is NOT a Saul stock today. It doesn't pass my criteria filter. But — and this is important — the trajectory of this business has my attention. Revenue growth is accelerating, not decelerating. The top customer cohorts are growing at crazy rates. And the AI inference opportunity looks genuinely durable. I'm putting this on my watchlist and watching very carefully.
DigitalOcean is a cloud infrastructure platform — what they now call an "agentic inference cloud" — serving small-to-mid-market companies, particularly AI-native and cloud-native businesses. Think of it as the anti-hyperscaler: simpler, cheaper, more predictable pricing than AWS/Azure/GCP, increasingly focused on AI inference workloads. They hit $1 billion in annualized revenue in December 2025.
| Quarter | Revenue | YoY% | QoQ $ | GM% | EBITDA M% | FCF M% | NDR | ARR |
|---|---|---|---|---|---|---|---|---|
| Q1 FY24 | $184.7m | 11.8% | +$3.8m | 59.1% | 40.2% | 18.6% | 97% | $739m |
| Q2 FY24 | $192.5m | 13.3% | +$7.8m | 61.0% | 42.4% | 19.4% | 97% | $781m |
| Q3 FY24 | $198.5m | 12.1% | +$6.0m | 60.2% | 43.7% | 13.2% | 97% | $798m |
| Q4 FY24 | $204.9m | 13.3% | +$6.4m | 61.5% | 41.9% | 17.9% | 99% | $820m |
| Q1 FY25 | $210.7m | 14.1% | +$5.8m | 61.4% | 41.0% | 16.3% | 100% | $843m |
| Q2 FY25 | $218.7m | 13.6% | +$8.0m | 59.9% | 40.9% | 26.1% | 99% | $875m |
| Q3 FY25 | $229.6m | 15.7% | +$10.9m | 59.6% | 43.5% | 37.0% | 99% | $919m |
| Q4 FY25 | $242.4m | 18.3% | +$12.8m | 58.7% | 41.0% | 11.1% | 101% | $970m |
FY25 Total: $901.4m (+15.5% YoY) | Adj EBITDA: $375m (42%) | Adj FCF: $168m (19%)
Trajectory: ACCELERATING. Revenue growth went from ~12% to 18.3% over 8 quarters. Sequential dollar adds accelerated from $3.8m to $12.8m. That is the trajectory I love to see.
| Cohort | ARR | YoY Growth | NDR | Churn |
|---|---|---|---|---|
| Total | $970m | +18% | 101% | -- |
| DNE (>$500/mo) | $640m | +30% | 102% | -- |
| $100K+ Customers (635) | 28% of rev | +58% rev growth | 102% | -- |
| $500K+ Customers | 17% of rev | +97% rev growth | 106% | -- |
| $1M+ Customers | $133m ARR | +123% rev growth | 115% | ZERO |
| AI Customers | $120m ARR | +150% YoY | -- | -- |
Now THOSE are numbers that get my blood pumping! The top of the funnel is growing at triple digits! $1M+ customers at $133m ARR, growing 123%, with ZERO churn! AI customer ARR at $120m growing 150%! And 70% of that AI revenue is from inference + core cloud services, NOT bare metal GPU rental. That's higher-quality, stickier revenue.
Record incremental organic ARR of $51m in Q4. Trailing twelve months at $150m, surpassing even their COVID peak. The RPO jumped to $134m, up 500% year-over-year. That's visibility I can get behind.
Revenue acceleration is real and building. 12% to 18%, with a credible roadmap to 25%+ exit rate in Q4 2026 and 30% for FY2027. This is not a company I'm watching decelerate and hoping for a miracle. Growth is clearly picking up.
The "growth engine within the growth engine" — top customer cohorts. The $1M+ customer cohort at 123% growth with zero churn is remarkable. These customers are scaling WITH DigitalOcean, not graduating to hyperscalers. That was the old knock on DOCN. They've debunked it.
AI inference is a durable, secular tailwind. Inference is different from training — it's driven by real end-customer usage, not VC-funded GPU hoarding. Post-product-market-fit AI companies need inference infrastructure that scales. The CEO's point about this being "real-world workloads" resonates with me.
70% of AI revenue from non-bare-metal. This is crucial! They're not just renting GPUs. They're providing a full-stack platform (databases, storage, networking, inference) that creates real stickiness. This differentiates them from the 31 other "Neo Clouds" that Goldman's analyst rightly pointed to.
Demand exceeds supply. Management said demand "far exceeds" current capacity. That means pricing power and high utilization as new capacity comes online. 31 megawatts ramping through 2026 is pre-sold.
Management credibility. They set 18-20% growth as a 2027 target at their April 2025 Investor Day. They pulled it forward to 2026. They hit the low end (18%) in Q4 2025 — two full years early. Now they're guiding 21% for 2026 and 30% for 2027. This team under-promises and over-delivers.
SBC discipline. Down from 12% to 9% of revenue, well managed. They've repurchased $1.6 billion in shares since IPO.
$22m ARR per megawatt vs. $9-12m for Neo Clouds. Superior economics from the full-stack approach.
Revenue growth is STILL only 18%. I typically want 30-40%+ for my portfolio. Yes, it's accelerating, but the 25-30% is still a promise, not a delivered result. I've learned the hard way that "next quarter will be better" is the most dangerous phrase in investing.
NDR at 101% is weak. Even recovering from 96%, this is nowhere near the 120%+ I look for. The best cohort ($1M+) is only 115%. This tells me the small-customer base (72% of revenue) is still churning or flat. The company depends on top-customer expansion and new customer acquisition to drive growth, not broad-based expansion. That's less durable.
Gross margins at 59% are below my floor. I want 70%+ for SaaS. DigitalOcean is infrastructure, not SaaS, so the comp set is different. But 59% gross margins mean less operating leverage. And management warned that AI mix shift will keep gross margins under pressure — "AI margins are lower than core cloud margins." The mix headwind is structural, not temporary.
The balance sheet gives me serious pause. $1.04 billion in net debt. 3.2x net leverage going to 4x+. $325 million in convertible notes due December 2026 that need to be refinanced or repaid. $130m in finance leases. This is NOT a clean balance sheet. For a $900m revenue company to carry over $1.2 billion in total debt is aggressive. I don't like debt. That's just the way it is.
Execution risk on capacity ramp. The entire acceleration story depends on bringing 31 MW online across 3 facilities during 2026. Supply chain delays, implementation timing, utilization ramp — any hiccup pushes the growth trajectory to the right. The CFO acknowledged "supply chain and implementation timing risks."
Competitive landscape. 32 companies in the SemiAnalysis inference benchmarking report. Hyperscalers (AWS, Azure, GCP) are also investing massively in inference. The "agentic inference cloud" branding is slick, but I want to see durable competitive advantages proven over time, not just marketing positioning.
Margin pressure in 2026. Adj EBITDA guided to 36-38%, down from 42% in FY25. FCF margin guided to 15-17%, down from 19%. Non-GAAP EPS guided to 0.75−1.00, down from $2.12 in FY25. They're investing for growth, which I understand, but this creates a "show me" period where growth must materialize to justify the margin compression.
Management was confident, detailed, and aggressive with forward targets. CEO Paddy Srinivasan clearly articulates the inference opportunity and their differentiation. The CFO (Matt Steinfort) was precise on the financial mechanics of the capacity ramp and capital structure. I was impressed by several things:
The analyst questions were probing (Goldman on competitive moat, BofA on visibility, Morgan Stanley on capacity timing), and management handled them well.
| Metric | Q1 FY26 | FY26 | FY27 (targets) |
|---|---|---|---|
| Revenue | $249-250m | $1,075-1,105m | ~30% growth |
| Revenue Growth | ~18-19% | ~21% | ~30% |
| Adj EBITDA Margin | 36-37% | 36-38% | -- |
| Adj FCF Margin | -- | 15-17% | 20%+ (unlevered) |
| Non-GAAP EPS | $0.22-0.27 | $0.75-1.00 | -- |
| Q4 Exit Rate | -- | 25%+ | -- |
| Rule of 50 | -- | -- | Weighted Rule of 50+ |
The quarterly cadence matters: Q1-Q2 at ~18-19% growth, Q3 ramp, Q4 exit at 25%+. If Q3 FY26 doesn't show meaningful acceleration (call it 22%+), the 25%+ Q4 exit becomes very hard to achieve.
This is my first analysis of DOCN, so there's no prior thesis to update. My priors going in were: "DigitalOcean is a stagnating small-cloud provider losing relevance."
Updated view: DOCN is a credible AI infrastructure pivot story with genuinely exciting top-customer dynamics (123% growth at the $1M+ tier, 150% AI ARR growth, zero churn at the top). The acceleration is real and building. BUT — 18% growth, 59% margins, 101% NDR, and $1B+ net debt keep this off my buy list today. The 2026 capacity ramp is the make-or-break execution test.
Watchlist. No position today. This doesn't meet my criteria for a buy. But I'm watching it more closely than most things on my watchlist, because the trajectory is right and the AI tailwind is real. If they deliver Q3 FY26 at 22%+ growth with NDR ticking up and margins holding, I'd consider starting a 3-5% trial position.
Best, Saul
As of early April 2026, DOCN trades around 85 − 88/share, marketcap 8.3-9.4B, enterprise value ~$9.1B.
| Metric | Value |
|---|---|
| EV/Revenue (TTM) | ~10.1x |
| Forward P/S (FY26E ~$1.09B) | ~7.3x |
| Forward P/S (FY27E ~$1.42B) | ~5.6x |
| P/E (TTM) | ~24x (vs. peer avg ~56x) |
| FY26 EV/EBITDA | ~13.2x |
The stock is up ~145% YTD in 2026 and trading above the analyst median target of $76.50 (range $57-105). At 10x TTM revenue for an 18% grower, the market is clearly pricing in the acceleration to 25-30%. If they deliver, the forward multiples (~5.6x FY27 revenue) look very reasonable. If they don't deliver the ramp, this is expensive. The valuation reinforces my "show me" stance — the stock already prices in success. I need to see it happen before buying. I'd rather pay up for confirmed acceleration than speculate on promised capacity ramps.