The IGV software ETF just entered a bear market. ServiceNow is down 50% from its highs. Salesforce and Workday have been crushed. And somehow, the little developer cloud that everyone wrote off is one of the only winners left standing.
Let’s set the scene.
It is February 2026. The “SaaSpocalypse” has officially arrived — as everyone now knows. On February 3rd, Anthropic launched Claude Cowork — a legal automation tool — and in a single day, roughly $285 billion in software market cap evaporated. Thomson Reuters dropped 16%. ServiceNow, Salesforce, HubSpot, Atlassian, Docusign, Asana, Workday, and Adobe all cratered in unison. Traders at Jefferies called it pure “get me out” selling.
The iShares Expanded Tech-Software ETF (IGV) — the sector’s primary benchmark — is down roughly 22-25% year-to-date. Software price-to-sales ratios have compressed from 9x to 6x, levels not seen since the mid-2010s. Nearly $1 trillion in enterprise software market value has been wiped out since mid-2025.
ServiceNow — which beat earnings for the ninth straight quarter — is still down 50% from its highs. Salesforce is down 40% over the past year. Adobe is down 35%. Workday is down 40%+. It’s rough out there.
And in the middle of all of this: DigitalOcean (DOCN) is sitting at $54.62. Up 11.5% year-to-date. Up 25% over the past year.
With a $5 billion market cap, 18% revenue growth accelerating to 21%, 42% EBITDA margins, and a path to 30% growth in 2027.
How? How is the little developer cloud that traded at $19.39 two years ago one of the only winners in the worst software selloff since COVID?
Let’s break it down.
The Carnage: A Quick Snapshot of the 2026 SaaSpocalypse
Here’s what the public cloud and SaaS landscape looks like right now:
The Wreckage (down significantly YTD or from 52-week highs):
- ServiceNow (NOW): Down ~34% YTD, down ~50% from 52-week high. Still growing 21%. P/E compressed to 24x. CEO buying stock personally. Stock around $104.
- Adobe (ADBE): Down ~35% over the past year. P/E compressed from 26x to 16x. Pivoting to “Generative Credit” pricing. At multi-year lows.
- Workday (WDAY): Down 40%+ from highs. Under intense pressure to prove AI doesn’t kill seat-based pricing.
- HubSpot (HUBS): Was down 51% in 2025. SMB CRM and marketing automation under AI pressure.
- Atlassian, Docusign, Asana, Monday.com: All caught in the indiscriminate selling.
- IGV (Software ETF overall): Down ~22-25% YTD in 2026. Entered a technical bear market.
The Survivors (flat to up) — It’s A Small Group 🙂
- DigitalOcean (DOCN): +11.5% YTD. +25% over the past year. $5B market cap.
- Oracle (ORCL): Up modestly, benefiting from AI infrastructure positioning. Though it’s down 45% from its September 2025 peak.
The Common Thread of the Carnage: Seat-Based SaaS Is Getting Repriced
The market’s fear is straightforward and, honestly, not irrational. If AI agents can do the work of 10 customer service reps, you don’t need 10 Salesforce seats. You need one. The per-seat business model — the foundation of the entire SaaS era — is being repriced in real time.
As Jason Lemkin wrote on SaaStr in January: the SaaSpocalypse isn’t AI killing SaaS. It’s the market finally pricing in the deceleration that started in 2021. Public SaaS growth rates have declined every single quarter since the 2021 peak. AI just gave the market “permission” to re-rate what the numbers had been screaming for three years.
So Why Is DigitalOcean the Outlier?
DigitalOcean just reported Q4 2025 earnings two days ago. The numbers were strong: $242 million in Q4 revenue (up 18%), $901 million for the full year, a record $51 million in incremental organic ARR in the quarter, and they crossed $1 billion in annualized monthly revenue in December. And importantly, it’s accelerating.
But the stock performance isn’t just about one good quarter. It’s about DigitalOcean being structurally positioned on the right side of the AI divide — while the rest of enterprise software is on the wrong side.
Here’s why:
1. DigitalOcean Sells Infrastructure, Not Seats
This is the fundamental difference. Salesforce, ServiceNow, HubSpot, Workday — they sell per-seat subscriptions to humans using software. When AI reduces the number of humans who need to use software, their revenue goes down.
DigitalOcean sells cloud infrastructure — compute, storage, networking, databases. When AI increases, more infrastructure gets consumed, not less. Every AI agent needs compute. Every inference call needs a server. Every AI-native startup needs a cloud provider.
The SaaStr “Great B2B Bifurcation” analysis from late 2025 identified this exact pattern: the 2025 winners were infrastructure plays (Palantir, MongoDB, Cloudflare, Snowflake) rather than application plays (HubSpot, Monday.com, Asana). Infrastructure benefits from AI. Applications get disrupted by it.
DigitalOcean is squarely in the infrastructure camp.
2. Consumption-Based Revenue, Not Seat-Based Revenue
DigitalOcean charges based on usage. More compute, more storage, more data transfer = more revenue. This is the same model that’s working for MongoDB (up 70% in 2025), Snowflake, and Cloudflare.
AI increases data volume. AI increases query complexity. AI increases compute demand. All of this flows directly into DigitalOcean’s revenue.
Seat-based models fight against AI growth. Consumption-based models ride it.
3. AI-Native Customers Are DigitalOcean’s Fastest Growing Segment
DigitalOcean’s AI customer ARR hit $120 million in Q4, growing 150% year-over-year and now representing 12% of total ARR. And critically, 70% of that AI revenue comes from inference services and general-purpose cloud, not just bare metal GPU rentals. These customers are buying the full stack.
The company is positioning as the “Agentic Inference Cloud” — purpose-built for companies running AI in production. And unlike the neo-cloud GPU rental shops (which have 70-80% customer concentration in their top 25 accounts), DigitalOcean’s top 25 customers represent just 10% of revenue. That’s a fundamentally more durable business.
4. The “Customer Graduation” Problem Is Solved
For years, the bear case on DigitalOcean was that customers start on DOCN and then “graduate” to AWS. The Q4 numbers kill that narrative:
Million-dollar customers are driving $133 million in ARR, up 123% year-over-year. Zero churn among million-dollar accounts over the last twelve months. Net dollar retention for $1M+ customers is 115%.
Digital Native Enterprises (post-product-market-fit companies) now represent 62% of total ARR and are growing at 30% year-over-year.
DigitalOcean didn’t just patch its churn problem — it built a growth engine out of upmarket expansion.
5. Profitable Growth at Scale
In a market where investors are suddenly obsessed with free cash flow and profitability, DigitalOcean delivers:
Full year 2025 adjusted EBITDA: $375 million at 42% margin. Net income: $259 million, up 207% YoY. Free cash flow: $168 million at 19% margin. Stock-based comp reduced from 12% to 9% of revenue (80th percentile of software peers). $254 million cash on hand, $300 million undrawn revolver, no material debt maturities until 2030.
Compare that to many SaaS companies that are still losing money or generating thin margins while their stocks get crushed. DigitalOcean is growing faster and more profitably than most of the companies in the IGV that are down 25%.
The Comparison That Matters Most
Let me put the DigitalOcean story in context against the broader cloud landscape:
DigitalOcean vs. the SaaSpocalypse Victims:
The pattern is stark. The companies selling software seats to humans are getting repriced downward. The company selling infrastructure that AI consumes is getting repriced upward.
DigitalOcean vs. the Other Infrastructure Survivors:
What about the companies in the “infrastructure benefits from AI” camp? MongoDB was up 70% in 2025. Cloudflare was up 80%. Snowflake was up 45-50%.
DigitalOcean’s advantage here is differentiated: it’s the only infrastructure player specifically targeting the SMB and mid-market AI-native segment at scale. MongoDB serves data. Cloudflare serves the edge. Snowflake serves analytics. DigitalOcean serves the full stack for the next generation of AI-native companies that aren’t big enough for AWS but need more than a GPU rental shop.
And unlike many infrastructure peers, DigitalOcean is already deeply profitable with a clear path to Rule of 50+.
The Forward Story: Why This Might Just Be the Beginning
DigitalOcean’s guidance tells a story of continued acceleration:
- 2026 revenue: $1.075-$1.105 billion (19-23% growth)
- Q4 2026 exit growth rate: 25%+
- 2027 growth target: 30%
- Path to Rule of 50+ (30% growth + 20%+ margins)
At their Investor Day last April, they targeted 18-20% growth by 2027. They hit 18% two full years early. Now they’re guiding to blow past those targets.
The growth is coming from 31 megawatts of new data center capacity ramping through 2026 — which will create some near-term margin pressure but fuel the acceleration into 25%+ exit rates. CEO Paddy Srinivasan explicitly called this a “physics problem” — the costs of new capacity come before the revenue, but the return profile is strong.
At ~5x forward revenue with accelerating growth and 40%+ EBITDA margins, DigitalOcean trades at a meaningful discount to slower-growing, more disruption-vulnerable SaaS peers that still trade at 6-12x revenue despite the selloff.
7 Lessons from DigitalOcean’s Positioning for SaaS Founders and Operators
1. Infrastructure Beats Applications in the Age of AI
If AI reduces the humans who use your software, you lose revenue. If AI increases the infrastructure your customers consume, you gain revenue. Build for consumption, not seats.
2. The “SaaSpocalypse” Is Really a Business Model Reckoning
The companies getting crushed aren’t getting crushed because their products are bad. ServiceNow is still growing 21%. Salesforce just did $41.5B in revenue. The issue is that the market is fundamentally repricing the terminal value of seat-based businesses. If you’re building a seat-based SaaS company right now, you need a credible answer for what happens when AI agents replace some of your end users.
3. Moving Upmarket Is the Highest-ROI Initiative in SaaS
DigitalOcean’s story is fundamentally about going upmarket — from $5/month droplets to million-dollar accounts. The $1M+ customer cohort growing at 123% with zero churn is the engine of the entire re-rating. If you’re stuck at low ACVs, your number one priority should be figuring out how to land and expand larger customers.
4. “We’re Not a GPU Landlord” Is a Strategy
DigitalOcean watched the GPU rental arms race and decided not to play. Instead of competing on raw hardware, they built a full-stack inference cloud. 70% of their AI revenue is from inference and cloud services, not bare metal. Differentiation beats commoditization, even in infrastructure.
5. Profitable Growth Is a Superpower in a Downturn
When markets turn, the companies with cash and margins have all the leverage. DigitalOcean’s 42% EBITDA margins, $254M in cash, and positive free cash flow let them invest aggressively in growth while others retrench. They’ve repurchased $1.6 billion in stock since IPO. In a bear market for SaaS, profitability isn’t boring — it’s survival.
6. NDR Over 100% Changes Everything
DigitalOcean’s NDR crossed back above 100% to 101% overall, and 115% for million-dollar accounts. This was the metric that bears cited for years as evidence the business was broken. Fixing NDR was the prerequisite for everything else — the re-acceleration, the upmarket motion, the AI positioning. If your NDR is below 100%, that’s your most important metric to fix.
7. Sometimes the Best Position Is Being Where Nobody’s Looking
Nobody was writing “DigitalOcean is the AI cloud play” in 2023. The stock was at $19. The narrative was completely dead. And that’s exactly when Paddy Srinivasan and team started building the positioning that’s now paying off. The best time to build your moat is when nobody believes it’s possible.
The Bottom Line
The 2026 SaaSpocalypse is real. The IGV is in a bear market. A trillion dollars in software market cap has been destroyed. ServiceNow is down 50% from its highs despite beating earnings. Marc Benioff is wearing a leather jacket and saying “SaaSpocalypse” on earnings calls.
And DigitalOcean — the developer cloud that was left for dead at $19 in 2023 — is up 12% YTD, guiding to 21% growth, heading toward 30% growth, trading at 42% EBITDA margins, and quietly building one of the most interesting AI infrastructure businesses in public markets.
It turns out that when AI eats the software world, someone still needs to sell the infrastructure that AI runs on. DigitalOcean figured that out before almost anyone else.
In every market cycle, there’s a company that surprises everyone. In 2026, in the middle of the worst software selloff in years, that company might just be the little developer cloud from New York.
Note: This is not financial advice. I’m not a financial advisor. Do your own research before making any investment decisions.
