From a dorm room to a Ubisoft acquisition

In 2002, an 18-year-old in the Netherlands started renting game servers to his friends from his dorm room. No pitch deck, no investors, no plan beyond fixing laggy games. Sixteen years later, Ubisoft — one of the world's largest game publishers — acquired the company that student built, i3D.net, along with its data-centre arm Smartdc. The deal terms were never disclosed. The playbook that got there is worth more than the number would have been anyway.

The Dorm-room Origin: Why Starting Small Was a Structural Advantage

Starting from a genuine user problem — laggy servers ruining games with friends — gave Stijn Koster something no funded competitor could buy: deep, authentic knowledge of what gamers actually needed. He was not theorising about latency. He was living it.

This matters more than it sounds. The game hosting market in 2002 was fragmented and immature. There was no AWS GameLift, no Azure PlayFab. Publishers were either hosting badly themselves or ignoring the problem. A student who understood the pain at a visceral level, and was willing to fix it for free at first, had a natural feedback loop that a VC-backed team in a glass-walled office simply could not replicate.

The early community traction — friends telling friends, then gaming communities noticing — also meant i3D.net grew into product-market fit rather than spending to acquire it. When EA Games came knocking, Koster already had proof that his servers worked. The dorm room was not a liability. It was the lab.

This is the same pattern you see in other infrastructure businesses that lasted: the founder was the first and most demanding customer. Compare this to how Konnect Insights bootstrapped to ₹85 crore without VC money — the founder's own operational frustration with existing tools was the product brief.

Owning the Stack: Why I3D.net Built Its Own Backbone Instead of Renting

The single most consequential decision i3D.net made was not to rent infrastructure from someone else. From the beginning, the team understood that performance could only be guaranteed if they controlled every layer: compute, connectivity, and software. They built Smartdc specifically to own their own data centres — not lease space in someone else's.

This is a painful, capital-intensive choice. Most hosting startups take the opposite path: rent capacity from a hyperscaler, mark it up, and call it a product. The margin on that model is thin and the differentiation is near-zero.

Flavio Kappelhoff, Director of Business Development at GPORTAL, articulated exactly why owned infrastructure wins on margin:

Owning the backbone meant i3D.net could genuinely promise low latency, because they controlled the routing end to end. A reseller of AWS capacity cannot make that promise and keep it — the moment traffic leaves their own network, they're at the mercy of someone else's peering agreements.

The Pivot That Mattered: From Gamers to Game Publishers

The company's second major inflection point was recognising that the real value in game hosting was not selling to individual players — it was selling to the publishers building the games those players were playing.

This is a classic B2B pivot, and it is where most infrastructure founders hesitate. Consumer revenue feels tangible: thousands of small paying customers, visible growth, community energy. But the unit economics of serving individual gamers are brutal — high churn, high support costs, thin margins per seat.

Publishers are the opposite. One contract with EA Games is worth thousands of individual server rentals. The sales cycle is longer, but the contract value, retention, and margin profile are incomparably better. i3D.net made this shift early, working with EA, THQ and Psyonix among others, and it is the primary reason the company could sustain over a decade of infrastructure investment without external capital.

Raphael Stange, CEO of Nitrado — one of the market's most active consolidators — describes what separates publishers who succeed from those who don't at launch:

This is precisely the gap i3D.net filled. Publishers needed a partner who understood gaming infrastructure at a technical depth that generic cloud providers didn't have. By the time AWS entered the market with gaming-focused offerings, i3D.net had over a decade of publisher relationships and operational muscle that no hyperscaler could acquire quickly.

The B2B pivot playbook — moving from fragmented consumer revenue to high-value enterprise contracts — is one of the most reliable levers in infrastructure. It is also explored in detail in how a logistics SaaS tripled revenue and hit its first profit by pivoting to B2B.

Early Entry: How 2002 Became a Moat by 2018

AWS and Microsoft Azure hold significant share of the game hosting market today, but in 2002, neither of those companies was in the game. The market was wide open, and the only way to build in it was to build slowly, with your own money, solving real problems for real users.

Early entry gave i3D.net three things that money cannot buy retroactively: publisher relationships forged before publishers had alternatives, operational knowledge of gaming-specific infrastructure that generic cloud teams never had to develop, and a reputation in the gaming community that predated the internet's review-culture era. By the time AWS arrived with a funded competitor product, i3D.net had over a decade of trust embedded in the market.

This is why bootstrapping a startup in a niche technical market can produce durable competitive advantage. The patience required to grow without external capital forces a discipline — serve customers profitably, reinvest carefully, own the assets — that funded competitors often skip in favour of growth-at-all-costs.

The Ubisoft Acquisition: What an Undisclosed Deal Still Signals

On November 30, 2018, Ubisoft announced it was acquiring i3D.net and Smartdc. The deal terms were not disclosed — but the strategic logic was clear: Ubisoft, one of the world's largest game publishers, needed a hosting partner it could trust for its own titles, and acquiring one with nearly two decades of publisher-grade infrastructure was faster and more reliable than building from scratch. Koster went on to become VP of Hosting at Ubisoft, leading both i3D.net and Ubisoft's own hosting services, before later moving into the role of chairman of i3D.net's supervisory board.

This is a pattern worth noting. The acquirer was not a private equity firm looking for a roll-up. It was an end-customer who had decided that owning the infrastructure layer was strategically important. That distinction matters for how the acquisition played out.

Max Podkidkin, CEO of BisectHosting — one of the few remaining large independent game hosting companies — has watched the consolidation wave closely:

The i3D.net exit to a strategic buyer rather than private equity is exactly the kind of outcome Podkidkin's logic implies is cleaner. Ubisoft wanted the capability, not a financial return. Koster had built something with genuine operational depth — not a financial engineering story — and that's what attracted a buyer willing to fold the whole team and infrastructure into its own hosting operation rather than strip it for parts.

Bootstrapping a Startup in 2026: What the I3D.net Playbook Still Teaches

The i3D.net story is not a nostalgia piece. The lessons are directly applicable to anyone bootstrapping a startup in a technical niche today.

First: start with a problem you personally feel. Koster was not researching market gaps. He was fixing his own gaming experience. That authenticity produced a product real users wanted before anyone asked them.

Second: own the assets that create your margin. Renting infrastructure from a hyperscaler is a shortcut that caps your upside. If your differentiation is performance, you eventually have to control the thing that delivers performance.

Third: pivot to B2B before you need to. Consumer traction is validation, not a business model. The moment you have publisher-grade proof that your infrastructure works, the B2B conversation becomes much easier — and the unit economics become dramatically better.

Fourth: early entry in a growing market compounds. Being a decade early, with owned infrastructure and publisher relationships, means you capture the compounding — not just the current year's growth.

The same patient-capital discipline that built i3D.net also underpins stories like Yes Madam, which stayed profitable for 9 years before raising ₹50 crore — proof that the bootstrapping playbook still works across industries.

Competing Against AWS: How Independent Operators Survive

AWS and Microsoft Azure hold significant market share in the global game hosting space, with effectively unlimited capital, global data centre footprints, and deep integrations with every major game engine.

How does an independent operator compete against that? The i3D.net answer — and the answer that GPORTAL, BisectHosting, and Nitrado are all executing variations of — is specialisation and ownership.

Generic cloud is optimised for general workloads. Game hosting has specific requirements: sub-20ms latency for competitive play, burst capacity for launch-day traffic spikes, DDoS protection tuned for gaming attack patterns, and anti-cheat infrastructure that sits close to the server. AWS can do all of these things, but a specialist who has done only these things for two decades does them better, faster, and often cheaper — because there's no hyperscaler markup sitting in their own cost base.

Is the Passion-to-business Path Repeatable in Infrastructure?

It's worth asking whether the i3D.net path is repeatable, or whether 2002 was a specific window that has since closed.

The honest answer is: the window for building a consumer game hosting company from a dorm room has largely closed. AWS and Azure have commoditised that layer. A student renting game servers today would be competing on price against hyperscalers with effectively zero marginal cost.

What has not closed is the window for deep specialisation within gaming infrastructure: anti-cheat, real-time communications, edge orchestration for mobile games, low-latency infrastructure for emerging markets. These are the 2026 equivalents of what Koster saw in 2002 — problems that generic cloud providers have not solved well, where a founder with genuine domain knowledge can build something defensible.

The passion-to-business path is not dead. It just requires finding the layer of the stack that the hyperscalers have not yet commoditised.

Conclusion

Stijn Koster's i3D.net is not an inspirational story about luck or timing. It is a precise playbook: start with a problem you personally feel, own the assets that create your margin, pivot to B2B before the consumer model caps your growth, and enter a growing market early enough to compound. Sixteen years of those decisions, made consistently, is what an end-customer like Ubisoft eventually paid an undisclosed sum to acquire.