How Small Industry Labs Can Thrive Through Open Research

Over the past 13 years, SIDN Labs has built collaboration models that are not so common in the world of internet infrastructure: a small, mission-driven tech company working in deep, long-term partnerships with universities. We published them as an ACM SIGCOMM CCR editorial (PDF).

Here, we summarise what we’ve learned from this journey - both the different collaboration models we’ve developed, and what it takes to successfully run an industry research team inside a small organisation.

Research at SIDN: How do you folks do that?

“Wait - you’re at a small industry lab and still manage to publish in top-tier academic venues? And you open-source your work? Plus you use your research results for your production services? How do you folks do that?”

We have been asked that question many times at scientific conferences and elsewhere. It makes sense: small organisations don’t typically have dedicated research groups, and open collaboration with academics is even less common. The questions have motivated us to explain how we have managed to operate as we do for so long.

Research is expensive, especially for a small organisation. To make long-term, open research possible, SIDN structurally allocates 6% of its annual revenue to SIDN Labs as part of its strategy. That dedicated budget allows us to work on research and ensures that research remains a core part of SIDN’s public-interest mission.

Collaboration models: 5 ways to work together

Our experience shows that there is no single way for industry and academia to collaborate successfully. Instead, five models have emerged at SIDN Labs over time, each suited to different situations and levels of maturity.

The most accessible entry point is the internship. These short, three-to-six-month projects allow college students to work with real production datasets, guided by both SIDN Labs staff and university supervisors. Internships are relatively low-cost for the company, offer students invaluable hands-on experience, and lead to master theses and occasionally peer-reviewed papers. Sometimes we also hire a student as part of our staff afterwards.

Another simple model is data sharing, which allows researchers to explore questions they could not otherwise study. OpenINTEL is one of the examples where we share data, sending a copy of the .nl zone file to the University of Twente daily for academic research.

A large share of our work happens through informal collaboration, which is by far the most common model we use. These are lightweight, contract-free partnerships where SIDN Labs researchers and academics jointly explore a topic and produce concrete outputs - for instance, with the goal of comparing phishing between .nl, .be, and .ie (PDF).

Because they have no administrative overhead and rely mainly on shared interest and trust, informal collaborations move quickly and often lead to some of our most influential results. Their flexibility is also their limitation: without formal structure, continuity is not guaranteed, and progress depends heavily on the motivation of the researchers involved.

To address that, SIDN Labs has invested in secondment positions, where the authors of this blog spend one day per week working at 3 Dutch universities (TU Delft, University of Amsterdam and University of Twente). Secondments create a bridge between our operational challenges and academic research agendas, and are a continuous model. They increase visibility among students and help us hone our research agenda.

Finally, we participate in funded research projects - some co-funded by SIDN, others financed through European or national programmes, and some commissioned by organisations such as ICANN or the Dutch government. These projects allow us to work with large consortia and build tools and knowledge that benefit wider communities. Their scale implies more coordination effort, but also provides the opportunity to grow our research network and explore new topics.

Why these collaborations matter for industry

For SIDN as a critical infrastructure operator, academic collaboration brings very practical advantages.

One is the ability to extract more value from the data we already collect but do not always have the capacity to analyse thoroughly. Students and researchers help illuminate this “dark data,” leading to insights that improve our operations.

Another benefit is independent scrutiny. Operators prioritise stability; researchers have the freedom to question assumptions, test alternatives and document unexpected behaviour. That complementary perspective leads to stronger evidence-based decision-making and more robust services.

And, because SIDN has a public-interest mission, the openness of our research matters. Publishing results, open-sourcing tools and contributing to RFCs ensure that our findings benefit not only SIDN, but also the wider internet community.

What academia gains from collaborating with SIDN

Academic partners benefit in multiple ways. They obtain access to real-world datasets that are rarely available in university environments. They can work on operationally relevant problems and test their ideas in an operational setting. And they gain direct contact with domain experts who have spent years operating .nl - expertise that is difficult to acquire through textbooks.

Such collaborations also offer researchers visibility beyond the academic world, with projects often presented at RIPE, IETF and ICANN meetings. The result is research that is both scientifically valuable and practically impactful.

Building a research team inside a small tech company

From SIDN’s perspective, building a research team has been an evolutionary process. We began by sponsoring PhD positions to create initial ties with universities. As the team grew, we developed our own research infrastructure, and we expanded internal expertise by increasing the size of our team. In parallel, we diversified collaboration models - from internships and informal projects to data-sharing. The final stage of maturation involves consolidating in-house expertise while maintaining strong academic links through secondments and continuous collaboration.

That phased approach has allowed us to maintain long-term research capacity while staying small, agile and closely connected to operational teams.

Looking ahead

Our experience demonstrates that leading, impactful, open research does not require the scale or resources of a global tech giant. With the right combination of collaboration models - each serving a different purpose - small public-interest organisations like SIDN can meaningfully contribute to the resilience and evolution of the Internet. We also hope that sharing our approach motivates other organisations to adopt similar models and even establish their own research teams.

For readers who want to dive deeper into the details, our full paper is available here (PDF).

This article was originally published over on the SIDN Labs blog. Note that the original was reviewed with help from LLMs.