The HCSS Datalab

Strategic decision-making is inherently subject to significant risk. Uncertainty makes it difficult to engage in long-term decision making, preventing you from taking bold action and realising your desired impact or potential. Within the HCSS Datalab we model complex phenomena, quantify and compare the impact of decisions, to provide insights into such risk. This allows us to gain a better understanding into the everchanging world of geopolitics, defence, and security. Using data and in-depth subject matter knowledge we can help to enhance decision-making and better communicate stories to in turn reach the desired audiences.

Datalab Academy Datalab Publications

Natural Language Processing

– We turn text into insight

Every day we collect and analyse large volumes of textual data at scale, from news and policy documents to speeches and reports, extracting patterns, signals and narratives that no single analyst could hold in view at once. Using NLP and generative AI, we map the lay of the land across strategic landscapes and scan the horizon for themes just beginning to surface. Those findings feed personalised intelligence reports, tailored to a specific question, region or portfolio, so that analysis reaches decision-makers in the form they actually need.

Surveying

– We measure attitudes at scale

We design and run large, representative, longitudinal population surveys that capture public attitudes, perceptions and trends across society. Fielded several times a year, they have built up one of the largest collections of Dutch population-survey data in the Netherlands and Europe, and provide a robust empirical foundation for tracking societal dynamics over time. The results inform our analyses, validate assumptions and ground strategic research in lived public experience.

Modelling

– We model complex landscapes

We build digital representations of geopolitical, security and economic environments to understand how actors, systems and dynamics interact. By combining network analysis, systems thinking and agent-based modelling, we explore scenarios, test assumptions and support decision-making under uncertainty. This includes generative agent-based modelling, in which populations of AI-driven agents reason, adapt and interact rather than follow fixed rules. These models translate complex realities into structured insights for analysis, policy discussions and serious games.

Products

Previous Next

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

2026 Rapport Cover testpage housestyle (3)

Research as a Product

– We turn quant to qual insights

We treat reports as carefully designed products that translate quantitative research into clear, qualitative insight that is credible, interpretable, and policy or scientifically relevant. By combining quantitative analysis with qualitative interpretation and contextualization, we produce publications that go beyond presenting results: they explain meaning, surface implications, and structure strategic or scientific debate.

Tool as a Product

– We build end-to-end solutions

We design, build, and maintain secure digital tools and platforms that support analysis, collaboration, and decision-making from start to finish. These tools are developed as reusable products rather than one-off project artefacts, and are actively adopted through onboarding, integration into workflows, and clear value propositions. This allows analytical insights to scale beyond individual projects and reach a broader set of users.

Previous Next

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Data as a Product

– We provide datasets

We create, enrich, and publish high-quality datasets as first-class products, designed for reuse by researchers, policymakers, and partners. By combining original data collection, enrichment, and transparent documentation, we make HCSS-owned data accessible through a publicly available APIs, following examples such as ACLED (operational) and POLECAT (terminated). This approach turns data into a durable asset that enables external research and collaboration with third parties.