The European video game industry has become a major cultural and economic force. From indie studios to global publishers, video games shape creative labour markets, digital innovation, and regional development across Europe.
Yet, despite its growing relevance, the industry remains surprisingly hard to measure. How many video game companies operate in Europe? Where are they located? Do they form regional clusters, and if so, what do these clusters look like?
These questions are at the core of the Gaming Clusters Across Multiple European Regions (GAME-ER) Horizon Europe project that developed a data-driven strategy to systematically map the European video game industry. Drawing on large online repositories, text-matching techniques and spatial analysis, GAME-ER aims to overcome the limitations of traditional statistics and provide a new macro-level picture of the sector.
Measuring the video game industry is not straightforward. Standard industrial classification systems (such as NACE or SIC codes) were not designed with creative and digital industries in mind. This problem affects cultural and creative industries in general, but it is particularly severe for video games.
At the European level, only one NACE code explicitly refers to video games (“Publishing of computer games”). In practice, however, game developers and publishers are scattered across many other codes, ranging from computer programming to artistic creation, motion picture, manufacturing of games and toys, or even specialized design. As a result, official statistics systematically underestimate the size of the sector and blur its spatial organization.
Similarly, national data often rely on trade association membership lists or ad-hoc surveys, which are rarely comparable across countries and tend to exclude smaller or less visible studios. The outcome is a fragmented and incomplete picture of an industry that is, by nature, at the intersection of different sectors fast-evolving.
GAME-ER tackles these issues by shifting the focus from administrative classifications to the information available on online repositories. Instead of asking how firms are classified in official registers, the project asks a different question: where do video game companies actually appear to operate based on their digital footprint?
The strategy combines a small number of rich sources. The core idea is to identify video game companies through specialised web directories, then match this information with business registers to obtain geographic and economic data. At the heart of this approach are three main data sources:
One of the most informative results of the mapping exercise concerns the distribution of video game companies across NACE codes. The dataset reveals that video game production in Europe spans more than thirty different NACE classifications, with no single code dominating across countries.
This empirical evidence confirms that the video game industry cannot be understood as a neatly bounded sector. Instead, it emerges as a highly fragmented and hybrid activity, embedded in multiple segments of the economy. This helps explain why official statistics struggle to capture it, and why policy interventions based on narrow classifications risk missing a substantial share of actual production.
By documenting this fragmentation quantitatively, GAME-ER provides a solid empirical basis to rethink how the video game industry is defined and measured in Europe.
The construction of the dataset follows two main steps.
Step 1: Identifying video game companies and their nationality
Company names from Mobygames are matched with OGDB using text-similarity techniques. Because company names often differ slightly across sources, the process combines multiple similarity algorithms after careful name cleaning. For companies that could not be matched automatically, large language models are used to retrieve missing information on country of origin. Through this process, GAME-ER identifies 9,583 video game companies operating in Europe between 1975 and 2024, of which 3,509 were active in the period 2020-2024. This figure already illustrates an important point: depending on the source and method used, estimates of the industry’s size can differ substantially.
Step 2: Geolocating companies at city and regional level
In the second step, European companies are matched with Orbis data using a similar text-matching strategy, allowing researchers to locate firms at the city level and aggregate them to NUTS-2 regions. The current dataset includes 4,067 European video game companies with reliable geographic information. This step also highlights key trade-offs. Matching rates vary across countries, and smaller studios are harder to trace in business registers. As a result, the final dataset tends to over-represent larger and more commercially visible firms. Rather than hiding these limitations, the GAME-ER approach explicitly documents them, making the resulting maps more transparent and interpretable.
The data-driven mapping developed in GAME-ER opens new analytical possibilities. It allows researchers to:
More broadly, it shows how digital traces and experimental data methods can complement official statistics when studying creative industries that do not fit neatly into existing classifications.
The mapping of the European video game industry is still a work in progress. Ongoing work focuses on validating matching procedures, improving business information and refining the characterisation of companies and clusters. Thanks to this approach, GAME-ER will deliver a comprehensive and replicable map of the sector, offering a new empirical foundation for research and policy discussions. In doing so, the project contributes to a broader agenda: making visible those creative industries that are economically significant, culturally influential, and yet statistically elusive.
