Mapping digital innovation in EU regions

A policy brief on regional contributions to R&I and cross-border collaboration in the digital ecosystem

The European Commission has put startups, research and innovation among the key priorities of its mandate, with the goals of closing the innovation divide between the EU and its global competitors and boosting competitiveness.

On May 28th, 2025, the European Commission launched the EU Startup and Scaleup Strategy titled “Choose Europe to Start and Scale”. Its aim is to make Europe a great place to launch and grow global technology-driven, innovative companies. The strategy is structured around five targeted actions to support and finance startups and scaleups, attractand retain the best talents, and facilitating access to infrastructure, networks and services.The strategy represents one of the flagship actions of the first pillar “Closing the innovation gap” of the recently launched Competitiveness Compass, aimed at creating a framework that can significantly influence all industries by providing a structured approach to enhancing competitiveness across various sectors.

Understanding regional positioning in EU research and innovation

To better understand how the landscape on which the strategy will have its effects, it is worth observing how the different regions in the EU are positioned in terms of research and innovation (R&I). The approach needs to be organised through homogeneous sectors, like industrial sectors. To help in this classification, the European New Industrial Strategy has introduced the concept of industrial ecosystem, a new approach to regroup economic activities. In total, 14 industrial ecosystems have been identified, each of them covering a whole range of economic activities and purposes. 

HIGHLIGHTS The digital ecosystem comprises nearly 6,500 research organizations across Europe,concentrated around major cities like Barcelona, Paris, and Berlin, that appear to drive regional innovation activities.‣ Network analysis allows to identify key EU regions that serve as vital nodes for knowledge exchange and collaboration, revealing both global and EU-specific strategic positions.‣ Top EU regions for R&I are München, Paris, and Madrid, demonstrating high levels ofintermediation and cross-border collaboration within the digital ecosystem.‣ Economic indicators show varying levels of GDP per capita, employment-to-population ratios, and startup presence in top EU regions, suggesting areas for enhancing innovation-led growth.

A first application of the Techno-Economic ecoSystem (TES) approach to the analysis of the 14 industrial ecosystems has been performed through the ATLAS [1], extending the analysis of the DigitalTechno-Economic ecoSystem (DGTES) included in [2] and [3]. Moreover, some of the indicators included in the ATLAS are derived from network analysis, encompassing the dimension of traditional metrics and opening to new faceted views on the data. This policy brief draws from some insights of the ATLAS and narrows down the analysis to NUTS 3 EU regions, offering a unique picture of the R&I landscape. The ecosystem chosen for the analysis is the digital one, the historical flagship of the DGTES approach.

R&I in the EU

The ‘usual suspects’ when thinking of R&I are research organisations, intended as research institutions and universities. The digital ecosystem (DGTES) accounts for almost 6,500 players of this type, spread across the EU as represented in  Figure 1. Most of the players are concentrated around the biggest European cities: Barcelona, Paris, Madrid, Wien, Berlin, Rome, and so on.

Strategic position of EU regions

The R&I environment is identifiable in the DGTES ecosystem through two different types of activities; research activities are intended as academic contributions of frontier research, such as publications and/or participation in high level

international conferences, while innovation activities correspond to patenting initiatives (filing of priority patent applications) and/or participation in innovative research projects (EU-funded projects).

Once all R&I activities pertaining to each EU region are identified, we need an indicator able to rank each region according to their strategic position in the EU network. Betweenness centrality (also referred to as intermediation) is an indicator typically used in network analysis and suitable for this task; it represents a measure of the control influence of a node in the network acting as a bridge, connector or bottleneck between two nodes (see Box 1 for details). We computed this indicator at NUTS 3 level for the whole EU, considering the rest of the World as one single entity. Moreover, in our analysis we only considered collaborative activities (intended as co-applications in patents and

co-authorships in academic publications), namely those that do not happen exclusively within the borders of a region but that involve at least another NUTS 3 region (or the rest of the World). The betweenness centrality values were then normalised to take values between 0 and 1.

Table 1 and Figure 2 show the top ten EU regions in terms of intermediation, meaning those regions that play a key connecting role in the R&I digital ecosystem. Table 1 presents the NUTS 3 regions in descending order of betweenness centrality, and the size of the bubble in the last column gives a flavour of the magnitude of each region strategic position; Figure 2 takes a step further, visualising the same top 10 regions grouped by country.

Table 1 – Strategic position in R&I of top 10 NUTS 3 regions – considering rest of the World

The regions of Paris (France) and Madrid (Spain) can exert amoderate influence, with betweenness centrality values between 0,6 and 0,3. The other seven cities only have a low influence in the digital ecosystem.

Figure 2 – Strategic position in R&I of top 10 NUTS 3 regions grouped by country– considering rest of the World

A second approach excludes the rest of the World from the analysis and computes the strategic position considering linkages between NUTS 3 regions only.

Table 2 – Strategic position in R&I of top 10 NUTS 3 regions – only including NUTS 3 regions

Indeed, the position of München, Kreisfreie Stadt appears weaker than before, with Madrid and Paris sharing the leading role among NUTS 3 regions. Despite the difference influence that the three regions can exert in the global and in the EU network, it is possible to identify the main drivers and entities behind this force.

In this case, we notice a mix of universities and private companies leading the activities (Centre National de la Recherche Scientifique CNRS and Sorbonne Université, Agencia Estatal Consejo Superior de Investigaciones Científicas and Universidad Politécnica de Madrid on one side, Orange SA, Atos Spain SA and Telefonica Investigación y Desarrollo SA on the other).

Figure 3 – Strategic position in R&I of top 10 NUTS 3 regions grouped by country – only including NUTS 3 regions

Additionally, some regions disappear from the top ten: Stockholm (Finland), Valencia (Spain), Dresden and Erlangen (Germany), meaning they represent areas that are more strategic on a global level. A closer look to these regions reveals that Valencia and Dresden R&I activities are mainly driven by EU funded projects, where a few entities collaborating with international ones can play a key role in the global context (Universitat Politecnica de Valencia and Universitat de Valencia for Spain, Technische Universitaet Dresden for Germany). Stockholm and Dresden regions appear to be driven by important international co-patenting activities with two main companies: Telefonaktiebolaget LM Ericsson for the 5G and connectivity digital area, and Siemens Healthcare GMBH on the topic of AI, respectively.

The strategic position of the first three regions is driven by EU funded projects carried out by several entities in collaboration with other intra-EU ones (Vrije Universiteit Brussels and Université Libre de Brussels, different universities in Barcelona, foundations and technological centres, as well as themain universities in Milan and a couple of Italian companies – Telecom Italia SpA and Hewlett Packard Italiana Srl). The strategic position of Zuidoost-Noord-Brabant is indeed due to the co-patenting intra-EU activities of several companies (Koninklijke Philips N.V., NXP B.V, Here Global B.V., Additive Industries B.V. and ASML Netherlands B.V.).

While most of the top ten regions are strategic on both levels (global and EU-only), this analysis already gives a flavour of how diversified the EU R&I environment is and how it allows to capture some peculiarities of different regions.

The top strategic regions in the EU

The identification of top regions in terms of R&I only represents the first step of the analysis, as it can give a feeling of the most influential areas in the EU. It is then crucial to deepen the analysis of top EU regions (identified through both approaches from the previous section), highlighting some economic features of those territories combined with some peculiar business presence (namely, startups).

Indeed, we think that the low presence of innovative startups in a region that has been identified among the topmost strategic ones in R&I could be the sign of the presence of opportunities for investors and innovators.

For each of the top EU regions in R&I, Figure 4 highlights three different aspects:

• Employment-to-Population (total) ratio (on the x axis), computed as the share of employment over the total population for year 2022 – source: Eurostat.

• Gross Domestic Product (GDP) at current market prices per inhabitant (on the y German regions behave very differently, with axis) for year 2022 – source: Eurostat.

• Number of startups in R&I (bubble size), identified as the number of startups in DGTES that in the period 2009-2024 present at least one innovation or research activity – source: DGTES.

The combination of the first two aspects allows to characterise the top EU regions in terms of R&I through a quadrant, while the bubble size of each region gives a flavour of the presence of innovative startups.

Figure 4 – Top NUTS 3 regions in terms of strategic position in R&I by employment-to-population (total) ratio, GDP per capita and number of startups in R&I

The figure allows to spot how the top three regions present completely different features; while Paris is characterised by high values of GDP per capita and employment-to-population ratio, Madrid region shows low GDP and employment-to-population ratio, while München sits in the middle. In terms of number of R&I startups, the three regions show a similar pattern, with a consistent presence of this type of business.

Low GDP and employment-to-population ratio are common features for the top Spanish regions (Valencia and Barcelona), placed at the bottom left corner of the quadrant. German regions behave very differently, with München acting as the top global region with substantial GDP, employment-to-population ratio and number of startups, but that gets overcome on the first two indicators by the Erlangen, Kreisfreie Stadt region, that on the contrary does not present any startup in R&I. As anticipated, the strategic global position in R&I of the latter region (even if low compared to the top three) could reveal a possibility for innovators and investors.

Regional Innovation Dynamics Across Europe

Among the remaining regions, Helsinki-Uusimaa excels with the highest number of R&I startups, while the arrondissement of Brussels-Hoofdstad only hosts one – another region that could hidden some business potential. The two regions are comparable in terms of employment-to-population ratio but show a gap of around EUR 18k in terms of GDP per capita in favour of the Belgian region. 

Of course, other external factors could influence the economic situation and business opportunities of the regions, but this analysis constitutes a first step towards unveiling some different perspectives on EU regions and their potential in R&I.

Conclusions

The analysis is a first attempt at mapping the R&I environment of the EU in the digital ecosystem at sub-regional level. The presence of several research organisations in an EU region is not enough for the region itself to be considered in a strategic position. This is due not only to the different nuances that characterise the R&I activities (research publications, co-patenting as well as EU funded projects), but also to the need of collaboration across borders.

Our approach can build a network for R&I considering only the collaborative activities (meaning those that involve more than one NUTS 3 region) and to compute an indicator (betweenness centrality) that characterises each region concerning their strategic (or not) role in a global and EU-only network.

Economic and startup Dynamics in strategic EU Regions

Results show how some EU regions can exert a key role both in the global and in the EU-only network, while some others invest a more specialised position. A focus on some features of the top EU regions in terms of GDP and employment-to-population ratio can unveil details on the economic environment in which the regions operate. Moreover, the presence (or absence) of startups with R&I activities in those regions can be intended as an opportunity for business. Indeed, as the DGTES approach works in a and research activities attributes to identify cumulative way, grouping all R&I activities that have been carried out throughout the years and their corresponding players, the absence of startups in a region that is considered strategic could be interpreted as a possible gap to develop new innovation and research activities through startups.

This first attempt has been applied to the digital ecosystem considering the long-term expertise that the DGTES team gained on the topic throughout the years, but the same analysis can be extended to other industrial ecosystems to discover the strategic role of EU regions and unveil new opportunities. At the same time, the analysis could be disaggregated by each of the 15 DGTES policy-relevant Digital Areas (DAs) – See in References [2] for the full description and list of DAs – to create a more tailored analysis and deep dive into specific topics that are part of the digital ecosystem.

Methodology

We identify research, innovation, business and investment activities related to forward looking digital technologies in each NUTS 3 EU region for years 2009-2024. Innovation activities are identified among patent filings sourced from European Patent Office’s Patstat, containing textual information on the invention description and its applicant, and innovative research projects from CORDIS data on Horizon Europe, Horizon 2020 and 7th Framework Programme for Research. Research activities are identified among publications on a selection of journals related to the digital ecosystem from Scopus. DGTES identifies activities relevant to the digital ecosystem through a comprehensive dictionary of keywords on digital technologies. We use information coming from innovation and research activities attributes to identifywho is innovating in the digital ecosystem (players), their geographical distribution, and the connections among them as they co-patent, collaborate on a publication or in an EU funded project forming a network of collaborations. These connections are not limited to one region but can also happen across borders. We exploit this network and generate indicators to analyse EU regions’competitiveness in the digital ecosystem. Startups are identified as DGTES players that during years 2009-2024 have received at least one venture capital (VC) early-stage deal, identified through the data source Crunchbase. Betweenness centrality is a network indicator that measures the relative position of a node and the extent to which it serves as a key broker between many other nodes. It allows to identify key players or influential nodes within a network, as it computes how often a node lies in the shortest path between two nodes. The value is normalised between 0 and 1; a region with a high intermediation score is considered a gatekeeper of knowledge flowing across the EU.

References

[1] Signorelli, S., Fabiani, J. and De Prato, G., ATLAS: An Analytical Tool for Linking and Assessing industrial ecoSystems – Measuring investment, innovation and the EU’s role in worldwide industrial ecosystems, Publications Office of the European Union, Luxembourg, 2025.

[2] Calza, E., Dalla Benetta, A., Kostić, U., Mitton, I., Vazquez-Prada Baillet, M., Carenini, M., Cira, P., De Prato, G., Righi, R., Papazoglou, M., Lopez Cobo, M. and Cardona, M., A policy oriented analytical approach to map the digital ecosystem (DGTES), EUR 31319 EN, Publications Office of the European Union, Luxembourg, 2022.

[3] Calza, E., Dalla Benetta, A., Kostić, U., Mitton, I., Moraschini, M., Vazquez-Prada Baillet, M., Cardona, M., Papazoglou, M., Righi, R., Torrecillas Jodar, J., Lopez Cobo, M., Cira, P. and De Prato, G., Analytical insights into the global digital ecosystem (DGTES), EUR 31538 EN, Publications Office of the European Union, Luxembourg, 2023.

ATTRIBUTION: Republished from the European Commission’s Joint Research Centre (JRC) publication: “Strategic NUTS 3 Regions in the EU: Analysing Research, Innovation and Startups in the Digital Ecosystem”, European Commission, Seville, 2025, JRC143080 under a CC BY 4.0 license. Copyrights: © European Union, 2025.