Europe’s Scaleup Geography and the Role of Access to Talent

Abstract

Although the concept of the ecosystem for entrepreneurs and startups dominates the discourse on promoting innovation, the geographical pattern of cities with startups remains largely uncharted territory. This study’s author first gives an overview of this concept’s roots in regional sciences and then outlines the contours of Europe’s scaleup geography by analyzing data on 12,500 scaleups (startups that raised more than 1 million euros) at the municipal level. Performance is marked by a strong West-East and North-South divide, and most scaleups and funding concentrate in a handful of cities. One may assume some convergence, however, from the upstream of locally available talent, especially in university towns. The study’s author concludes that scaleup ecosystems in Europe have only developed into a self-reinforcing system in a few cities. Although they have become self-sustaining in a multitude of Western and Northern European cities, large cities in Southern Europe and Central and Eastern Europe lag behind, not least because of the weak upstream of university talent.

The Entrepreneurial Ecosystem: From a Metaphor to an Analytical Approach

Although startups play a vital role in innovation and have become one of the main drivers of current industrial revolution, researchers know little about their geographical distribution. This, however, largely contradicts the fact that use of today’s buzzword “entrepreneurial ecosystem” (Brown & Mason, 2017; Brown & Mawson, 2019)—a term conceived by Moore (1993) primarily as a metaphor rather than as a research and policy concept—aims to place the entrepreneur center stage (Acs, Stam, Audretsch, & O’Connor, 2017; Audretsch, Cunningham, Kuratko, Lehmann, & Menter, 2019). More precisely, those using this approach largely focus on analyzing and supporting that institutional environment which supports and enables the creation of new firms and businesses (De Meyer & Williamson, 2020; Isenberg, 2010; Spigel, 2017).

The first drawback is that the list of such institutional actors is exceedingly long, including, for example, universities, research institutes, technology centers, large multinational, nonprofit organizations, incubators, accelerators, business organizations, banks, venture capital, angel investors, and governmental organizations. Yet skilled labor and talents are also a prerequisite, and cultural factors, including success stories and societal norms, may play a major role. From this broad spectrum, scholars have made numerous references to the pivotal role of universities (Heaton, Siegel, & Teece, 2019) and in particular to knowledge transfer and spillover through university spin-offs (Graham, 2014; Grimaldi, Kenney, Siegel, & Wright, 2011; Scott, 2002). The publicly funded and operated R&D and innovation agencies—such as the Defense Advanced Research Projects Agency (DARPA) in the US, the Finnish Funding Agency for Technology and Innovation (TEKES) in Finland, or the European Institute of Innovation and Technology (EIT) in Europe—form another widely studied area, not least because they are embedded in the idea of public-private partnership for innovation (Block, 2008; Mazzucato, 2013). Still, in both cases, policymakers face the same challenge: How can they foster new technological breakthroughs by promoting and subsidizing those young companies and startups that are not yet even born? Kay (2011) puts it succinctly: “[I]f an industry is to advance, much—perhaps all—innovation will come from businesses that don’t yet exist” (pp. 9–10).

A further and more serious drawback is that when applying the ecosystem metaphor to innovative technology regions with a vibrant entrepreneurial culture and critical mass of startups, a fundamental contradiction arises. The word “ecosystem” is associated with stability, resilience, and organic development, whereas innovative entrepreneurs and especially startups in the sense of Schumpeter’s creative destruction hew more closely to the disruptive character of innovation. Researchers have shown that startups are those companies from which one expects technological innovations and innovative products, whereas large corporations usually incrementally and continuously develop their products with systematic R&D work (Baumol, Litan, & Schramm, 2007; Christensen, 1997; Tirole, 2017). Yet these Schumpeterian dynamics and the ceaseless struggle between the incumbent companies with old technologies and the frontier firms with new technologies (Aghion, Antonin, & Bunel, 2021; Phelps, 2013), between breakthrough innovation through startups and incremental innovation driven by incumbent companies, hardly fit the ecosystem metaphor (De Meyer & Williamson, 2020; Fransman, 2018).

The Battle of Narratives

Although the term “ecosystem” in relation to entrepreneurs, startups, and innovation stretches back to the mid-2000s, this strongly policy-oriented application of the concept is deeply rooted in the regional sciences of previous decades. As early as the 1990s, Storper (1997) outlined the “Holy Trinity” of regional economic development as the three fundamental elements of technology, institutions, and the region with its features jointly interacting to influence the economic development of a given region. A decade later, Etzkowitz (2008) introduced the “Triple Helix” as an analytical framework for innovation clusters, according to which innovation can be found where the activities of the three principal actors of university sphere, industry, and state intersect, or where their border areas mutually overlap, and hybrid organizations are created (Ranga & Etzkowitz, 2013).

Today’s policy thinking on promoting innovation revolves around the concept of open innovation, whose advocates emphasize the cocreation of knowledge in multiplayer networks (Andersen, de Silva, & Levy, 2013; Hutton, 2015). This is the result of a decades-long journey in four distinguishable phases. They are clearly recognizable by the frequencies with which the related terms are mentioned in the literature, taken from Google Books Ngram Viewer that records a great bulk of books (see Fig. 6.1).

Fig. 6.1

Frequencies of terms for the different policies promoting innovation in printed sources, 1960–2019 (frequencies in %). Source: Data retrieved December 29, 2021, from Google Books Ngram Viewer (See https://books.google.com/ngrams/). Design by author

The 1950s and the 1960s were the decades of science policy, characterized by the hope that the results of state-financed basic research would spread to the economy almost automatically. In the 1970s and the 1980s, technology policy received more attention and the state no longer heavily subsidized research projects, but rather some technologies in a broader sense. Then came the decades of innovation policy, in the 1990s and 2000s, with the catchphrase “knowledge transfer,” and thus the basic problem arose of finding institutions and organizations that could facilitate the flow of technological knowledge towards economic actors.

To speak of the new stream of open innovation since the early 2010s is to acknowledge the simple fact that no organization developing a new product or technology today can claim that they know everything and have no need for the knowledge of others. In a stricter, corporate management sense, as Chesbrough (2003) first used the concept, companies open up their boundaries to their environment in R&D. In a broader context, today new product and new technology is developed in a dense network of interactions, sharing, cocreating, and cofunding of numerous actors, including of course the state, the business sphere, the universities, the small startups and large corporates, the financial sector, and the supporting institutions, whereas the Internet and unlimited global and mobile accessibility enable almost everybody to join the network of open innovation.

As for other theoretical roots, researchers of industrial clusters have found a direct route into the concept of the entrepreneurial ecosystem (Spigel & Harrison, 2018). This goes back to the 1990s and Porter’s seminal paper (1998), in which he underscored the prominent role of geographical concertation of interconnected companies and firms in related industries and associated organizations (i.e., universities and research institutes, financial intermediaries, and business-related services). At the same time, case studies proliferated on technology-based industrial clusters, such as Silicon Valley and Boston’s Route 128 (Kenney, 2000; Lee, Miller, Hancock, & Rowen, 2000; Saxenian, 1996), and on technopoles and sciences cities (Castells & Hall, 1994), such as the Research Triangle in North Carolina, Tsukuba and Kumamoto in Japan, the “Silicon Fen” around Cambridge in the UK (Koepp, 2002), and Sophia Antipolis in France.

Further roots lie in the concept of the industrial district that emerged in the 1990s. The reasoning behind this was that small companies—if they embed themselves in regional networks while cooperating and competing with each other—can be very innovative and make the entire region more competitive (Pyke & Sengenberger, 1992). A prime example is the so-called “Third Italy,” which refers to the north-east and central parts of the country, in which numerous industrial districts have developed through locally embedded collaborations of small and medium-sized companies mainly specializing in craft-based manufacturing (Pyke, Becattini, & Sengenberger, 1990). Nevertheless, here again success is rooted in many factors, such as the localized knowledge production and spillover trough interaction between the firms (Maskell & Malmberg, 1999), their networks (Camagni, 1991) and their flexible specialization (Piore & Sabel, 1984). Undoubtedly, all these ideas are indebted to a large extent to Marshall (1919) and his theory on industrial districts.

Looking back over the past few decades, the narratives surrounding regionally anchored cooperation and competition between different actors promoting new technologies, innovation, and new companies seem to have changed radically. With a simple glance at the frequencies with which the related terms are mentioned in the literature—taken from Google Books Ngram Viewer—one can clearly see that whereas the ecosystem concept has won the battle of narratives, the theory of industrial districts and clusters, which are more closely linked to a seemingly outdated industrial policy, has gradually lost its relevance (see Fig. 6.2).

Fig. 6.2

Frequencies of terms for the different concepts for regionally embedded innovation in printed sources, 1990–2019 (frequencies in %). Source: Data retrieved December 29, 2021, from Google Books Ngram Viewer (See https://books.google.com/ngrams/). Design by author

There are many reasons for this shift, but most of them are related to changes in technology. On the one hand, today’s almost ubiquitous digital technology makes the regionally embedded interconnectedness of different actors faster and cheaper than ever. Yet the same digital technology allows companies, especially startups, to scale and grow faster and more economically than ever before. In short, policies that encourage innovation and entrepreneurial ecosystem require relatively little investment but promise high returns. However, ample evidence exists that policymakers virtually everywhere in Europe are aiming to create their own Silicon Valley: “[T]aking on a name, and perhaps establishing some business incubators or building a few semiconductor firms, PC factories, or software houses, is not enough” (Lee et al., 2000, p. 3). Even if an extensive literature on ecosystems for innovation, entrepreneurs, and startups supports these efforts, in practice it is hardly possible to implement these concepts without market forces, the flesh-and-blood startup founders, and venture capitalists (Lerner, 2009).

Why the Scaleups?

In this study, I focus on startups in terms of innovation and entrepreneurial ecosystem, for multiple reasons. In a comparison between startups and big corporates, Tirole (2017, p. 443) rightly pointed out that today “innovation happens more and more in small entrepreneurial startups rather than in large companies.” Corporate management is interested in safeguarding the market of their existing products, so why should they support intra-company development of those new products that would eventually eat up the market opportunities of the previous ones? Startups also have the upper hand against corporations with those innovations that require mainly intellectual capital and relatively low capital investment (where corporations will always hold the trump card). They also have advantages in areas with strong competition for users and consumers, where the market is not covered by a few large enterprises. They often win in fields where innovation does not require deep scientific knowledge or expensively equipped laboratories, and as Phelps (2013) underscores, innovation is not the preserve of the elite—most of the time, innovation is not rocket science or high-tech.

However, digital technologies increase startups’ chances enormously. Once again, a quick glimpse into the frequencies with which the terms “digital technologies” and “startups” are mentioned in the literature reveals that they have been going hand in hand and that their effects are mutual (see Fig. 6.3).

Fig. 6.3

Frequencies of terms digital technologies, startups, and open innovation in printed sources, 1970–2019 (frequencies in %). Source: Data retrieved January 31, 2022, from Google Books Ngram Viewer (See https://books.google.com/ngrams/). Design by author

Digital technologies lower the barriers to market entry, and thus open up more opportunities for startups; vice versa, those startups are driving the development of digital technologies. In addition, digital technologies facilitate extraordinarily the combination of different business and technology fields, which is the very essence of innovation and thus of great potential benefit for startups.

Back in the middle of the last century, Schumpeter not only glorified the entrepreneur as the engine of development, creating new products, new methods of production, or new forms of industrial organization (Schumpeter, 1942/2003, p. 82), but he was also aware of what new means in most cases, as “innovation combines factors in a new way, or that it consists in carrying out new combinations” (Schumpeter, 1939/1989, p. 62). Today Ridley (2020, p. 250) formulates this insight more generally, emphasizing that every innovation is recombinant, and “every technology is a combination of other technologies, every idea is combination of other ideas,” it is digital technology which makes these combinations easier, faster, and cheaper. As Brynjolfsson and McAfee (2014, p. 78) recognize, “the true work of innovation is not coming up with something big and new, but instead recombining things that already exist.” By listing many well-known examples from Google’s self-driving car to Facebook and Instagram, they conclude that “digital innovation is recombinant in its purest form” (Brynjolfsson & McAfee, 2014, p. 81).

Yet startups are not only benefiting from this shift towards recombinant and open innovation—they are also taking advantage of the increasing role of intangibles in the modern economy, from software to intellectual property rights, from brand value to large databases. Haskel and Westlake (2017, 2022) underline that in our age when investment in intangible assets becomes increasingly important, a crucial property of intangibles, the synergy, has a critical impact on innovation. Because ideas and other ideas go well together, especially in technology, intangibles are often particularly valuable when properly combined with other intangibles. This is precisely what paves the way for startups, which are typically involved in the innovation process when knowledge and human capital are the assets to be leveraged. Another advantage of intangibles, especially in relation to digitized assets or platforms with network externalities, is that the companies relying heavily on them can grow exponentially and scale globally at unprecedented speed (Azhar, 2021). All of this combined is generating a winner-take-all frenzy, the rise of superstar firms (Aghion et al., 2021; Autor, Dorn, Katz, Patterson, & van Reenen, 2020), and a growing gap between the front-runners and laggards, as the latter are usually engaged in tangible economy.

The advantages startups hold in digital technologies and innovation are obvious and, in this study, I apply Graham’s (2012) approach and use the term “startup” in line with my main information source, the Dealroom, whose authors define a startup as “a company designed to growth fast” (Wijngaarde, 2021). This would allow one to avoid arbitrary thresholds for various metrics such as age, technology, funding structure, market value, or employment structure of firms. However, for deeper research, the problem arises that there is a skewed distribution of startups that somewhat follows the power law when one examines the relationship between the amount of funding and the number of startups, and the vast majority of startups tend to receive very minor funds or no funds at all (Cséfalvay, 2021). Similarly, only a tiny fraction of startups is responsible for the bulk of innovations and technological breakthroughs, whereas most are caught in the early stages of launching a new business with a marketable product.

For this reason, by analyzing the growth stages of young companies, Flamholtz and Randle (2015) distinguish the “organizational scaleups,” which are those startups that have already received significant funding, developed a marketable product and viable business model, and therefore are able to grow quickly. For a startup to qualify as a scaleup, the various startup ecosystem ranking institutions (Dealroom & Sifted, 2021; Durban, 2021; Erasmus Centre for Entrepreneurship, 2021) set numerous criteria to be met, such as annual growth, number of employees, or annual turnover. Yet what they have in common is that scaleups are those startups that have already raised at least US$1 million in funding.

Whereas Ries (2011, p. 27) famously defined a startup as “a structure designed to create a new product or service under conditions of extreme uncertainty,” scaleups have already passed the stage of extreme uncertainty. To quote another often-cited definition of Blank and Dorf (2020, p. xvii), who describe a startup as “a temporary organization designed to search for a repeatable and scalable business model,” scaleups have already found their business model and have marketable products. In short, scaleups are successful startups that are economically relevant and have growth prospects and as such can make a significant contribution to the entrepreneurial ecosystem of a city or a region.

Why the Cities?

It is evident that policies targeting the ecosystem for innovation, entrepreneurship, and startups include increasingly place-based measures. The crucial question is, however, what kind of places are best suited today for establishing such an ecosystem, and, in particular, how to stimulate its dynamics (Bailey, Pitelis, & Tomlinson, 2018).

In this context, Florida (2017) stressed that the recent “urban shift of the high-tech startup companies und talent is a real sea change” (p. 42). On the one hand, it was a long-awaited phenomenon and, on the other hand, a contradiction to the period from the 1970s to the turn of the millennium, when high-tech industries, venture capital investment, and startups moved to the edges of suburbs like Silicon Valley or Boston’s Route 128. How, however, apart from a few previously established corporate campuses of today’s digital giants, the startups are leaving—as Kotkin (2000) puts it—the “Nerdistan,” the sprawling, car-oriented suburban periphery with office parks, for the vibrant and dense cities with creative milieu. Whereas the venture capital investment and venture capital-backed startups of the 1980s and 1990s clustered around the fringe of suburban areas, today it is the city that is becoming a booming “startup machine” (Florida, Adler, King, & Mellander, 2020).

Cities have always been the centers of knowledge production and transfer, so they offer an almost natural fit for startups. What is new is their comeback, and the drivers beyond are again increasingly technological. Since the beginning of the last decade, society has been experiencing the Fourth Industrial Revolution. with new technologies such as artificial intelligence, big data analytics, blockchain, biotechnology, and nanotechnology (Schwab, 2016); with new means of production such as digitization, robotization and automation; and with the overarching economic shift from tangible to intangible assets and investments. One of the common denominators of these technologies is that they are less geared towards hardware and more towards software and intangibles, and thus do not necessarily require large office spaces or manufacturing capacities, the easy and cheap availability of which once fueled the rise of the suburban periphery. Consequently, startups are now moving from the suburban areas to the cities to benefit from the dense network and cluster of universities, research institutes, venture capital funds, high-tech services, and the creative milieu. As Florida and Mellander (2016) summarize this shift: “[T]he suburban model might have been a historical aberration, and innovation, creativity, and entrepreneurship are realigning in the same urban centers that traditionally fostered them” (p. 14).

Research Questions, Data, and Methodology

As a backdrop for this brief overview of startups, scaleups, and the entrepreneurial ecosystem, let me lay out the two objectives of my study.

The first is to analyze the European scaleup landscape in terms of municipal performances and to look in detail at the territorial distribution of scaleups across the European cities. Examining the well-known startup ecosystem rankings (Dealroom & Sifted, 2021; Erasmus Centre for Entrepreneurship, 2021; Startup Genome, 2021; StartupBlink, 2021), one can conclude that a few large cities dominate the landscape. However, my aim is to include every European city with considerable scaleup performances in order to provide a deeper insight into the geographic pattern.

The second objective is to investigate how access to locally available talent affects this landscape. Does it reinforce the trend to concentration, or does it even weaken this tendency? Do the cities with good access to talent have a chance to compete with the big scaleup cities? Or, conversely, does poor access to talent pose an obstacle for scaleup cities to strengthen their position in the European scaleup city landscape?

My main source of information to answer these questions is the Dealroom.co, a leading global platform for intelligence on startups whose authors provide comprehensive data on venture-backed startups in every country throughout the world, with a detailed breakdown by location, industry, technology, funding, founders, investors, and market value. As I am focusing on scaleups, which I here define as startups that raised more than €1 million in funding, my team members and I retrieved a total of 13,851 scaleups headquartered in Europe from the Dealroom database. As for the territorial distribution of scaleups, we applied the EU-OECD classification of Functional Urban Areas (FUAs) (Dijkstra, Poelman, & Veneri, 2019; OECD, 2021). A FUA consists of a city (core) and its commuter zone and thus encompasses the economic and functional expansion of the city, with the great advantage of available corresponding economic data, such as population and GDP.

To analyze the European scaleup landscape, we matched the 13,851 scaleups retrieved from Dealroom database with their respective FUAs by using the Tableau software. We then applied three variables to measure cities’ performance at the FUA level in terms of scaleups: the number of scaleups, the total funding of scaleups, and the number of scaleups with a market value more than €200 million. Based on this, we performed a cluster analysis to filter FUAs with considerable performance in terms of scaleups; in particular, we applied k-means algorithm, and this resulted in total of 166 FUAs (consisting of 12,472 scaleups), which were arranged in six clusters (see Table 6.1).

Table 6.1 Descriptive statistics of the scaleup city clusters in Europe, 2021

Whereas Global scaleup cities excel in every way and play an important role not only at the European but also at the global level, Top European scaleup cities perform less well in terms of the scaleups’ numbers and market values and occupy a leading position only in Europe. Top European Emerging and Emerging scaleup cities feature relatively strong funding but lag far behind in growth, measured by the number of scaleups with a market value of more than €200 million. In contrast, regional and local scaleup cities perform very weakly in all aspects.

Towards Europe’s Scaleups Geography

Skewed Distribution of Scaleup Cities in the European Scaleup City Landscape

The skewed distribution of startups and scaleups, with a few companies concentrating most of the funding and the vast majority receiving very little, is also reflected in the landscape of European scaleup cities. Of the 166 scaleup cities, only a handful—global and top European scaleup cities, 15 in total—concentrate 61% of the European scaleups, 71% of their funding, and 68% of the scaleups with a market value of more than €200 million (see Table 6.2).

Table 6.2 The distribution of scaleup city clusters according to their main performance variables

Nevertheless, a scaleup city’s development is a lengthy and complex process influenced by a number of crucial factors. When a city begins to concentrate startups and scaleups and an ecosystem with universities, risk capital, entrepreneurial expertise, and supportive institutions evolves, the first challenge is to maintain them to make the development self-sustaining. Yet regional science researchers—particularly those studying industrial clusters and districts (Castells, 2000; Saxenian, 1996) and more recently innovation and entrepreneurial ecosystems (Engel, 2014)—have long proven that once a critical mass of these factors is reached, the ecosystem evolves into a self-reinforcing system that is able to attract startups, scaleups, investments, and talents, first from a larger region and later from around the world.

Therefore, on the one hand, competition between scaleup cities is about to grow to the point where the ecosystem becomes self-sustaining; on the other hand, beyond this point, it is also about to globally attract main resources, primarily talent and capital. Looking at the figures on the distribution of scaleup cities in terms of performance indicators (see again Table 6.2), the ecosystem’s development is in the initial stages in the vast majority of European scaleup cities. Most of these are trying to develop a self-sustaining ecosystem, whereas only few scaleup cities have reached the point where development becomes self-reinforcing and increasingly attracts global resources.

West-East and Nord-South Gaps

Although the distribution of scaleup cities by performance indicators conforms to the widely held claim that ecosystems are concentrated in a few hubs that hold the overwhelming majority of scaleups and funding, with a detailed analysis one can paint a different picture—one with strong territorial gaps. Europe is marked by a deep West-East and North-South divide, and even large metropolitan areas in Central and Eastern Europe and in Southern Europe lag far behind when it comes to the number of scaleups and the funding they raised.

In terms of the number of scaleups, the landscape is dominated by the large Western European capitals, which also fall in the cluster of Global and top European scaleup cities (see Fig. 6.4). With almost 5000 scaleups combined, Global scaleup cities—London, Paris, Berlin and Stockholm—concentrate 40% of scaleups in Europe. Top European scaleup cities—for example, Barcelona, Copenhagen, Dublin, Helsinki, Madrid, Amsterdam, Munich, Cambridge, Manchester, Oxford, and Zurich—host over 2500 scaleups, forming a further 20%.

Fig. 6.4

The number of scaleups across the European scaleup cities, 2021. Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

In striking contrast, the 15 scaleup cities of Central and Eastern Europe—for example, Prague, Budapest, Tallinn, Vilnius, Gdansk, Poznan, Wroclaw, Cracow, Warsaw, Bucharest, Bratislava, Ljubljana, Riga, Sofia and Zagreb—offer a total of only 443 scaleups, equal to 3.5% of all European scaleups. Just for comparison, this lies above the corresponding values of Dublin (339 scaleups), but below those of Stockholm (489). Similarly, capitals in Southern Europe—Rome, Athens, and Lisbon—have put together fewer than 130 scaleups, which is in line with the values of Lausanne or Edinburgh.

In terms of total scaleup funding, however, the West-East and North-South divide is more pronounced (see Fig. 6.5). In the southern part of Europe, scaleups receive a relatively high level of total funding only in Barcelona (€5 billion), Madrid (€3.1 billion), and Milan (€2 billion), whereas the funding raised by the scaleups of Rome, Athens, and Lisbon jointly amounts to less than €1 billion (equal to the values of Toulouse or Malmo). Nevertheless, these numbers fall orders of magnitude below those of London (€57 billion), Paris (€22 billion), Berlin (€20 billion), and Stockholm (€13 billion), and also below those of Amsterdam (€7 billion) or Munich (€7 billion). With the West-East divide, the scaleups of Central and Eastern Europe have notable total funding only in Bucharest (€1.9 billion) and Tallinn (€1 billion), whereas they received less than €500 million in major capitals such as Warsaw, Prague, and Budapest, and less than €100 million in Riga, Bratislava, and Ljubljana. The combined total funding of scaleups in the 15 cities of Central and Eastern Europe comes to just about €6 billion, which corresponds to a mere 2.5% of all funding of European scaleups. For comparative purposes once more: This is equivalent to the funding of startups in Dublin alone.

Fig. 6.5

The total funding of scaleups across the European scaleup cities, 2021 (in million €). Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

Trends for Convergence Only in Western and Northern Europe

Despite the almost oligopoly of very few scaleup cities, taking the size of the economy into account somewhat balances the picture (see Fig. 6.6). In terms of funding density—measured as total funding of scaleups (million €) per US$1 billion GDP—Global scaleup cities take the lead: Berlin (€85.9 million), Stockholm (€81.4 million), and London (€67.3 million), whereas Paris (€24.2 million) seems to be an exception. At the top of Europe, however, stand towns with world-class universities, for example, Cambridge with €268.1 million funding per US$1 billion GDP and Oxford with a corresponding value of 190.5. In addition, there are very high funding densities in other university towns, such as in Lausanne (115.7), Basel (55.3), Grenoble (37.1), Malmo (31.3), Geneva (32.9), and Leiden (22.1). Capitals in the Baltic region also excel when it comes to funding of scaleups relative to the size of the municipal economy, as in Tallinn (53.9), Helsinki (50.3), and Vilnius (26.4).

Fig. 6.6

Funding density across the European scaleup cities, 2021 (scaleups total founding (million €) per US$1 billion GDP). Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

Yet Southern Europe’s scaleup cities—with the exception of Barcelona (€22.9 million funding per US$1 billion GDP)—have low funding densities: see Madrid (9.1), Milan (6.9), Athens (4.0), Lisbon (1.8), and Rome (0.9). Similarly, Central and Eastern Europe has only one capital with a noteworthy funding density, Bucharest (34.1), whereas the scaleups receive significantly less funding than one would expect given the size of their economies in other major cities of the region, such as in Prague (3.8), Warsaw (2.4), and Budapest (2.9).

In short, convergence marks the funding of scaleups relative to cities’ economic power in Western and Northern Europe; in smaller university towns particularly, scaleups receive more funding than one would expect given the size of their economies. However, this trend can hardly be observed in the scaleup cities of Central and Eastern Europe and Southern Europe.

Access to Talent in the Scaleup Cities of Europe

Locally Available Talent as a Driving Force Behind the Performance of Scaleup Cities

Turning to my second research question—how access to locally available talent affects the scaleup city landscape of Europe—I have analyzed three variables: the number of startup founders who attended a university in the city; the number of startups created by founders who attended a university in the city; and the number of those founders who attended a university in the city and raised more than €10 million in funding. As for investigating the overall relationship between indicators of performance indicators and access to talent, I applied a linear regression model (y = mx + b) across the entire sample of 166 scaleup cities, designed the regression trend line, applied a 1-percent significance level (a = 0.01), and computed the coefficient of determination (R²). With respect to the number of scaleups in the scaleup cities, all variables of access to talent fit extremely well (R² values from 0.82 to 0.86). Similarly, there is a strong—though somewhat weaker—relationship between the total funding of scaleups and the access to talent (R² values from 0.66 to 0.73), and the performance variable of the number of scaleups valued at more than €200 million also correlates with the access to talent variables at the same level (R² values from 0.67 to 0.73).

Decoupling the Eastern and Southern Parts of the Continent

Given these strong correlations, it is unsurprising that the territorial landscape is again marked by the decoupling of the Eastern and the Southern parts of the continent, and that students create successful startups in the large Western and Northern European cities, such as Paris with more than 2150 founders, London with 1850, followed by Amsterdam, Berlin, Barcelona, and Stockholm with founder numbers ranging between 500 and 700 (see Fig. 6.7). Students in Madrid, Munich, Dublin, Copenhagen, Milan, Utrecht, Helsinki, Rotterdam, Zurich, and Vienna are also active in creating startups, with the numbers of founders who attended the cities’ universities falling between 240 and 430. Traditional university towns have remarkably high values, as Cambridge and Oxford each have close to 500 founders, Malmo almost 200, and Leuven just fewer than 100. However, in capitals of Central and Eastern Europe, such as in Prague, Budapest, and Bucharest, the numbers of startup founders who attended a university in these cities are very low, between 65 and 75, with the exception of Warsaw, which has almost 200.

Fig. 6.7

Number of founders who attended a university in the scaleup city, Europe, 2021. Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

This landscape becomes more diverse if one examines the founder density as measured by the number of founders who attended the universities of scaleup cities per 100,000 inhabitants (see Fig. 6.8). On the one hand, Global scaleup cities such as Stockholm (23.3 founders per 100,000 inhabitants), Paris (16.7), London (14.8), and Berlin (13.2), as well as top European scaleup cities in the Southern part of the continent, such as Barcelona (13.5) and Madrid (6.3), hold rather modest values, whereas the Scandinavian, the Baltic, the Dutch, and the German scaleup cities have higher number of founders than expected based on their population size. In addition, some top European scaleup cities are even ahead of Global scaleup cities in this regard, as is the case for Amsterdam (25.6), Dublin (19.2), Helsinki (18.7), Copenhagen (18.5), and Zurich (18.1).

Fig. 6.8

Founder density of university students across the scaleup cities in Europe, 2021 (number of founders who attended a university in the scaleup city per 100,000 inhabitants of the city). Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

The university towns once again lead Europe, with Cambridge sporting the highest founder density (133.2 founders per 100,000 inhabitants), followed by Oxford (89.4), Maastricht (63.7), and Lausanne (49.8). Even smaller university towns have a relatively high density, as in Leuven (40.1), Aarhus (35.3), Malmo (27.6), Grenoble (24.5), and Leiden (24.3). In contrast, capitals in Central and Eastern Europe have some of the lowest values: Warsaw has 6.0 founders per 100,000 inhabitants, Prague 3.3, Bucharest 3.1, and Budapest 2.1.

Convergence in Scaling Opportunities in Western and Northern Europe

The West-East divide also appears in terms of scaling opportunities, indicated by the variable number of those founders who attended a university in the city and raised more than €10 million in funding (see Fig. 6.9). Paris and London are by far the largest places in Europe for students to scale their startups and the corresponding figures range from 550 to 750. Cambridge, Oxford, and Stockholm offer good opportunities for scaling, with the number of founders who studied in these cities’ universities and received more than €10 million in funding falling between 180 and 200. Munich, Dublin, Barcelona, and Copenhagen are also popular places for growing scaleups, and the values here lie between 110 and 130.

Fig. 6.9

Number of founders attended a university of the scaleup city who raised more than €10 million in funding, Europe, 2021. Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

Yet by examining the scaling rate of the scaleups of student founders—measured as the number of founders who attended a university in the scaleup city and raised more than €10 million in funding compared to the total number of founders who attended a university in the given city—one can see that big cities do not hold the monopoly on pools of university students with entrepreneurial spirit (see Fig. 6.10). In Europe on average, around one in four founders who attended a university in the city raised more than €10 million in funding (26.8%). Almost every second founder in Lausanne and Cambridge and every third founder in Oxford, Zurich, Dublin, Paris, Stockholm, Munich, Copenhagen, and London have been able to grow and scale and raised more than €10 million in funding. Yet it is striking that some large scaleup cities—such as Berlin, Amsterdam, Rotterdam, Utrecht, and The Hague—despite the huge number of founders who have attended a university of these cities, offer very weak and below-average opportunities to scale and grow, as only around one in ten of these founders received more than €10 million in funding.

Fig. 6.10

The scaling rate of the scaleups of student founders in selected European scaleup cities, 2021 (cities with more than 200 founders who attended a university in the scaleup city, scaling rate (%) = the number of founders who attended a university in the scaleup city and received more than €10 million in funding in relation to the total number of founders who attended a university in the scaleup city). Source: Author’s own calculation based on Dealroom data retrieved June 2, 2021 from https://dealroom.co and EU-OECD FUA classification. Design by author

Not only are there significantly fewer scaleup founder students in the capitals of Central and Eastern Europe, but those that there are also struggle with scaling, as only about one in five founders raised more than €10 million in funding. Respectively, the scaling rate is 24% in Bucharest, 20% in Tallinn, 19% in Prague, 17% in Budapest and Bratislava, 15% in Warsaw and Riga, 10% in Vilnius, and 8% in Ljubljana. In other words, when startups founded by university students turn to scaleups in this region, they usually lack the capital and market to growth.

Concluding Remarks

In this study, I have reinforced the widely held claim that startup ecosystems are concentrated in a few hubs and that in Europe only a handful of scaleup cities hold the vast majority of scaleups and funding. However, with detailed analysis I have also revealed deep West-East and North-South divides, with major metropolitan areas in Central and Eastern Europe and Southern Europe lagging far behind in both the number of scaleups and the funding these scaleups have raised. Signs of convergence appear only in Western and Northern Europe, and university cities in particular perform remarkably well with respect to the number and funding of scaleups relative to their population and economic size. This is partly due to the good access to locally available talents that universities can provide, whereas in the scaleup cities of the lagging Central and Eastern European and Southern European region, students’ weak engagement in entrepreneurship hampers the ecosystem’s development.

Moreover, I have shown that the European scaleup city landscape is shaped by some strict rules. Firstly, size matters. Large European cities host not only huge number of scaleups but provide many funding and scaling opportunities. Researchers have long proven that big cities have better conditions for entrepreneurial ecosystems due to economic agglomeration effects triggered by larger population and greater densities. Yet not every large European city can benefit from this. The regional concentration of top universities, startups and scaleups, venture capital, entrepreneurial know-how, and supporting institutions tends to develop first a self-sustaining and then a self-reinforcing system which, after reaching a critical mass, is able to attract investment and talent from all over the world. In Europe, the startup ecosystems in most of Global and top European scaleup cities have reached this critical mass and now their ecosystems appear to be evolving on their own, yet only few have turned into a self-reinforcing system.

Secondly, location matters as well. Size is not the only factor, as long-lasting West-East and North-South development disparities also prevail in the European scaleup city landscape, especially when one compares the performance of the scaleup cities with their population and economic size. In addition, the large cities of Southern and Central and Eastern Europe not only feature significantly fewer scaleups than the Western and Northern parts of the continent, but scaleups in these regions also struggle to access finance and handle scaling and growth. In short, although the concentration of the entrepreneurial ecosystems with strong scaleup performance is the dominant trend, it is one deeply embedded in Europe’s economic and territorial disparities.

Thirdly, knowledge matters too. A high number and ample funding of scaleups as well as good opportunities for scaling and growth are not a prerequisite for large cities, since many smaller towns in Western and Northern Europe can offer them an adequate ecosystem. Towns with world-class universities, in particular, are becoming serious competitors of the big players in the European scaleup city landscape. Although there are undoubtedly many factors influencing the performance of scaleup cities, I have shown that one such determining factor is the upstream stemming from the university students in the cities in question. As creating startups is almost a unique university cultural “genre,” it comes as no surprise that university towns also have the highest values in every respect, be it in terms of number of founders, the amount of funding they raised, or densities relative to population. In contrast, the startup activities of the university students in some large cities are rather modest, whereas the East-West and North-South divide still predominates in this area.

In short, scaleup cities in Southern and Central and Eastern Europe largely lack the upstream of university students, which is partly why their scaleup performances are lagging far behind. University cities, especially in Western and Northern Europe, on the other hand, have very good scaleup performance due to the extremely high level of student engagement in creating startups. This is one reason why one can observe some signs of convergence in their scaleup city landscape. The big scaleup cities are, however, in a unique position. Their size has raised them to a stage where the startup ecosystem becomes a self-sustaining—in a few cases even a self-reinforcing—system. Hence, despite having relatively modest upstream from their own universities, particularly in relation to the size of their population and economy, they can attract talent from all across Europe.