1 Introduction

Entrepreneurship is characterized by high risk and potentially high rewards, prompting many early-stage startups to seek external capital from investors. This capital is essential for business growth and closely tied to survival and future performance (Cassar 2004). Besides, equitable access to funding for entrepreneurs is regarded as essential for economic development (Breuer and Pinkwart 2018; Morazzoni and Sy 2022; Tykvová 2018). Recent studies of the startup ecosystem in Germany indicate a persistently high demand for capital, yet the investment climate is tense (Kollmann et al. 2023).

Angel investors play a vital role in this context, bridging the gap between private capital and formal investors, while offering additional support to startups through their networks and experience (Brettel 2003; Hohl et al. 2021; Schmidt et al. 2018). They have long been considered one of the most critical sources of funding for entrepreneurs (Bygrave and Reynolds 2006). At the same time, the decisions of angel investors are characterized by a high degree of uncertainty, as they must decide which ideas, entrepreneurs, and businesses to invest in. With limited prior investor involvement in the business, they rely heavily on their subjective experience and make inferences from the observable information about the startup (Boulton et al. 2019; Ewens and Townsend 2020). Numerous recent studies underscore that angel investors, compared to larger venture capital firms, emphasize criteria related to the entrepreneurial team (Granz et al. 2020). Yet, making assumptions based on team characteristics renders their decision-making susceptible to systematic error and bias.

Over the past decade, practitioners and researchers have studied investor decision-making processes by analyzing secondary data, conducting surveys, and observing real investor behavior in startup pitch competitions (Boulton et al. 2019; Hohl et al. 2021; Jeffrey et al. 2016; Maxwell et al. 2011; Sohl 2022). Borrowing from social perception theory (Snyder et al. 1977) and stereotyping concepts (van Knippenberg and Dijksterhuis 2000), much of this research explores investor preferences and systematic bias based on observed superficial characteristics of entrepreneurial teams, such as age, gender, or ethnicity. While relatively unexplored in this context, physical attractiveness is assumed to similarly influence investor decisions. Stereotypically driven decisions pose challenges for both investors and startups, resulting in reduced accuracy of investment decisions, suboptimal investment outcomes for investors, and discrimination against individuals and entire groups of entrepreneurs (Boulton et al. 2019; Jetter and Stockley 2023; Morazzoni and Sy 2022; Wickham 2003). Tangibly, recent studies substantiate the systematic misallocation of venture capital to male-led businesses, and removing the gender funding gap would both create equal opportunities for women and significantly increase the overall economic productivity (Morazzoni and Sy 2022). It is thus economically essential to understand the occurrence and determinants of decision biases to address them directly.

While existing research on bias in entrepreneurial finance predominantly focuses on North American angel investors (Balachandra 2020; Boulton et al. 2019; Hohl et al. 2021; Hussain et al. 2023; Poczter and Shapsis 2016), the decision-making of angel investors in other regions remains underexplored. To address this gap, some recent studies analyze angel investor decisions in various European countries such as e.g., the UK, Italy, France, and the Czech Republic (Capizzi et al. 2022; Croce et al. 2017; Färber and Klein 2021; Skalicka et al. 2023). However, they mostly use secondary data or surveys, failing to identify investor bias in relation to superficial characteristics of entrepreneurial teams. This study aims to fill this research gap and complement the European perspectives by exploring the decision criteria and bias of German angel investors in the popular televised startup pitch competition Die Höhle der Löwen.

Previous studies suggest that overall, the investment behavior of German angel investors is comparable to that of angel investors in the US and the UK (Brettel 2002). However, there are some notable differences: They usually allocate smaller portions of their wealth to startup investments (Brettel 2002) and the overall investment volume is smaller, with Germany having 2–3 times fewer seed financing rounds per million citizens compared to the US (Schlüter 2018). Interestingly, due to different economic and venture capital histories, German angel investors are found to be more risk-averse and prefer to find suitable entrepreneurs over identifying innovative ideas (Schlüter 2018; Stedler and Peters 2003), which may further increase their susceptibility to systematic bias based on stereotypes.

Regarding the characteristics of entrepreneurial teams in Germany, diversity remains a significant issue (Hirschfeld et al. 2022; Kollmann et al. 2023). For example, the proportion of female founders in Germany stands at 20.7%, with no recent positive trend. This suggests that women, often having fewer networks with relevant contacts, face greater challenges in the current economic climate, contributing to the existing gender funding gap. Therefore, understanding whether German angel investors exhibit systematic preferences based on representative founder stereotypes is crucial from an economic perspective. Given the evidence that they emphasize team characteristics over business characteristics, we expect them to show the same or even stronger bias compared to previously investigated angel investors in the US and the UK.

To address this question, this study analyzes investment behavior observed across seasons 1–10 of the German televised pitch competition Die Höhle der Löwen to explore the impact of age, gender, ethnicity, and physical attractiveness of entrepreneurial teams on the funding decisions of angel investors. While controlling for various potential factors, our results indicate that the likelihood of receiving offers and securing deals with German angel investors is positively influenced by the age, diverse ethnicity, and physical attractiveness of entrepreneurs. These findings are novel and empirically unique to angel investors in the German setting. Additionally, resulting valuations in deal negotiations are significantly lower for teams with female entrepreneurs and older teams, aligning with prior results from US angel investors.

2 Theoretical background and hypothesis derivation

2.1 Information asymmetries of startup pitch competitions

Private angel investors and venture capital firms are constantly looking to identify the most promising investment opportunities. The capital they provide to startups is typically tied to equity shares in the business. While this arrangement offers the potential for significant returns, it also entails substantial financial risk (Maxwell et al. 2011; Sohl 2022). Moreover, angel investors often provide additional resources, operating experience and network to supported startups, increasing the cost and risk of their investments (Schmidt et al. 2018; van Osnabrugge and Robinson 2000).

The decision architecture in pitch competitions is particularly complex (Maxwell et al. 2011). Within a few moments during and after the pitch of a startup, investors not only must choose whether to invest or not, they also need to decide on the amount of capital they are willing to provide and the share of equity they require in return for their investment. The business valuation resulting from these parameters is a complex function of the expectations about the future performance of the startup (Wickham 2003). Furthermore, equity shares are intertwined with the level of influence in the business, establishing the relative power of investors for future strategic decisions. Importantly, this multistage decision-making process is shaped by considerable information asymmetries between the angel investors and the entrepreneurs (Glücksman 2020; Sohl 2022).

Information asymmetry is a well-established concept in economic and social psychological theory, which can be applied to explain why entrepreneurs have much better knowledge of the status of their own business than investors do (Courtney et al. 2017; Graebner 2009). Angel investors in this setting can only leverage the information from the pitch, draw conclusions from the responses that founders give to their questions, and use their own knowledge of the respective market. As a result, angel investors decide based on the limited information that are available to them, and which they assume to be predictive of future business success (Harrison and Mason 2017). This includes the parameters of the venture mentioned by entrepreneurs, such as the degree of innovation and functioning of the product, growth potential in the market, the financial situation (e.g., past revenue, margins), competition, and protectability (Courtney et al. 2017; Croce et al. 2017). For these criteria, however, objective information is impossible to obtain for investors in the pitch setting, so they have to rely on statements of the entrepreneurs. The information asymmetries present in the pitch setting can thus result in angel investors relying on further signals they observe and, at least partially, make investment decisions based on their subjective judgement and attributions about the characteristics of entrepreneurs and entrepreneurial teams (Courtney et al. 2017; Harrison and Mason 2017; Tarillon et al. 2023).

2.2 Personal characteristics of entrepreneurs as decision criteria for investors

“I invest in people, not in products” is a typical statement from angel investors in the German startup pitch competition Die Höhle der Löwen (Cremer 2022). Prior research corroborates that strong teams of entrepreneurs can compensate for less compelling business aspects (e.g., revenue) and analogously, pitches are rejected more often for reasons related to the characteristics of the founder and management team, and less often for lack of business innovativeness (Croce et al. 2017; Mason et al. 2017; Moritz et al. 2022; Tarillon et al. 2023). Angel investors consistently show a much greater emphasis on the entrepreneurial team than venture capital firms (Granz et al. 2020), and geographic studies underline that the first personal impression of the entrepreneurs is the key decision factors for angel investors in Germany (Brettel 2003; Schlüter 2018; Stedler and Peters 2003). For instance, the personal characteristics of entrepreneurial teams that are presumably linked to venture success include professional experience, presentation skills, and entrepreneurial passion (Clark 2008; Fernández-Vázquez and Álvarez-Delgado 2020; Gielnik et al. 2018; Lavi and Yaniv 2023).

First, prior industry experience and a track record of successfully building startups in the past positively influence investment decisions, given that experienced entrepreneurs are more likely to correctly identify new opportunities for creating wealth and successfully establish businesses (Baron and Ensley 2006; Gielnik et al. 2018; Ucbasaran et al. 2009). Furthermore, relevant professional experience is also associated with having larger professional networks that can be leveraged for partnerships and business development.

Second, effective presentation skills, preparedness, and charisma that are observable during pitches, have a significant positive impact on investor decisions (Clark 2008; Daly and Davy 2016; Huang et al. 2023; Lavi and Yaniv 2023). Presentation skills are presumably relevant to the entrepreneurs’ capability of attracting people, partners, and customers in the future. Interestingly, angel investors are shown to be unaware of, or at least reluctant to acknowledge, the level of impact the presentational skills had on them (Clark 2008).

Third, angel investors value affective entrepreneurial passion and emotional appeal in their decision-making (Fernández-Vázquez and Álvarez-Delgado 2020; Hsu et al. 2014; Warnick et al. 2018). Entrepreneurial passion reflects the degree of intrinsic motivation that entrepreneurs have towards building the business and can be considered a willingness to make additional efforts when suffering setbacks, indicating resilience (Warnick et al. 2018). Interestingly, findings suggest that angel investors emphasize entrepreneurial passion even more than traditional VC investors, while considering the economic potential of the business as being less important (Hsu et al. 2014).

It can thus be considered reasonable for angel investors in this setting to base their decisions on the perceived strength of entrepreneurial teams. However, it remains difficult to deduct objective information about the true skills and personality of the entrepreneurs based on their performance in the startup pitch. As a result of insufficient information as well as time constraints in the context of pitch competitions, investors may leverage decision heuristics related to stereotypes attributed to the more superficial characteristics of entrepreneurial teams (Harrison et al. 2015; Smith et al. 2010).

2.3 Superficial categorization and stereotyping in social perception theory

Decision heuristics are applied to simplify the complex judgmental task of investment decisions into more simple cognitive operations (Holcomb et al. 2009; Jeffrey et al. 2016). Because of this, angel investors may develop preferences based on easily observable characteristics of the entrepreneurs. To better understand why these superficial characteristics, such as age or gender of entrepreneurial teams, can serve as heuristic cues and impact investor decision-making, it is useful to draw on social perception theory (Cook 2021). In the past decades, this area of social psychology and behavioral science has explored the cognitive processes behind social perception and attribution. A significant part of the theory examines the roles of stereotypes and bias which impact the accuracy of interpersonal perception, as well as the resulting behavior towards others (Cook 2021). Specifically, it was established that individuals categorize others based on superficial characteristics, which is now considered an automated cognitive process (van Knippenberg & Dijksterhuis 2000). Some of the earliest empirical research focused on racial categorization (Katz and Braly 1933), and the idea of racial stereotypes was quickly adapted to other observable attributes of individuals. Critically, stereotypes play a crucial role in maintaining the often disadvantaged position of the stereotyped group through self-fulfilling mechanisms (Snyder et al. 1977) and are highly prevalent and influential in all settings of social interaction (Cook 2021; Taylor 1981).

Studies show that superficial personal characteristics such as age, gender, and race are usually the first to be noticed in social interaction and strongly influence all subsequent information processing (Snyder et al. 1977; van Knippenberg and Dijksterhuis 2000). Stereotypes related to these characteristics often affect individuals unconsciously, and decision-makers in the context of startup investments are shown to attribute their influence to intuition and gut feeling (Haines et al. 2003). This is problematic, given that stereotypes can be highly inaccurate and, independent of their general validity, result in biased interactions aiming to confirm the stereotype (Snyder et al. 1977). This observation corresponds to the pitfalls of the representativeness heuristic introduced by Tversky and Kahneman (1974), which suggests that individuals make objectively less accurate decisions when judging individuals whose characteristics are representative to certain subgroups.

2.4 Specific stereotypes and bias of German angel investors: hypothesis derivation

Applied to the observed context of investor decision-making in pitch competitions, social perception theory and stereotyping suggest that investors look for superficial characteristics they deem representative of successful entrepreneurs, even when these characteristics are not directly related to business success. Underscoring this, Wickham (2003) experimented on investors’ appraisal process of startups, including relevant details of the business as well as details of the background of the founders that are irrelevant to the venture’s operations. They found that judgements of future success were less precise overall when decision-makers were presented with information on personal characteristics that are representative of specific founder stereotypes. The results further suggest that angel investors favor ventures with low success probabilities when they are presented together with representative and stereotypical context information regarding the entrepreneurs. This indicates that specific superficial characteristics of entrepreneurial teams compensate for more rationally relevant criteria in their investment decisions.

In practice, this means that angel investor decision-making is often influenced by the demographics or appearance of the entrepreneur, with preferences for stereotypically representative groups, relating to age, gender, ethnicity, and physical attractiveness (Boulton et al. 2019; Brooks et al. 2014; Jetter and Stockley 2023). Assuming that subsequent startup success is not directly driven through these superficial personal characteristics, the use of stereotypes leads to suboptimal decision outcomes for investors and is a significant source for decision bias (Harrison et al. 2015; Morazzoni and Sy 2022). This includes false positive decisions, such as overconfident investments into startups that later fail, which can result in the total loss of the initial investment. Analogously, false negative decisions can occur, leading to missed investment opportunities of startups who turn out to be successful. Clearly, both types of decision error are associated with significant cost for angel investors.

Importantly, systematic bias towards certain demographic groups also creates discrimination in the startup economy. If pitches are rejected based on the representativeness of the entrepreneurs’ demographics, minorities face significantly higher difficulties when seeking funding for their business. In this regard, recent research on the so-called gender funding gap shows that stereotyping and discrimination against female founders is a prevalent issue in many economies, including Germany (Balachandra et al. 2019; Cicchiello and Kazemikhasragh 2022; Henry 2020; Jetter and Stockley 2023; Prokop and Wang 2022).

In the context of startup pitch competitions, the stereotypes relating to personal characteristics of entrepreneurial teams likely not only impact the probability of receiving an offer or making a deal with angel investors, but also influence the resulting deal valuations. Building on the model of Boulton et al. (2019), who analyzed decision bias of US angel investors in the format Shark Tank and found mixed results, our study of German angel investors investigates these biases using a similar multivariate approach by including the entrepreneurs’ age, gender, ethnicity as independent variables. Augmenting the perspective of superficial characteristics further, we also include physical attractiveness as independent variable. Further characteristics of entrepreneurial teams and their pitches, such as professional experience, are used as control variables. In contrast to Jetter and Stockley (2023), who focused exclusively on entrepreneur gender as the main predictor of investor decisions, we believe that the multivariate approach is essential because the impact of these personal characteristics cannot be evaluated independently, considering potential correlations of e.g., age and attractiveness (Granleese and Sayer 2006), or age and experience (Baron and Ensley 2006). Importantly, given the more risk-averse profile of German angel investors and their pronounced emphasis on identifying promising entrepreneurial teams over ideas (Schlüter 2018; Stedler and Peters 2003), as well as a slightly different entrepreneurial demographic (Kollmann et al. 2023), we may find investor bias which was not apparent in studies using US samples (Boulton et al. 2019; Jetter and Stockley 2023; Poczter and Shapsis 2016). In the following sections, we rely on prior research on investor bias and the stereotypes related to the specific superficial characteristics of entrepreneurial teams in Germany to derive each hypothesis.

2.4.1 Age bias

Social perception theory establishes that humans automatically categorize each other based on their perceived chronological age (Nelson 2005). The discrimination of individuals based on stereotypes associated with age is termed ageism, or age bias. Based on this, it is assumed that angel investors differentiate between younger and older entrepreneurs, allowing this superficial characteristic to influence their decision-making, with the hypothesis that they favor teams of younger entrepreneurs over older teams (Boulton et al. 2019). To better understand this mechanism, scholarly attention towards the relationship between age and entrepreneurial behavior as well as entrepreneurial success has recently increased (Gielnik et al. 2018; Zacher et al. 2019; Zhao et al. 2021).

On the one hand, society associates older age with a few positive attributes. Older individuals are stereotypically regarded as wise and knowledgeable (Nelson 2005) and are believed to have more experience as well as larger networks (Bowen et al. 2020). Regarding entrepreneurship, it is assumed that older entrepreneurs have more social and financial capital, which is useful when starting a new business (Weber and Schaper 2004). Parker (2009) further substantiates the link between age and entrepreneurial activity, finding that most entrepreneurs start businesses at middle to higher age, especially for those who want to mainly employ themselves (Harms et al. 2014; Kautonen et al. 2014). Older entrepreneurs with greater experience are better at identifying new venture opportunities, and are more likely to progress from initial intention to the actual formation of a business (Baron and Ensley 2006; Gielnik et al. 2018; Ucbasaran et al. 2009). Moreover, Zhao et al. (2021) show a weak positive relationship between older age and overall venture success. Their meta-analysis reveals that older age is particularly positively related to subjective and financial success as well as firm size.

On the other hand, however, angel investors are assumed to show a preference for younger entrepreneurs relating to stereotypes and the general prejudice towards older adults in society: Young individuals, are generally associated with higher levels of motivation, productivity and persistence, and considered as more energetic and willing to learn (Bowen et al. 2020; Nelson 2005). They may be perceived as more willing to accept the risk and long working hours involved with early-stage entrepreneurship. Older people, in comparison, are overall considered less adaptive and to have less physical and cognitive strength (Nelson 2005). Based on this set of attributes, discrimination of older individuals is a recurring issue in the business context (Nelson 2005; Zhao et al. 2021). In the general business context, ageism manifests in the preference of employers for hiring younger applicants, independent of their individual qualification (Abrams et al. 2016; Kaufmann et al. 2017; Rothermund and Meyer 2009).

Prior research in the context of angel investors suggests that investors may prefer younger entrepreneurs for their presumably higher resilience, their ability to perform under pressure as well as their openness to feedback in a mentoring relationship, and establishes that older teams are less likely to receive an offer from angel investors in the US pitch competition Shark Tank (Boulton et al. 2019). Additionally, the preference for younger teams may be linked to the belief that younger people have a longer time horizon (Gielnik et al. 2018). Compared to older entrepreneurs, younger entrepreneurs are more likely to build a business to employ other people (Kautonen et al. 2014). This idea suggests they have more time left to invest in the growth and long-term success of their venture and is supported by the findings of Zhao et al. (2021), who show a positive link between younger age and venture growth in their meta-analysis. Finally, in Germany, the average entrepreneur is 36.7 years old, which is considerably younger than the average of the general German working population at 43.3 (Kollmann et al. 2023). These two figures highlight why investors’ stereotypical image of an entrepreneur in Germany would be younger rather than older. Based on the prior research on the general well-documented stereotypes associated with age, and the skewed age distribution among German entrepreneurs, investors in the German sample may hence overlook the value of experience and wisdom that older entrepreneurs can bring and are expected to prefer younger entrepreneurs.

H1. Angel investors in Germany show a preference towards (teams of) younger entrepreneurs.

H1a. Younger entrepreneurs are more likely to receive investment offers from angel investors.

H1b. Younger entrepreneurs are more likely to reach a deal with angel investors.

H1c. In case of a deal, younger entrepreneurs obtain higher valuations for their business than older entrepreneurs.

2.4.2 Gender bias

Automatic gender categorization is another well-established mechanism in social perception theory (van Knippenberg and Dijksterhuis 2000). Accordingly, angel investors are assumed to differentiate their investment decisions based on gender of the entrepreneurial teams, and specifically to show a preference for male entrepreneurs. This is linked to male gender being more representative in the (global as well as German) entrepreneurship ecosystem (Hirschfeld et al. 2022), and established by traits associated with gender stereotypes. For instance, some traits that are beneficial for entrepreneurial activity, such as risk-taking, boldness, and aggressiveness, are traditionally more associated with men (Baughn et al. 2006). Female entrepreneurs, in comparison, are be assumed to have additional responsibilities not directly related to the business due to gender stereotypical roles (Baughn et al. 2006). Signaling an entrepreneurial attitude to potential investors resulted in lower funding for female entrepreneurs, while the same attitude proved to be beneficial for men (Malmström et al. 2020). Prior research also underscores the stereotypes that women are judged as less able to lead a startup (Edelman et al. 2018) and less eligible to partner with for a business venture (Balachandra et al. 2019; Lim and Suh 2019). Furthermore, investors may gauge women’s ventures to be less likely to succeed in the market due to potential discrimination and bias from customers and suppliers (Bates 2002). Ultimately, business pitches by female entrepreneurs are considered to be less persuasive, and in interaction, women are asked more risk-related questions, while men are asked more potential-related questions (Balachandra et al. 2019; Brooks et al. 2014). Notably, recent studies show that the bias against female entrepreneurs is shown to be particularly strong for socially attributed characteristics (Pistilli et al. 2023), when angel investors are politically conservative (Chen et al. 2023), and when objective information about prior experiences and qualifications is not available (Tinkler et al. 2015), which is often the case in the context of televised pitch competitions.

As a result, female entrepreneurs are demonstrably less likely to receive venture funding compared to male entrepreneurs (Färber and Klein 2021; Greene et al. 2001), which results in further underrepresentation of women as entrepreneurs. In Germany, current data suggests only 20.7% of individual entrepreneurs are female, and the majority (61.3%) of entrepreneurial teams are male-only (Kollmann et al. 2023). Bias against female entrepreneurs is well-documented in research on applications for bank loans (Muravyev et al. 2009), equity-based crowdfunding (Prokop and Wang 2022), and investments from angel investors in startup pitch competitions. While analyses on offer and deal likelihood in these competitions yield mixed results, teams with female entrepreneurs consistently receive less funding and lower valuations than male entrepreneurs (Boulton et al. 2019; Hohl et al. 2021; Hussain et al. 2023; Jetter and Stockley 2023; Poczter and Shapsis 2016).

Some argue that this gender funding gap is self-imposed: For instance, women are assumed to enter less attractive business sectors or have a lower propensity to seek external capital in the first place (Becker-Blease and Sohl 2007; Poczter and Shapsis 2016). However, recent research indicates that the funding gap is not proportional to the entrepreneurial activity of genders and cannot be explained by differences in other venture-related characteristics (Henry 2020). Comparatively, the gender funding gap is more pronounced in Germany than in the US across entrepreneurial activity, investment frequency, and total invested capital percentages. Out of all nine countries analyzed in the study, Germany holds the largest gender gap in average deal size, with women-led businesses in Germany receiving the smallest proportion of total deals and invested capital (Henry 2020). Thus, while studies on gender bias among US investors yield mixed results (Boulton et al. 2019; Hohl et al. 2021; Jetter and Stockley 2023), it is expected that German investors exhibit gender bias with a preference for male entrepreneurial teams.

H2. Angel investors in Germany show a preference towards (teams of) male entrepreneurs.

H2a. Male entrepreneurs are more likely to receive investment offers from angel investors.

H2b. Male entrepreneurs are more likely to reach a deal with angel investors.

H2c. In case of a deal, male entrepreneurs obtain higher valuations for their business than female entrepreneurs and mixed teams.

2.4.3 Ethnicity bias

Research in the context of social perception theory has established that individuals automatically notice and categorize the ethnic background of other individuals (Snyder et al. 1977). Specifically, evidence suggests generally negative judgement towards ethnic out-groups, referring to individuals with ethnic backgrounds that are different from one’s own (Maddox 2004; Snyder et al. 1977). Accordingly, it is assumed that angel investors’ decisions are impacted by a bias against ethnically diverse entrepreneurs.

Discrimination of ethnic minorities is linked to racism and associated stereotypes, affecting perceptions both explicitly and implicitly (Snyder et al. 1977). Within the entrepreneurial economy, ethnicity bias is evident in the underrepresentation of ethnic minorities (Parker 2009). There is evidence that minority founders face discrimination from banks when applying for credit and loans (Blanchard et al. 2008), and have significantly lower chances of receiving crowdfunding (Younkin and Kuppuswamy 2018). In Germany, robust evidence exists for ethnic bias in hiring practices, particularly against applicants of Turkish decent (Kaas and Manger 2010).

It is conceivable that angel investors in pitch competitions show similar discriminatory behavior. This is especially critical for angel investor panels of non-diverse backgrounds, because they perceive minority founders as the ethnic out-group (Maxwell 2011). Bias against minority founders may stem from negative stereotypes, such as perceived weaker political skills among non-native speakers, resulting in lower likelihood and amounts of funding (Huang et al. 2013), and lower rates of business survival (Li and Johansen 2021). Existing positive stereotypes relate to the perceived stronger work ethic of minority groups (Hsin and Xie 2014), but these do not appear to counterbalance the existing negative ethnicity bias. Specifically, analysis of US angel investor behavior in Shark Tank indicates that black entrepreneurs receive fewer offers and lower valuations than others (Boulton et al. 2019). In Germany, results from the recent nationwide report of minority founders indicate that startups led by ethnic minority entrepreneurs face significant challenges to realize their ambitions related to venture capital funding and encounter disadvantages in debt financing (Hirschfeld et al. 2023).

H3. Angel investors in Germany show a preference for (teams of) ethically non-diverse entrepreneurs.

H3a. Ethnic minority entrepreneurs are less likely to receive investment offers from angel investors.

H3b. Ethnic minority entrepreneurs are less likely to reach a deal with angel investors.

H3c. In case of a deal, ethnic minority entrepreneurs obtain lower valuations for their business than entrepreneurs with non-diverse backgrounds.

2.4.4 Attractiveness bias

The fourth variable to explore in this context is physical attractiveness, with the assumption that angel investors favor for more physically attractive entrepreneurial teams. Alongside age, gender, and ethnicity, physical appearance is a key characteristics susceptible to stereotyping according to social perception theory (Dion et al. 1972). Research in social psychology has established a strong universal consensus on what constitutes physical attractiveness (such as symmetric faces with large eyes, small noses, prominent cheekbones, large smiles, and full hair) which shows high consistency between raters across genders and cultures (Cunningham et al. 1995; Dion et al. 1972). Correspondingly, there is robust evidence that humans treat others differently based on their physical appearance, known as lookism, or attractiveness bias: Overall, higher levels of attractiveness are stereotypically associated with more positive characteristics (Dion et al. 1972; Eagly et al. 1991; Lorenzo et al. 2010; Rosar et al. 2014; Snyder et al. 1977). In interpersonal perception, higher attractiveness is related to socially desirable traits as well as personal skills, such as general intelligence, competence, work ethic and professional success, as well as overall happiness. A recent study also highlighted that more moral traits are attributed to attractive individuals (Klebl et al. 2022). This general positive attractiveness bias applies universally for both female and male individuals (Dion et al. 1972).

In a professional setting, facial appearance and attractiveness positively influence the likelihood of being invited to job interviews (Kaufmann et al. 2017; Watkins and Johnston 2000) as well as being promoted (Heilman and Saruwatari 1979). Moreover, recent research indicates a robust “beauty premium” in terms of earnings for men and women in the German labor market (Hellyer et al. 2023). There is also some evidence for a negative stereotype related to attractiveness, with highly attractive women sometimes perceived as less competent for leadership positions (Heilman and Saruwatari 1979) and attractive women deliberately “dressing down” in the workplace to be taken more seriously (Granleese and Sayer 2006). However, more recent studies show the impact of this negative stereotype of attractiveness to be inconsistent (Braun et al. 2012; Johnson et al. 2010).

In the realm of entrepreneurial finance, there is still a substantial research gap regarding the impact of attractiveness. Initial evidence from Brooks et al. (2014) suggests a systematic attractiveness bias in pitch settings: Attractive male entrepreneurs were rated as significantly more persuasive, even when the content of the pitch was exactly the same. Interestingly, this effect was not found for female entrepreneurs. Addressing attractiveness, Smith and Viceisza (2018) included app-based attractiveness measures of entrepreneurs as a control variable in their analysis of Shark Tank pitches but did not report its impact on deal outcomes. Evidence from televised startup pitch competitions in the US indicates a positive correlation between physical attractiveness, attributed likeability, and overall funding success (Huang et al. 2023). Moreover, a recent experimental study found that male investors assessed prerecorded pitches more positively when delivered by a more attractive women, even though the content and form of the pitch is identical (Schreiber et al. 2024). Notably, this preference was linked to an increase in cortisol levels. Assuming the appearance of entrepreneurs is not causally related to the quality of the idea or its future success, any unequal treatment based on physical attractiveness can be considered the result of a bias (Brooks et al. 2014; Huang et al. 2023). Considering the emphasis of German angel investors on personal characteristics of entrepreneurs (Schlüter 2018), we thus assume that they also show an attractiveness bias, favoring more attractive entrepreneurial teams.

H4. Angel investors in Germany show a preference for (teams of) physically attractive entrepreneurs.

H4a. Physically more attractive entrepreneurs are more likely to receive investment offers from angel investors.

H4b. Physically more attractive entrepreneurs are more likely to reach a deal with angel investors.

H4c. In case of a deal, physically more attractive entrepreneurs receive higher valuations than physically less attractive entrepreneurs.

3 Method

3.1 Approach: observing televised startup pitch competitions

This study leverages observations of business angel decision-making in televised startup pitch competitions; most famously known are the formats Shark Tank and Dragons’ Den. These formats have been adapted globally and all share the same premise: real entrepreneurs pitch their businesses to investors, who then discuss investments of their personal capital. These formats are statistically and economically significant (Boulton et al. 2019) and provide a rich resource to extend existing entrepreneurial finance research, which often relies on survey data prone to response bias.

Startup pitch competitions offer unique laboratories for investigating the impact of founder characteristics on angel investors’ decision-making. They provide information about the entrepreneurs’ and startups’ backgrounds from an external perspective (Pollack et al. 2012), and include interactions and valuations from the investors’ perspective when deals are made (Sherk et al. 2019). Due to their rich data and wide viewership, these formats have garnered significant attention from entrepreneurship scholars over the past decade. Researchers such as Maxwell, Jeffrey and Levesque have coded episodes from Shark Tank and Dragons’ Den to analyze specific decision heuristics of angel investors (Jeffrey et al. 2016; Maxwell 2011; Maxwell et al. 2011; Maxwell and Lévesque 2014). However, particular concerns regarding internal and external validity must be addressed.


Sample selection bias. Both the angel investors and entrepreneurs appearing on these shows are selected by the production company, which pre-screens startups to ensure only the most promising ones get to pitch. This selection process must be considered when generalizing the results to the broader startup ecosystem. Furthermore, since business valuations are only available for startups that secure a deal, there is potential for selection bias in the empirical findings. This study addresses this concern using appropriate econometric approaches and reporting Heckman-adjusted measures. On the positive side, the pre-selection can reduce variance in the dataset, as startups in the competition are at similar development stages—most have formed a business and created a product but are still early enough to qualify for angel investments. This allows for better comparison of decision-making across startups in the sample.


Unobserved decision criteria. The observed interactions, typically around 20 min per startup, are edited versions of the entire interaction between entrepreneurs and investors, with some material cut in post-production (Cremer 2022). Thus, investor decisions might be influenced by criteria not observable in the aired version, potentially affecting the validity of statistical estimates. However, investors and production firm members have confirmed that crucial moments are always included and relevant aspects of the interaction are never cut (Cremer 2022). Additionally, the angel investors receive no prior information on the startups before production, ensuring that their decisions are based solely on the pitch.


Investor motivation and social desirability. In decision-making research, it is essential that individuals are motivated to make the right choice given the information available. Compared to traditional experimental settings (e.g., with university students as subjects), the reality and external validity of the observed interactions in these formats are high. Investors use their own money to make real investments in real companies, financially independent of the show’s production company (Cremer 2022; Maxwell 2011; Poczter and Shapsis 2016; Smith and Viceisza 2018). Thus, angel investors are financially motivated to make optimal decisions and invest in promising startups with high market potential and future returns. Production firms confirm that interactions are unscripted and that investors are not asked to make investment choices for entertainment purposes (Maxwell 2011; Maxwell et al. 2011; Poczter and Shapsis 2016; Pollack et al. 2012). Notably, the televised nature of the setting might reduce the influence of stereotypes and bias regarding entrepreneurs’ characteristics given investors’ awareness of social desirability and the audiences’ perception of discrimination. However, given that they invest their personal capital, they are still motivated to make optimal decisions. Furthermore, bias might be even stronger in private VC setting, where pitches and negotiations occur behind closed doors.


Entrepreneur motivation. This study assumes that startups pitching in the competition seek capital and intend to secure a deal with angel investors. The televised context and the interaction with famous angel investors may further attract entrepreneurs who seek publicity and advertisement for their product without intending to surrender equity (Blaseg and Hornuf 2024). However, all applicants invest significant effort into the pitch and seek positive feedback. Strategic partnerships with investors increase the chance of receiving subsequent capital and achieving long-term success (Kerr et al. 2014; Maxwell 2011). Therefore, the share of entrepreneurs not wanting a deal is considered small. To address this, our analysis includes the offers extended by angel investors that did not result in a deal.


Subjective observer bias. When leveraging video material for research, best practice involves employing third-party observers to watch the interaction and code measures based on pre-defined criteria (Poczter and Shapsis 2016). While some variables are unambiguous, such as the pitch outcome (deal vs. no deal) and agreed-upon investment amount, variables like physical attractiveness and ethnicity are more subjective. For instance, untrained observers may code an individual as attractive, which may not align with the investors’ perceptions. This subjective observer bias can impact measurement validity and pose ethical challenges, potentially reinforcing stereotypes. It is important to acknowledge this as a main limitation of the research approach and mitigate these issues by employing multiple observers and defining clear, standardized criteria to reduce subjective variation.

3.2 Sample

This study examines business angel decision-making in the startup pitch competition Die Höhle der Löwen (German for “The Lions’ Den”; abbr. DHDL). DHDL is the official German adaptation of the formats Shark Tank and Dragons’ Den. It continuously attracts attention from the German startup scene (Cremer 2022) and is considered very popular with a broad viewership. The total observed sample consists of investment decisions on N = 553 startups, i.e., all startups that appeared on the 100 episodes airing between August 2014 and October 2021 (seasons 1–10). Video material of all pitches was made available by the production firm and is publicly accessible through online streaming. The material covers a total of 183 h of footage, equating to approximately 20 min per startup, including pitch, interaction, and investor decisions. The authenticity of startups that appeared and investment deals made on the show were confirmed by members of the production team.

3.3 Procedure

A team of four independent observers was recruited for coding. All observers were fluent in German and research assistants at a German university. The team was trained to watch and code all startup pitches, interactions, and business angel decision-making according to a predefined coding scheme.

The coding scheme was developed to include variables relevant to our hypotheses and was tested and refined with the team over a period of two weeks. As in previous studies of televised pitch competitions (Maxwell 2011), the order of the variables in the coding sheet was set to reflect the order of the entire interaction. This approach allowed the coding of all variables in real-time and before knowing the outcomes, thus eliminating hindsight bias and increasing the objectivity of the resulting dataset. For transparency, the coding sheet is included in Appendix A1. Following the best-practice approach from previous studies (Poczter and Shapsis 2016), every single pitch was double-coded by at least two observers. This method reduces the probability of coding errors and increases the objectivity of rating variables. Subsequently, individual coding sheets were aggregated into one large master dataset for analysis. All statistical analyses were performed using Stata 18 statistics software.

3.4 Measures

3.4.1 Independent variables

To reflect the personal characteristics of every startups’ entrepreneurial team, the proposed independent variables—age, gender, ethnic diversity, and physical attractiveness—were coded at the individual level for every single entrepreneur present in the pitch. Subsequently, these individual variables were aggregated for entrepreneurial teams to create team-level measures. The specific coding process for each variable is detailed below.


Age. The chronological age of each individual entrepreneur present at the pitch was coded as a continuous variable. Entrepreneurs’ ages were displayed to the audience at the introduction of each startup. To allow comparison of single entrepreneurs and entrepreneurial teams of varying sizes, the arithmetic mean of entrepreneurs’ ages was computed per pitch, resulting in the aggregated variable team age. The consideration of age on a team level is based on prior research analyzing startup pitches (Boulton et al. 2019).


Gender. Gender of each individual entrepreneur present at the pitch was coded by observers as a dichotomous variable (male, female) based on self-presentation. This allows analysis of all-male, all-female, and mixed teams with varying shares of female members. Subsequently, following prior research on gender funding gaps in televised startup pitches (Hohl et al. 2021; Poczter and Shapsis 2016), and to align sizes of compared groups, entrepreneurial teams were considered female on the team-level if they included at least one female entrepreneur.


Ethnic diversity. Observers were trained to rate a team as ethnically diverse if at least one entrepreneur identified to have a diverse ethnic background. For this study, ethnic diversity was coded as a dichotomous variable. In the German setting of the pitch competition, ethnic diversity refers to e.g., Arab, Black, Latino, Asian, and overall international backgrounds of entrepreneurs from a German perspective (Boulton et al. 2019). On this note, the investor panel in the observed seasons consists of entirely ethnically non-diverse individuals, with the only exception of Vural Öger, who is of Turkish descent and appeared in seasons 1 and 2. German-speaking entrepreneurs with Austrian or Swiss backgrounds were not considered ethnically diverse. Consistency analysis confirmed nearly perfect interrater agreement for ethnicity ratings (Cohen’s κ = 0.96).


Physical attractiveness. All observers rated the physical attractiveness of each individual entrepreneur considering face, body, and overall appearance in the pitch. As in prior research on the impact of attractiveness (Mueser et al. 1984; Parks and Kennedy 2007), a Likert scale from 1 (very unattractive) to 10 (very attractive) was used. Based on the literature and adhering to ethical research practices, clear criteria were established in the coding team as to which attributes to consider (see coding sheet in Appendix A1), mitigating the impact of subjective observer bias. Second, to increase objectivity of the results, the average estimate of all raters was computed. Prior research found a strong universal consensus regarding what is perceived as attractive (Cunningham et al. 1995; Dion et al. 1972; Reis et al. 1980). Correspondingly, ICC analysis (Shrout and Fleiss 1979) of the coding sheets confirmed average interrater correlations of 0.71 for attractiveness ratings, which can be considered substantial for ratings of social evaluation (Dion et al. 1972; Koo and Li 2016). For comparison of teams, the arithmetic mean of the entrepreneur’s physical attractiveness ratings was derived per team, resulting in the variable team physical attractiveness.

3.4.2 Dependent variables


Offer. After coding characteristics of the pitch and interaction, observers coded if an offer to invest was made by at least one of the angel investors (Offer = 1, No offer = 0). Investors were able to make offers that differed from the request in terms of equity and investment amounts. Not all offers from investors resulted in a deal with the entrepreneurs.


Deal. If the offer resulted in a deal, it was recorded as another binary variable (Deal = 1, No deal = 0). The deal made on production day between investors and entrepreneurs is considered a declaration of intent. It is not legally binding and conditional upon a due diligence process. However, to reflect investor decision-making based on the observable criteria, this research focused on the deals made in the show as main outcome variable.


Deal valuation. For every deal made between angel investors and entrepreneurs, observers recorded the agreed-upon investment amount and equity that would be transferred in return for the investment. Resulting from this, a proxy for the valuation of the business can be derived based on the recent deal via extrapolation. For example, if the deal stipulates 150,000 EUR of investment for 15% equity, the valuation resulting from the deal is 1 million EUR.

3.4.3 Control variables

Additionally, other characteristics of the entrepreneurs, the business, and the pitch are considered and recorded which may be correlated with both independent and dependent variables. If not included, the estimated models may suffer from omitted variable bias. Thus, to mitigate this issue, the following control variables—that had already been used in prior research in the context of televised pitch competitions—were included in the empirical analysis. Consistency checks confirmed satisfactory inter-rater-reliability for all interval-scaled measures.


Team size. In DHDL, some startups are presented by single entrepreneurs while others are pitched by teams of up to five entrepreneurs. Team size is thus included as a continuous variable in all analyses to account for the impact of multiple people presenting the product and the potentially stronger position in the negotiation of deal valuation. Prior research assumes that angel investors have a preference for teams with multiple entrepreneurs over single entrepreneurs, considering that larger teams have more diverse skillsets and a stronger layer of social control within the team (Boulton et al. 2019; Nuñez 2015). Evidence from the context of televised pitch competitions shows that team size is not indicative of offer probabilities but positively linked to resulting deal valuations (Poczter and Shapsis 2016).


Relevant professional expertise. A measure was developed to control for the impact of the entrepreneurs’ professional expertise relevant to the product or industry of the business. Prior research suggests that relevant professional expertise and the track record of past entrepreneurial success is a relevant decision factor for angel investors (Cairnes 2016; Maxwell 2011; Maxwell et al. 2011; Moritz et al. 2022), and expertise is correlated with age of the entrepreneur (Baron 2009; Boulton et al. 2019). Observers rated the level of expertise on a scale from 1 (no relevant professional expertise) to 3 (significant expertise).


Presentation skills. In the pitch competition setting, presentation skills, including entrepreneurial charisma and rhetoric strategies influence the first impressions of angel investors and are considered to impact the funding success (Clark 2008; Daly and Davy 2016; Huang et al. 2023; Lavi and Yaniv 2023). Plus, presentation skills are sometimes associated with gender (Balachandra et al. 2019). Thus, a measure for presentation skills of the entrepreneurs was included as a control variable. For each pitch, observers rated the team’s presentation skills on a Likert scale from 1 (low) to 5 (high).


Entrepreneurial passion. Emotional appeal and storytelling of the pitch, communicated passion and personal enthusiasm for the project have been found to impact the business success and decision-making of angel investors (Breugst et al. 2012; Cardon et al. 2009; Warnick et al. 2018). As entrepreneurial passion is sometimes associated with female gender (Balachandra et al. 2019), or diverse ethnic backgrounds (Hsin and Xie 2014), it was included as a control variable. For each pitch, observers rated passion on a Likert scale from 1 (low) to 5 (high).


Entertainment value. Given the televised setting of the pitch competition, it is possible that some startups in the sample were selected by the production company based not on their quality but for entertainment value of the product or the pitch. This could result in e.g., the presentation of particularly ridiculous concepts or polarizing products, asking the investors to make a fool of themselves trying out a product for show appeal, focusing on heartbreaking personal background stories or pitching together with young children or animals banking on the cuteness factor (Poczter and Shapsis 2016; Smith and Viceisza 2018). To mitigate this effect and a potential selection bias related to the televised setting, a corresponding measure was included as a control variable. Observers rated the entertainment value of each pitch on a Likert scale from 1 (low) to 5 (high).


Business model. The televised setting of the pitch competition and the potential for startups to present their products directly to end-consumers attracts a larger number of businesses with a business-to-consumer (B2C) model compared to business-to-business (B2B) models (Cremer 2022), limiting the representativeness of the sample to the overall startup scene in Germany. Additionally, given that angel investors in this sample may have a general preference for one of the business models, in this case presumably B2C based on the televised setting, the difference needs to be accounted for in the analysis. Thus, prior research includes the B2C vs. B2B categorization into investor decision-making models (Antretter et al. 2020; Blohm et al. 2022). Based on the description of the product and targeted industry, all startups were hence categorized into B2C or B2B models to account for selection bias of the sample and investor preferences.


Season number. As in previous analysis of televised pitch competitions (Blaseg and Hornuf 2024; Boulton et al. 2019; Smith and Viceisza 2018) the number of the season the pitch aired in was included to control for the varying macroeconomic conditions over time and to account for different investor group compositions throughout the seasons. Additionally, it is a way to filter out learning effects of entrepreneurs and angel investors from following the show over multiple years (Poczter and Shapsis 2016).

4 Empirical findings

4.1 Summary statistics

In the observed ten seasons of DHDL, a total number of N = 553 startups were pitched by 994 entrepreneurs. The average team size was M = 1.80 (SD = 0.76), ranging from 1 to 5 members. Notably, 206 startups (37.25%) were pitched by single entrepreneurs. On average, entrepreneurial teams were 36.41 years old. Regarding gender and ethnic diversity, 211 (38.15%) teams included at least one female, and 154 (27.85%) teams included a member with an ethnically diverse background. Table 1 provides an overview of all founder and pitch characteristics included in the analysis. A correlation table of these variables is included in Appendix A2.

Table 1 Descriptive statistics for team and pitch characteristics of DHDL pitches in seasons 1–10
Full size table

Overall, 59.31% (328) of startups received an offer from at least one of the angel investors after the pitch, and 52.08% secured a deal. For those that secured a deal, the average investment was M = 196,677 EUR (SD = 190,421 EUR) for an average equity share of M = 26.15% (SD = 9.95%). Average of the individual resulting deal valuations is M = 929,900 EUR (SD = 1,196,390 EUR). The average relative bid-ask-spread for startups that secured a deal was M = -31.53% (SD = 25.89%), referring to an average absolute discount of -590,251 EUR in valuation. An overview of all deal characteristics is included below in Table 2.

Table 2 Descriptive statistics for deal characteristics of DHDL pitches in seasons 1–10
Full size table

4.2 Multivariate analysis and hypothesis testing

The regressions below demonstrate the impact of the entrepreneurial team’s personal characteristics on the outcomes of the pitch. Considering possible dependencies and correlations of individual variables (e.g., age and attractiveness, or age and experience), this multivariate analysis is crucial for understanding angel investor bias along our individual hypotheses. A visualization of the research model with observed effects is included in Appendix A3.

4.2.1 Founder-related bias regarding offer and deal likelihood

To test the first and second set of hypotheses regarding the impact of personal characteristics on the likelihood of angel investors extending an offer and agreeing on a deal, logistic regression models were estimated (Table 3). The logistic regression estimates the probability of a binary event based on a given set of independent variables, assuming linear relationships. The results suggest that, in contrast to our hypotheses (H1a, H1b), German angel investors prefer to make investment deals with teams of older rather than with younger entrepreneurs, z = 2.96, p < 0.01. Furthermore, female gender has no effect on offer or deal likelihood (H2a, H2b). The results also suggest a slight preference towards making deals with ethnically diverse teams, which contradicts the original hypothesis (H3b). Lastly, the likelihood to receive offers and reach deals is strongly influenced by the physical attractiveness of entrepreneurs, z = 2.27, p < 0.05 for offers (H4a), and z = 2.34, p < 0.05 for deals (H4b), respectively.

Table 3 Logistic regression of offer and deal likelihood on characteristics of entrepreneurial teams
Full size table

Notably, of all included variables, the entrepreneurs’ presentation skills have the strongest impact on securing a deal with the German angel investors. In addition, the positive impact of season number indicates that, overall, more deals are made towards to later seasons of DHDL.

4.2.2 Founder-related bias regarding deal valuations

For all entrepreneurs that secure a deal with the angel investors, the resulting valuations indicate the investors’ degree of confidence regarding potential future success of the startup. To test the third set of hypotheses, the impact of personal characteristics on deal valuations is examined using OLS (ordinary least squares) regression models. This is a common approach to explore the significance of predictors and appropriate in this case based on the size of the cross-sectional dataset and given that we assume linear relationships between the interval-scaled criterion deal valuation and the set of predictors. Furthermore, the adequacy of models was determined checking for homoscedasticity and multicollinearity. A skewness and kurtosis test for normality indicates a significantly skewed distribution of absolute deal valuation, Χ2(2) = 211.48, p < 0.01. Thus, prior to the estimation of the second regression model, the dependent variable deal valuation was log-transformed (ln[x + 1]) to adjust for skewness. Results indicate that log-transformed measures can be considered normally distributed, Χ2(2) = 5.67, p > 0.05.

Table 4 includes both the regression of absolute deal valuation for the sake of interpretability of coefficients (1), as well as the regression of log-transformed deal valuation (2). All following interpretations of individual coefficients and their respective significance levels are based on model 2.

Table 4 OLS regression of resulting deal valuation on characteristics of entrepreneurial teams
Full size table

The results suggest that the age of entrepreneurial teams negatively influences the valuation that German angel investors are willing to agree on in case of a deal, i.e., that younger entrepreneurs receive significantly higher valuations (H1c), t = 2.79, p < 0.01. Moreover, the previously hypothesized gender bias of investors is revealed relating to deal valuation: Teams that include a female member receive significantly lower valuations from the angel investors than male-only teams (H2c), t = 1.97, p < 0.05. In contrast to the remaining hypotheses, further results suggest that the angel investors are not biased towards ethnically diverse (H3c) or more attractive entrepreneurs (H4c) regarding their valuation of the business.

Besides, the further results indicate that relevant professional expertise, presentation skills and the size of the entrepreneurial team positively influence deal valuations in the observed sample. Notably, deal valuations are not influenced by the degree of entrepreneurial passion, entertainment value, B2B business model or season of the pitch.


Robustness checks. First, we address potential multicollinearity in all regression models, which can lead to false positive results. The variance inflation factors (VIF) for all variables in the regression models average at VIF = 1.18, with the highest VIF for physical attractiveness being 1.37. This indicates absence of multicollinearity. Additionally, we followed Kalnins’ (2018) suggestion to check the absolute correlations among independent variables, none of which exceed ± 0.3 (correlation table included in Appendix A2). This suggests that correlation between regressors due to unobservable common factors and resulting Type I error (false positive) from multicollinearity are not major concern for the validity of our results.

Second, the subsample analyzed for deal valuations excludes startups that did not secure a deal with investors. These cases are likely not missing at random, indicating a presence of selection bias in this analysis. A likelihood ratio test for independence suggests that deal valuation equations are not independent from the overall probability of securing a deal, Χ2(1) = 0.04, p = 0.83. Thus, estimates are adjusted using a Heckman selection model, with a first-stage regression of deal likelihood on the same variables followed by a regression of deal valuations as a second stage (model 3). Comparing both regression results suggests that the significant effects from the unadjusted model for the hypothesized personal characteristics remain robust when including a first-stage selection equation. Notably, the difference in significance of presentation skills indicates that the positive effect of presentation skills on deal valuation in the unadjusted model is linked to the selection into the subsample of deals in the first stage.

Third, alternative model specifications were estimated (displayed in Appendices A4, A5, and A6, respectively) to mitigate omitted variable bias and ensure that the significance levels of hypothesized independent variables were not dependent on the set of control variables included in the model. Initially, following previous research (Boulton et al. 2019), deal valuations were regressed with bootstrapped standard errors to demonstrate robustness independent of distributional assumptions. We also tested our hypotheses with regression models that either excluded the hypothesized main predictors or omitted control variables (such as presentation skills). Results across all alternative specifications consistently support the direction and significance of coefficients. However, as anticipated, excluding these variables results in poorer model fits compared to the full models.

5 Discussion

This study contributes to the existing research with a nuanced perspective on angel investor decision-making in Germany. By observing a substantial sample of startup pitches and investor negotiations in the format Die Höhle der Löwen, it provides deeper insights into funding decisions. Descriptive statistics show that the demographic profile of the observed early-stage startup entrepreneurs aligns closely with that of the broader German startup founders population, considering age and gender composition of entrepreneurial teams (Kollmann et al. 2023). Inferential statistical analyses further underscore specific preferences among angel investors in Die Höhle der Löwen tied to stereotypes about superficial personal characteristics of entrepreneurial teams (Table 5).

Table 5 Overview of empirical findings by hypothesis
Full size table

5.1 Key findings and contextualization

In terms of age, the results indicate that German angel investors prefer older entrepreneurial teams over younger ones when extending offers and making deals. This contrasts with the initial hypothesis that investors generally favor younger entrepreneurs across all dependent variables (Table 5). Interestingly, it also diverges from the results of Boulton et al. (2019), who found that older entrepreneurs are significantly less likely to receive offers in the US format Shark Tank. The comparison suggests German investors’ age bias reflects distinct entrepreneurial stereotypes which are different from those of American angel investors. For instance, the preference for making deals with older entrepreneurial teams may be linked to the assumption that older entrepreneurs are better at identifying successful venture opportunities (Gielnik et al. 2018; Zhao et al. 2021) and that investors value the experience and wisdom that is stereotypically associated with older age (Nelson 2005). This finding is significant given increasing prevalence of “gray entrepreneurship”, where older individuals pursue self-employment (Harms et al. 2014).

Conversely, the analysis of resulting deal valuations highlights a nuanced perspective among German angel investors. As hypothesized, they are inclined to offer higher valuations to younger entrepreneurial teams. This preference likely stems from the perceived longer runway for growth and longer time horizon to achieve venture success (Gielnik et al. 2018; Kautonen et al. 2014; Zhao et al. 2021). Moreover, given that younger entrepreneurial teams have more energy to grow the business than older teams, it is plausible for angel investors to accept the higher risk involved with larger deal valuations. This is a novel finding, as Boulton et al. (2019) did not find age to significantly impact deal valuations in the sample of American angel investors, indicating a unique age bias towards younger entrepreneurs of German angel investors.

With respect to gender, German angel investors do not demonstrate a direct bias in terms of making more offers to male entrepreneurs (Table 5). This finding contradicts the initial hypothesis based on recent reports of female underrepresentation in Germany (Hirschfeld et al. 2022; Schlüter 2018) and previous studies indicating gender discrimination by angel investors in private settings (Balachandra 2020; Färber and Klein 2021). However, our results align with studies of US televised pitch competitions, which similarly did not find significant gender disparities in offer and deal probabilities (Boulton et al. 2019; Hohl et al. 2021; Jetter and Stockley 2023; Poczter and Shapsis 2016). The results suggest that gender bias in these formats may be mitigated by public scrutiny, given the ongoing discourse on discrimination against female entrepreneurs in Germany (Hirschfeld et al. 2022). It is also noteworthy that these formats typically feature a higher proportion of female investors (40% in a panel of 5), which is significantly above the global market average of around ~ 25% (Sohl 2020), potentially influencing gender-related deal probabilities. Future research should explore the interaction between funding for female entrepreneurial teams and the gender diversity of angel investors across Germany, particularly regarding potential gender similarity biases (Becker-Blease and Sohl 2007; Ewens and Townsend 2020; Jetter and Stockley 2023).

Besides, evidence of gender bias emerges when considering the German angel investors’ significantly lower valuations of female-led businesses. This finding mirrors results from Shark Tank research in the US, where substantial gender differences in business valuations were observed (Hohl et al. 2021; Jetter and Stockley 2023; Poczter and Shapsis 2016). Considering that female entrepreneurs have been shown to lead more profitable businesses than their male counterparts (Morazzoni and Sy 2022), this disparity indicates a systematic bias. Jetter and Stockley (2023) argue that the funding gap may partly stem from higher requested business valuations of male teams. The authors suggest that this reflects higher self-confidence of male entrepreneurs, and there is additional evidence that female entrepreneurs undervalue their firms due to lower self-assessment of their abilities (Hohl et al. 2021; Jetter and Stockley 2023; Kirkwood 2009; Poczter and Shapsis 2016). Despite these differences in requested valuations, male entrepreneurs still manage to secure deals (Poczter and Shapsis 2016), highlighting a partly self-imposed gender funding gap. This finding resonates with broader research indicating higher levels of confidence and overconfidence among men (Barber and Odean 2001; Lundeberg et al. 1994), also contributing to the observed gender disparities among German investors.

Contrary to our expectations and findings from other settings (Blanchard et al. 2008; Boulton et al. 2019; Younkin and Kuppuswamy 2018), our results suggests a slight preference among German angel investors for making deals with entrepreneurial teams of diverse ethnic backgrounds (Table 5). This positive bias may stem from stereotypes about the stronger work ethic and resilience of ethnic minorities (Hsin and Xie 2014), influencing investors’ perceptions of commitment to business success. On the other hand, given the televised setting, German angel investors in DHDL might be particularly sensitive to perceptions of bias against ethnic minorities and potential allegations of racism. It is thus conceivable that angel investors overcompensate in their decision-making, resulting in the observed small positive effect. Nonetheless, the non-significant effect of ethnicity on deal valuations suggest that this overcompensation does not translate into higher business valuations for startups led by ethnically diverse teams. At the same time, this findings contrast with Boulton et al. (2019), who found significantly lower valuations for black entrepreneurs in the US sample. Hence, in direct comparison, a potential ethnicity bias in the sense of racial discrimination is less pronounced in Germany than in the US.

Entrepreneurs’ physical attractiveness, as hypothesized, significantly influences the likelihood of angel investors extending offers and making deals with the team (Table 5). This aligns with prior research on the positive stereotypes associated with attractiveness in general (Dion et al. 1972) and attractive entrepreneurs in particular (Brooks et al. 2014; Schreiber et al. 2024). Our findings further expand on this prior work by demonstrating the positive impact of attractiveness bias for both male and female entrepreneurial teams. Given that our study is the first to prominently explore the impact of entrepreneurial attractiveness in the setting of televised pitch competitions, direct comparison of German investors’ attractiveness bias to those of other geographies would require further research. Nevertheless, assuming physical attractiveness does not result in stronger entrepreneurial skills and startup success, our results indicate a systematic bias. Importantly, the effect of entrepreneur’s physical attractiveness is limited to offer and deal probabilities, as attractiveness does not influence deal valuations among German angel investors. This suggests that rational considerations prevail when deciding on financial terms.

Additional variables were included to reflect criteria potentially linked to personal characteristics and pitch outcomes. As anticipated, strong presentation skills positively impact deal and offer probabilities. Entrepreneurial passion, however, does not directly affect deal probability and valuation. Moreover, team size significantly influences deal valuations. This likely reflects the importance of human capital for angel investors (Pollack et al. 2012). Larger teams often possess a broader range of skills, which increases investor confidence and willingness to offer higher valuations. Moreover, larger teams may have an advantage in equity negotiations, aligning deal valuations more closely with their initial requests (Poczter and Shapsis 2016).

Controlling for the televised context, the entertainment value of pitches negatively impacts deal probability and valuations. This corroborates prior research indicating that startups selected by the production company for their entertainment value may not necessarily represent the most promising investment opportunities and are then rejected by investor panels at the observed stage (Smith and Viceisza 2018). Moreover, this underscores angel investors preference for startups that emphasize business potential over entertainment appeal, given the personal financial stakes involved.

Finally, evidence suggests that deal probabilities increase with the number of seasons, indicating a rising trend in investor activity across successive seasons. This may reflect an increasing pool of promising startups attracted to the format as its popularity format grows. Additionally, startups may benefit from a learning effect, studying old episodes and improving their pitching techniques compared to earlier seasons.

5.2 Theoretical implications

Social perception theory posits that people use stereotypes linked to perceived superficial characteristics to make assumptions about others, which is an automated cognitive process in interpersonal perception (Cook 2021; van Knippenberg and Dijksterhuis 2000). Given the information asymmetries in pitch competitions, this process can lead to biased decision-making by angel investors and the discrimination of entrepreneurial groups. This study contributes to this debate showing that angel investors in Germany are significantly impacted by the superficial characteristics of founders.

Theoretical perspectives in interpersonal perception assume different cognitive processes based on the varied impact of decision-making criteria (Cook 2021). Our partly bidirectional results (Table 5) suggest that angel investors apply different stereotypes and heuristic processes for extending offers vs. deciding on a reasonable business valuation. The decision to offer a deal is more intuitive and influenced by superficial characteristics such as physical appearance. In contrast, valuing a business is a more complex decision driven by reasoned preference for younger and male entrepreneurs. This interpretation, corresponding to Kahneman’s (2011) framework of a fast and a slow cognitive system as made popular by, may be fruitful for future exploration.

The televised setting may further impact investor behavior. Angel investors are likely aware of the most prominently discussed biases around gender and ethnicity, and do not want to appear discriminative. To manage their public impression, they might actively adjust their decision-making against the appearance of sexism and racism when deciding whom to offer an investment. On the other hand, biases related to age and attractiveness are less prominently discussed in the startup ecosystem and less obvious to the audience, so investors do not adjust for these biases as actively. The potential public scrutiny and likely changes in the behavior of angel investors offer another interesting area for further exploration.

5.3 Limitations and directions for future research

Subjective measures such as attractiveness and ethnicity present challenges for data collection and interpretation. Studying these sensitive attributes responsibly is crucial to mitigate ethical concerns that can arise from making assumptions about the perception of others, e.g., the reinforcement of stereotypes (Schreiber et al. 2024). Like previous studies of televised pitch competitions, we relied on a limited number of observers to assess these characteristics, which can introduce subjective observer bias (Boulton et al. 2019; Poczter and Shapsis 2016). We cannot be certain that the observers’ assessment of attractiveness and ethnicity matches the perception of angel investors, which can limit the internal validity. Ethical research practices could be improved by surveying angel investors’ subjective perceptions directly, adding a significant number of observers from diverse backgrounds, and combining human coding with machine-based ratings of ethnicity and attractiveness (Smith and Viceisza 2018).

Statistical limitations are another potential concern as our use of multivariate logistic and linear regression models assumes linear relationships, which may limit the fit and the explanatory power. Future research could examine if relationships between investor decisions and attributes like age or attractiveness are nonlinear, e.g., suggesting an optimal team age or optimal level of attractiveness when applying for venture funding, or if there are interaction effects indicating an optimal combination of attributes for success.

To understand if founder-related bias of angel investors results in systematically suboptimal investment outcomes, our results only provide a first perspective. We concentrate on the analysis of data exclusively derived from pitch recordings and not diluted by additional third-party data or qualitative valuations (Hohl et al. 2021). Future studies could include data on long-term outcomes to provide a more comprehensive perspective on startup failure and success in relation to the investor decision. For instance, observable indicators of long-term success include business survival, later funding rounds, team growth, website traffic and customer satisfaction (Smith and Viceisza 2018). This would allow the identification of a superficial bias resulting in the over- or underestimation of entrepreneurial teams.

Furthermore, this study focused on a small panel of expert investors in a televised competition, as in prior studies (Hohl et al. 2021; Maxwell 2011; Poczter and Shapsis 2016). It is possible that these investors are not representative of the general population of angel investors. To increase generalizability, future studies should examine gender, age, and ethnic diversity of investors, and include information about the investors’ reputation, political attitudes, wealth, education and experience, given that these are shown to influence the investment decisions (Moritz et al. 2022), their likelihood for co-investments (Maus et al. 2023) and the subsequent performance of startups (Blaseg and Hornuf 2024). For instance, conservative angel investors show a stronger gender bias against female entrepreneurs than liberal angel investors (Chen et al. 2023). Besides, the impact of social desirability in the televised setting may influence the occurrence of bias, with the assumption that they have an even stronger impact in private settings (Boulton et al. 2019). Future research could compare investor decisions across different settings to see if transparency acts as a control mechanism for stereotypical bias.

To generalize findings across industries and other pitch settings, a more diverse sample of startups is required. Most startups in the DHDL format offer consumer products (Poczter and Shapsis 2016). To understand and predict the effects of team characteristics on business angel decision-making across Germany, other industries and especially more B2B startups should be included in future samples. It is conceivable that investors show different bias depending on the industry, e.g., favoring older entrepreneurs in more traditional industries. Lastly, survivor bias may impact our sample, which only includes startup ideas that made it through the initial formation phase and the pre-selection process. To address this, we recommend including a perspective on all startups that originally applied for funding.

5.4 Practical implications

Decision bias of investors in relation to superficial characteristics of entrepreneurial teams impacts the startup ecosystem, often resulting in disadvantages for specific founder groups regarding funding accessibility. This can negatively affect the overall diversity of entrepreneur demographics, which may negatively impact innovation of diverse new products and services offered to the market. Beyond the implications for theory and research, this work can thus provide valuable impulses for practitioners in the startup ecosystem, both globally and for Germany specifically.

5.4.1 Entrepreneurs

When seeking funding, younger as well as less attractive entrepreneurs must be attentive of the prevalent stereotypes and potential bias against them impacting offer and deal probabilities. Older and female entrepreneurs, in turn, should be vigilant about a bias affecting business valuations and more confidently justify their worth. This is crucial, given that stereotypes have a self-fulfilling nature according to social perception theory (Snyder et al. 1977), and it is easy for affected individuals to succumb to them. For instance, the bias against female entrepreneurs can be partially mitigated by emphasizing their prior experience and qualification (Tinkler et al. 2015). Attending tailored pitch training workshops could help them enhance their presentation strategies and highlight their achievements supported by evidence over superficial characteristics. Furthermore, the televised setting of DHDL may act in favor of ethnically diverse entrepreneurial teams, as the German angel investors may be expected not to appear as discriminatory, increasing the likelihood of an offer. Ethnic minority entrepreneurs may find that such public platforms provide a promising route to gain visibility and secure early-stage funding.

5.4.2 Angel investors

Angel investors in pitch competitions can apply measures to be more inclusive and objective in their decision-making. First, to avoid biased pre-selection, investors can implement diversity policies and actively seek out founders from a diverse range of backgrounds. On that note, Hohl et al. (2021) suggest raising the number of female entrepreneurs and angel investors in public pitch competitions to serve as role models for aspiring female entrepreneurs. Second, angel investors should focus on structured evaluation criteria including the factors that are most important for the success of startups. For instance, they could arrange blind auditions to eliminate the impact of superficial characteristics (Goldin and Rouse 2000). Third, investors should expand their networks, build relationships with other investors and advisors, and actively seek out different perspectives. They could create collaborative elaboration panels with members from varied backgrounds to provide less biased feedback. Overall, they should train their awareness, constantly reflect on decision outcomes, and consciously recognize situations that might provoke bias. As a result, angel investors can contribute to the creation of a more inclusive, equitable and ultimately more profitable startup ecosystem.

5.4.3 Policymakers

Policymakers in Germany and throughout Europe can further increase their engagement towards a more equitable startup ecosystem. Prior research suggests that state financing of research phases and investing in entrepreneurship education can foster an inclusive entrepreneurial culture in Germany (Fuerlinger et al. 2015). Considering our findings, educating female and older entrepreneurs on business valuation and negotiation with investors may reduce the funding gap. Governments can build inclusive grant and funding schemes that incentivize diversity and inclusion. Moreover, facilitating support networks for different underrepresented groups may increase their success in early-stage funding. With their central role, policymakers could increase the transparency in funding decisions by monitoring and reporting investment data and developing feedback mechanisms to investors, where investments are reviewed for bias so that investors can adjust their decisions. Finally, they can develop specific policies that mandate or encourage inclusive practices among investors and within the startup ecosystem.

5.5 Conclusion

Building on psychological principles from social perception theory and exploring the practical consequences of stereotyping in economic decision-making, this study highlights the great societal interest in improving the objectivity and decision quality of angel investors. Bias related to superficial characteristics can have detrimental effects: entrepreneurs may be unfairly denied funding due to personal attributes, even if those are irrelevant for future their venture’s success. On the other hand, biases may prevent angel investors from investing in potentially profitable ventures, and at the same time increase the risk of overconfident judgements about ventures that ultimately fail.

By analyzing data from ten seasons of the German televised pitch competition Die Höhle der Löwen, this research reveals how superficial characteristics of entrepreneurial teams influence angel investor decisions. Complementing and extending prior evidence from US samples, our findings indicate that deal likelihood is positively related to older age and diverse ethnicity, while resulting deal valuations are positively linked to younger age and male gender. Additionally, this study is the first to provide evidence that physical attractiveness positively affects deal likelihood but not deal valuation. The bidirectional results suggest different cognitive processes are active in the distinct stages of angel investor decision-making, affecting the impact of representative entrepreneurial stereotypes.

In conclusion, the significance of public pitch competitions like DHDL for the German startup ecosystem should not be underestimated: These formats provide researchers with valuable data to understand the mechanisms of both entrepreneurial and investor decision-making. They increase the visibility of diverse entrepreneurial groups and provide access to capital and publicity for early-stage startups. Finally, they enable a broad audience to hold investors publicly accountable for their decisions. Therefore, they effectively contribute to the ongoing public debate about decision bias and fairness in the startup ecosystem.