The Startup Selection Process in Accelerators: Qualitative Evidence from Turkey

3.1 Research Context

Creating interfaces such as accelerators to establish and enhance innovation is a critical component of Turkish technology and innovation policy. Such interfaces as organizations, connected firms and entrepreneurs, and the professionals working in the interfaces define an environment where technology-based entrepreneurship flourishes.

As of 2020, there are 57 accelerators in Turkey (Investment Office 2020) in various sizes and forms, most of which are located in either Ankara or Istanbul. Accelerators have been well received by the other actors of the entrepreneurship ecosystem. The number of accelerators increased approximately 10-folds from 6 in 2010 to 57 in 2020. However, when the fieldwork started in 2016, there were 24 accelerators in Turkey. Most accelerators in Turkey were formed to organize certain elements (intangible services to obtain fast traction) of the entrepreneurship support activities already present in prominent universities in Ankara and Istanbul. Due to law, some accelerators are organized under the university technology park or the university’s technology transfer office. Only a few of the accelerators are private or sponsored by corporations, municipalities, or non-government organizations.

3.2 Sample Selection

Accelerators differ from incubators in many aspects (Cohen and Hochberg 2014; Cohen et al. 2019; Galbraith, McAdam, and Cross 2019; Pauwels et al. 2016; Sansone et al. 2020); for instance, the duration of the program is shorter than incubators, and they accept startups in cohorts. Due to the high number of applications, most of the accelerators use a staged selection system. At the first stage, applications are evaluated based on information provided by applicants through a web-based system. Then, at the second stage, the successors of the first stage receive an invitation for an interview (Yin and Luo 2018).

As a first step, we made a list of accelerators that we would interview for this research by collecting secondary data from newspapers, magazines, public agencies, and the websites of accelerators. We considered three criteria for selecting cases:

1. a history of continued operations: Some accelerators have sustainability problems, some cease to exist, and some are very small in size. The selected accelerators have a certain scale and continuing operations for more than a year (see Table 1, column 2).

2. the duration of the acceleration: It ranges from several weeks to one year (see Table 1, column 3).

3. cohort-based selection: The selected accelerators have well-defined application and selection processes in cohorts (see Table 1, column 4).

We identified about 15 accelerators that fit the above criteria in Turkey (out of 24 that existed when the fieldwork started in 2016). The accelerators in Ankara and Istanbul were contacted, and 10 accelerators agreed to participate in the research. All accelerators in our sample use a staged selection system where at least in the second stage, a jury or a committee decides on the applicants’ faith. Turkish Economist (Ekonomist 2018) reports about more than six thousand startups that have benefited from the acceleration services as of January 2018, 75% of which were officially registered in six accelerators in our sample (Accelerators B, C, D, E, F, and I in Table 2). Table 2 provides brief information on the selected accelerators, the role of the interviewee, and the date of the interview.

3.3 Data Collection and Analysis

The data used in the empirical research was collected from two sources: primary data came from the transcribed semi-structured in-depth interviews and site visits, while secondary data was collected on literature review and archival data (such as accelerators’ websites, blogs, newspapers, newsletters). Archival data were mainly used to select accelerators, prepare for the semi-structured interviews, open up new questions in interviews, and supplement interview data. Thus, data triangulation has secured our interviews to be carried out with a representative group of accelerators. This approach of using supplementary data to improve a primary set of collected data has been widely used in the literature (e.g., Sklyara et al. 2019).

We interviewed accelerator managers at their offices face-to-face and observed their operations and their relationship with startups at these visits. All interviews were conducted from March 2016 to January 2017. During these interviews, a set of questions were asked to collect information about the organizational structure of accelerators, history of operations, general processes before and after the selection, detailed information about selection and selection committees, selection criteria, the definition of an ideal startup. The length of the interviews was between 45 and 90 min. All of them were recorded and transcripted.

As consistent with inductive research methodology (Bryman and Bell 2007), first, we tried to understand how startups are selected by different actors such as VCs, business angels, incubators, and accelerators. After an intense literature review on startup selection, the authors reviewed the transcripts and field notes. Two of the authors made the first contacts with accelerator managers, field visits, and face-to-face interviews. An iterative process was used to analyze the information embodied in transcribed interviews and compare and contrast the analysis with the theory. We conducted a peer evaluation process (Miles and Huberman 1994). All authors contributed to coding by carrying out independent parallel analysis of data items, but all codes are based on consensus (Bryman and Bell 2007). Thus, we did not conduct an inter-judge reliability test.

We searched for themes that we drew from previous research; however, we embraced the new ones. This iterative process involved drafting themes and sub-themes that emerged from face-to-face interviews and the data collected from secondary resources. We looked for emergent codes and statements in the transcripts, grouped these codes under sub-themes and themes, and re-read the transcripts to edge the sub-themes and interpret the findings (Bryman and Bell 2007; Denzin and Lincoln 2000; Stake 1995). The first read of the transcripts was based on the themes from the literature and produced various codes (column 2, Table 3) that derives from actual statements (column 1, Table 3). Then these codes were refined and matched to sub-themes (column 4, Table 3) (Gioia, Corley, and Hamilton 2012). Once a draft code-sub-themes and themes structure was ready, a second read was conducted for the robustness of the analysis. The final analysis produced 21 codes grouped under five sub-themes of selection (columns 2 and 4, Table 3). Previous research that scrutinizes the role of sub-themes in startup selection is also provided in Table 3 (column 3). Finally, these sub-themes are grouped under two main themes of startup selection, “business opportunity” and “team” (column 5, Table 3). We reviewed each case and checked how these five sub-themes appear in the interviews at the final stage. The common recurring sub-themes can be viewed as criteria of startup selection to accelerators.

4.1 The Engagement of Various Actors through Selection Committees

Accelerators aim to attract high-quality teams in each cohort. Social media and websites are the basic channels for reaching potential entrepreneurial teams. They collect applications on different occasions via web-based forms as well. In the first step, evaluations are made using the information in these forms, and most applicants are eliminated based on available or missing information. Application forms include questions which help applicants to better explain the entrepreneurial opportunity they have discovered, and their business model. One interviewee (Interviewee D) mentioned “what you make different and how you make it different? Who are the customers? What are your unique resources? Applicants have difficulties answering such questions”. This sharp trimming down makes further screening affordable. If acceleration experts are not sure about the appropriateness of a startup, they contact applicants face-to-face or on the phone to get the details about the entrepreneurial opportunity and the team. Sometimes they are given advice on how to improve their projects. By spending time with startups in this process, accelerators get closer to startups.

In the second step, selection committees make the final selection. The size of the selection jury ranges from three to 40. Most of the accelerators in Turkey invite angel investors, industry representatives, sponsor organizations, mentors, or influential serial entrepreneurs to their selection committees. By doing so, they open up the communication channels and increases the interaction between investors and accelerators and also between investors and startups. Accelerators have the chance to learn more about the concerns and expectations of investors and the type of startups they are willing to invest. In these selection committees, norms and understandings regarding the new ventures are exchanged, discussed, and opened to negotiations. Accelerators interact vividly with potential investors. As stated by one interviewee, “inviting investors to our selection committees is a signal to investors: t hese startups will be brought to you when they are seeking investment. Now, you need to tell us what you want. Then we will find those startups for you” (Interviewee D).

The interaction between investors and accelerators is not limited to the selection process. It continues in different forms; for example, some investors serve as mentors in accelerations, and can benefit from information flow from accelerators about the startups under their tenants. These interactions help them select the most appropriate startups to invest and close the information gap between investors and startups. Thus, accelerators build trust on behalf of the investors that the accelerator will bring them what they want (as emphasized by Interviewee D). These interactions are also valuable to accelerators to learn more about the appropriate signals a startup needs to vibrate to access resources. In that way, they can sense the products, markets that investors are most interested in, or the team qualifications that are concerned by investors and the appropriate ways of showing off these qualifications.

Our research indicates that program managers and experts make their decisions very quickly, especially at the first round of selection, where they need to downsize hundreds of applications into reasonable amounts. Nevertheless, the experts in accelerators have a chance to spend more time with applicants to observe them. They learn more about entrepreneurs’ backgrounds, motivations, what they did so far for their project, and their future projections and work together to understand how teams take the feedbacks and respond to them. For example, “we are giving small assignments; such as a competitor analysis. Their returns and works in this process give us a lot of clues about how they are serious. Most of the time, an application form is not enough” (Interviewee E). Accelerators trim the most appropriate entrepreneurial teams and opportunities that are likely to be invested. They eliminate red-flagged startups, which cannot vibrate the “right” signals. Thus, accelerators clean the noise and find out the most appropriate new ventures for investors by reducing information asymmetry.

Accelerators bring investors into the startup selection process. Accelerators prefer to engage investors in the selection process to assess better startups’ potential to achieve product-market fit and quick scalability. Additionally, it helps them build relations with investors that could provide private funding as quickly as possible since it is a matter of survival and the key to scaling fast for startups. As stated by Interviewee A, most of these accelerators have close connections with angel investors and most of these investors serve as mentors to startups: “we are well connected to angel investor networks… In one of the mentorship sessions, one of our angel investor mentors very much liked the presentation of a startup. He wanted to work with them voluntarily. Now they are working together 2–3 days a week. He has not invested yet but wants to mentor them for now”.

In connection with other supporting mechanisms, selection mechanisms applied by accelerators generate benefits for both parties by reducing the information asymmetries among investors and startups. They eliminate the red-flagged startups at very early stages for investors by bringing investors’ concerns and expectations forward in the selection mechanisms and training startups to vibrate positive signals to investors. In this environment, private investment actors can shape the new ventures’ context and growth (Baum and Silverman 2004).

4.2 Selection Criteria for Entrepreneurial Teams

Due to high uncertainty and ambiguity, it is difficult to decide based on future projections of an entrepreneurial project. Therefore, most of the experts/managers of the accelerators emphasize that they do not consider the financial projections of a startup for selection and screening. From the interviews, two main selection criteria emerge: the entrepreneurial opportunity and the team. Although these two criteria are common in the selection, there are still differences among accelerators. The criteria applied to startups are essential to understand how new ventures are selected by accelerators and the impact of selecting entrepreneurial activities and the ecosystem.

The first criterion is the entrepreneurial opportunity, consisting of concerns about the problem, value proposition, innovativeness, and market-related issues such as market size. The second criterion covers founder and team characteristics. It ranges from the technological and business-related competencies of founders to their passion. Figure 1 depicts salient signals which are considered by accelerators while selecting entrepreneurial teams to nurture.

Figure 1:

Criteria applied to startups in the selection and screening process.

Source: Interviews.

The market addressed by the entrepreneurial team is critical in the evaluation process. One of the interviewees suggests that “the problem gives us an idea about the market” (Interviewee A). However, how the designated problem is solved is as important as the problem. The solution needs to be innovative and offer something new to customers. Nonetheless, acceleration managers are aware of the difficulties of creating a new market. One of the interviewees explained that “it is meaningless to create a product which does not have an existing market” (Interviewee H). This point reminds us the concept of legitimate distinctiveness (Navis and Glynn 2011). Having an innovative idea is a positive signal to attract the attention of accelerator experts. However, this innovative solution must address an existing market that is easily identifiable by investors and other stakeholders. Creating an entirely new market is challenging, time-consuming, and expensive. The novelty of entrepreneurial opportunities is appreciated up to a limit. When the entrepreneurial opportunity is latent, target customers and market size are not identified. Then, innovativeness and novelty can even turn into an obstacle. In such a situation, the signal does not fit the receiver’s expectations.

Similarly, as the time needed to commercialize the entrepreneurial opportunity increases, the probability of being selected by any accelerator decreases. This criterion limits the technology fields that accelerators prefer. They mostly prefer technologies and products that can be developed quickly and have an easily identifiable market. A working prototype is the most valuable signal to measure the required time to market, which is a step toward being selected.

Interviewees frequently mention that team characteristics and composition are the most important factors they consider in screening and selection. Our research reveals that acceleration experts consider many different factors related to entrepreneurial teams. These factors consist of characteristics such as passion and commitment exhibited by entrepreneurs, fields of education, knowledge about markets, preparedness, coachability of teams, diversity in entrepreneurial teams, and chemistry among team members. The technical and business-related capabilities of entrepreneurial team members are interpreted as valuable signals of how well the team fits the opportunity. Acceleration experts generally do not have expertise in technological fields, and therefore cannot provide mentoring or training to support startups in relevant problems. Thus, the teams capable of solving technical problems are also seen as critical for the sustainability of startups. For accelerators, the educational background of founders is a signal to measure the capabilities of entrepreneurial teams. The signal of business-related capabilities is not formal education or relevant industry experience but instead team members’ attitudes and willingness to engage in business-related activities.

In most cases, technical competencies create a disadvantage for entrepreneurial teams because they focus too much on the product’s technical qualifications and ignore customers. How team members work together, how long they know each other, and how they allocate the responsibilities within the team are also valued as “right” signals. One of our interviewees reveals that “we are looking for people who are very good at technical issues but also they can easily communicate with people, be persuasive and be capable in bargaining, and develop themselves in selling and marketing” (Interviewee E).

Two signals are remarkably voiced in the interviews: passion revealed by entrepreneurial teams and whether they are coachable or not. Though it is difficult to observe and measure entrepreneurial passion, we identified several signals regarding teams’ commitment to the entrepreneurial process’s overwhelming requirements and preparedness. The willingness of entrepreneurial teams to dedicate their full time to entrepreneurial activities is taken as a positive signal. The personal investments of team members in their project is also a “right” signal. One of the interviewees emphasizes that “if an entrepreneur does not risk her savings for this project, it means that she does not believe in her project” (Interviewee E). Signals of commitment are critical for accelerators to distinguish applicants who persist in their entrepreneurial endeavors. As previous research emphasizes, entrepreneurial passion is manifested in behaviors and narratives of teams (Chen, Yao, and Kotha 2009; Pollack, Rutherford, and Nagy 2012). We find that building a prototype, doing market research, building networks, benchmarking the product with existing competitors, even having some ideas about target customer segments, markets, and telling a consistent story about the emergence of business are perceived as the signals of preparedness and entrepreneurial passion by the interviewed accelerators. The flexibility of entrepreneurs and their willingness to collaborate with managers/experts are seen as eminent signals of coachable entrepreneurial teams. In literature, coachability is defined as “the degree to which an entrepreneur seeks, carefully considers, and integrates feedback to improve his or her venture’s performance” (Ciuchta et al. 2018, p.861). Our interviewees emphasize that entrepreneurs must be open to collaboration with accelerator managers, experts, and mentors to improve their entrepreneurial projects further. They must be open to constructive criticism and pivot their projects when it is necessary. Interviewee D states that “in order to benefit from the services provided by our program, they have to be open to criticism and change… if an entrepreneur does not want to try what we advise her to improve the project, we do not want to work with her”.

Due to the high uncertainty and ambiguity, it is difficult to make realistic projections for experts about a proposed entrepreneurial project. They focus on signals vibrating the quality of the teams, the fit of the team to the opportunity, and the founders’ passion and commitment. If the experts observe positive signals in a team, they are more likely to invest in time and effort. Interviewees emphasize their unwillingness to spend their limited resources on teams that lack passion and commitment or those that would eventually disband due to disputes among the team members.

The criteria applied by accelerators resemble those applied by investors to make investment decisions (as shown in Table 3). In most cases, accelerators include a wide range of audiences (i.e., entrepreneurs, business angels, VCs) in their selection process (Clarysse, Wright, and van Hove 2016; Leatherbee and Gonzalez-Uribe 2018). Intense interaction and exchange between investors and accelerators in the selection process and inviting investors to the selection process increase the probability of uniform understandings of the appropriate or acceptable new ventures. In the selection process, accelerators reproduce common understandings about an appropriate and eligible startup (Meyer and Rowan 1977). Signals of legitimate organizations are created and reproduced in every instance by accelerators, investors, mentors, and startups. Some signals become more observable and easily recognized and interpreted by receivers, while others are ignored. In sum, our study shows that accelerators act as a substitute for investors while selecting the startups to nurture.

Inexperienced entrepreneurial teams are not aware of the right signals to vibrate to access funding. During the acceleration process, as one of our interviewees briefly explains, “the proposed entrepreneurial project is shaken, deconstructed and reconstructed by entrepreneurs in collaboration with program experts, mentors, and trainers” (Interviewee G). The entrepreneurial narrative is rewritten in this process in collaboration. Accelerators help entrepreneurs to reconstruct their entrepreneurial opportunities and narratives. Startups learn how to vibrate the right signals and vibrate them in the right form (Busenitz, Fiet, and Moesel 2005; Higgins and Gulati 2006; Kirsch, Goldfarb, and Gera 2009). Accelerators coach startups to vibrate positive signals in an appropriate form. Therefore, accelerators tend to select the most coachable, open to collaboration, passionate, and willing to be committed startups. As the number of successful startups increases, the reputation of accelerators and their capabilities to mobilize external resources increase too. This reputation works as a positive signal to possible investors.

5.1 Implications for Theory

This paper offers two contributions to the literature on accelerators. First, the findings help us understand the selection in accelerators as a complex process involving multiple actors. Accelerators overcome the uncertainty in their context by involving various actors in the selection process. The collaborations within the selection committees facilitate a futile ground for communication among accelerators, startups, and investors. This research also exhibits the criteria applied by accelerators in Turkey to select startups to nurture. Our findings show that accelerators focus on opportunity and team, with specific emphasis on team members’ characteristics, including coachable, open to collaboration, passionate, and willing to be committed entrepreneurs. These findings resemble how VCs and angel investors select their own invested entrepreneurs (Huang and Pearce 2015; MacMillan, Siegel, and Narasimha 1985). Second, our findings support previous research that suggests that accelerators are more connected to investors in the selection process (Clarysse, Wright, and van Hove 2016; Leatherbee and Gonzalez-Uribe 2018). But they also exhibit how these connections help the quick scalability of startups in accelerators. By including investors in selection committees and applying selection criteria resembling those of investors, accelerators substitute investors and trim down the startups which cannot attract investment and hence cannot quickly scale. Furthermore, by preferring coachable, passionate, and willing to be committed entrepreneurs, accelerators spend their resources on those that they can train to reach milestones faster and therefore vibrate positive signals to prospective investors.

5.2 Implications for Practice

There are some practical implications of accelerators behaving as receivers and signalers in the selection and screening process. On the investors/resource holders’ side, investors’ engagement in the selection process in accelerators may lead to a well-functioning and more effective selection. Moreover, new venture teams who are less suffered from information asymmetry can manage their impressions over resource holders, increase their chance to access private funds (Maxwell, Jeffrey, and Levesque 2011), and help their accelerator gain a reputation in the ecosystem. In a developing country context, this would increase the number of business angels and the total amount of seed funding provided to startups, attracting more foreign VCs and increasing the number of seed-funded startups as observed in the Turkish case (Investment Office 2020). The interaction among different actors of the entrepreneurship ecosystem would increase and become more institutionalized.

However, some unintended practical implications may also be observed. Bringing investor selection and selection criteria backward to earlier stages of the startup life cycle may confront the entrepreneurial teams’ market selection biases very early. For example, they might eliminate some disadvantaged groups, e.g., female lead entrepreneurial teams (Lall, Chen, and Roberts 2020; Yang, Kher, and Newbert 2020). Also, the early involvement of investors might be biased to entrepreneurial teams with radical product/service innovations or deep-tech innovations in which the problem/need is latent, and the market is not foreseen. Since the entrepreneurs with radical ideas may not be easily coachable (Ciuchta et al. 2018), they can be dropped out of the system where the resources are scarce. The risk for accelerators might be selecting startups with low impact and identical or resembling products/services in motivation to increase the number of startups funded by private investors. In the long run, this problem may lead to a decrease in the efficiency of the acceleration system and may create a reverse impact on their reputation. It would be beneficial for accelerators to be self-reflective about their selection procedure and revise it to ensure variety in each cohort.

In sum, accelerators bring many advantages to stakeholders of the entrepreneurial ecosystem. Startups get early access to finance, while investors identify worthy startups. Accelerators become critical mediating agents in the ecosystem thanks to their efforts to reduce uncertainty and increase seed funding efficiency.

5.3 Limitations and Future Research

This explorative study has four main limitations that might become opportunities for further studies. First, this study exploits a limited number of in-depth interviews with accelerator managers in Turkey. Hence, our findings’ generalizability may increase if similar studies covering accelerators in other countries are conducted to compare findings across different contexts. Second, the study focuses on the selection process in accelerators, while other activities such as networking and mechanisms to improve startups’ market presence and survival might also influence the performance of startups. Third, our research does not consider the negative impact of the selection processes in accelerators. Bringing investor selection to the early stages may lead to the institutionalization of “a certain type of startup.” As observed in our interviews, the selection mechanism emphasizes picking the winners. Such a selection mechanism might create biases against some startups that develop radical or disruptive technologies with high market uncertainties and longer time-to-market, eventually reducing heterogeneity in the entrepreneurship ecosystem. Finally, this study interviews accelerator managers, while future studies should interview other stakeholders involved in the selection process and entrepreneurs who are selected into accelerators. An additional research avenue might be conducting quantitative studies to understand the impact of different selection processes on the final performances of the accelerators.