Unraveling the Selection Phase of Business Incubators: Proposal for a Conceptual Model and Future Research Agenda

Abstract

The selection of start-ups by business incubators (BIs) is one of the main processes of these organizations that aim to promote entrepreneurship and economic development. Through a systematic literature review of the BI selection phase, following the PRISMA guidelines, a new conceptual model is proposed, delineating findings across three dimensions: the incubatee–incubator alignment and BI strategy, the selection criteria considered, and the decision-making methods used. The conceptual model proposed here represents the first conceptual model focused exclusively on the selection of start-ups by BIs, mapping practices in terms of start-up selection that hold important practical implications for BI managers. Considering the growing need to align economic performance with environmental and social objectives, the start-up selection process by BIs also plays a key role in fostering sustainability-oriented ventures. This fact increases the relevance of this selection phase, not only in terms of operational efficiency, but also as a potential lever for sustainable regional development. Findings emphasize the need for future research that integrates all three dimensions simultaneously, explains the selection process at an operational level, and assesses the importance of this phase for BI performance.

1. Introduction

BIs are organizations that aim to support start-ups at a stage in their life cycle marked by great uncertainty, lack of resources, and quite high failure rates [1]. This support is provided through a series of tangible and intangible services [2] that can be divided into infrastructures, operational services, business support, networking, financing, event organization, educational and self-organization [3]. The incubation process typically unfolds in three phases: the pre-incubation selection phase, the incubation phase and the exit and post-incubation phase. The pre-incubation selection phase, which is the focus of this study, involves start-ups applying to the incubation program, the BI collecting detailed information about these candidates, through interviews, for example, and the BI deciding which candidates are selected to join the BI. Following this, the incubation process also includes the incubation phase, which is the core of the whole process, the one that lasts the longest and where the BI’s various services are provided to the incubatees with the aim of their growth and development. Understanding how the internal capabilities of a BI can be used to provide these services effectively has therefore become relevant, especially when aligned with sustainable development goals [4]. The last phase of the incubation process is the exit from BI and the post-incubation period that follows, where the graduated company is expected to be able to grow and survive outside of BI.

In recent years, the role of BIs as an instrument for promoting sustainability has gained relevance beyond the classic roles of promoting entrepreneurship and regional economic growth that are associated with these organizations. BIs have the potential to incorporate into their selection decisions elements that favor start-ups developing environmentally responsible products, promoting social innovation or adopting circular economy models [4,5]. In this way, and considering that many BIs are publicly funded, they can help tackle global challenges such as climate change, resource scarcity and inequality. This aligns the selection phase with broader political objectives, including several of the United Nations’ Sustainable Development Goals (SDGs), in particular SDG 8 (decent work and economic growth), SDG 9 (industry, innovation and infrastructure) and SDG 12 (responsible consumption and production).

The importance and inclusion of the selection phase in the business model of business incubators has been addressed by some scholars [6,7]. In general, these studies build on the assumption that the use of specific criteria and processes in this phase will allow the identification of start-ups with a lower probability of failure and greater growth potential, both in terms of employment and financial indicators. The selection of start-ups to enter an incubation program is a critical factor for BI performance [8] and a fundamental step in the incubation process and in the start-up-venture capitalist relationship [9]. This process includes the selection of a set of start-ups that BI managers believe will benefit from the incubation process and its support, which will not only increase the likelihood of survival and obtaining a grant or venture capital [10] but also achieve higher levels of growth while reducing costs [11].

This selection phase is part of the BI model [6,7], and its necessity derives from BI’s limited capacity in terms of services and office space [12] in conjunction with the high demand they face. This calls for the implementation of rigorous selection processes [13] that will lead to an improvement in the performance of the start-ups and the BI itself [14,15]. In addition to this justification based on an operational capacity-demand management problem, since the selection phase has clear implications for the performance of the BI itself, it impacts the BI’s long-term viability as an institution. The importance of this phase is not limited to the start-up–BI, but spreads to the whole entrepreneurial ecosystem and the surrounding region [16,17]. Nonetheless, despite these arguments in favor of a careful selection process, the BI can relax its selection criteria in periods when there are very few applicants to avoid problems of unused capacity [18]. An adequate selection impacts the performance of the incubated companies and, ultimately, the performance of the incubator itself. Moreover, the importance of studying this topic derives from the positive impact that greater efficiency in the selection of start-ups would have not only on the start-ups and BI themselves, but also on innovation and regional economic development.

Despite this importance, the literature on the BI selection phase is fragmented and not aligned with the effective selection practices of BI managers, since they often use processes that are not structured in terms of selection criteria [19] and make decisions based on non-rational elements [20]. The literature on the selection decision of BI is predominantly based on mathematical models such as Multiple Criteria Decision Making (MCDM) [21,22] or multi-objective models [23], which, due to their sophistication, may not represent the level of rationality followed by BI managers who mainly opt for intuitive, non-systematic decision-making processes [20,24]. In addition, there are gaps in the literature in what refers to the impact that BI’s strategic vision has on this selection phase, as well as regarding the lack of a framework encompassing the various dimensions of this selection phase in a single model or studies considering the perspective of start-ups, since they are the first ones to play this selection “game” when choosing a BI to apply.

In response to these gaps, the purpose of this study is to map and systematize studies on the selection of start-ups by BIs, to develop a conceptual model for this selection phase, and to delineate promising directions for future research within the framework of this proposed model and for the BI selection phase in general. This study focuses on the selection of early-stage start-ups, which typically fall within the category of micro or small enterprises. Although the conceptual model is deliberately developed in broad terms to encompass different types of BIs, economic sectors, and regional contexts, it is particularly relevant for BIs targeting this type of enterprise. The relevance of this study stems from the absence of theoretical approaches in literature on the subject of the BI selection phase, namely a conceptual model that encompasses, systematizes and integrates the various dimensions of the BI selection phase. Such a model would address a theoretical gap in the literature and guide scientific research on the subject of the BI selection phase and contribute to rationalizing and strengthening the practical actions of BI managers.

This paper is divided into six sections, starting with this introduction. The second section presents the research methodology used for the systematic literature review on the BI selection phase, which is presented in the third section, detailing the studies focused on this topic and their contributions. The fourth section presents a theoretical model based on three dimensions in which the contributions of literature on this subject can be framed: selection strategy, selection criteria and selection decision. In the fifth section, several directions for future research are proposed, framed not only in a general perspective on the selection phase, but above all in each of the dimensions present in the model. The final section presents the conclusions of this study as well as its theoretical and practical implications and limitations.

2. Research Methodology

A systematic literature review was used to consider the suitability of this method to synthesize the literature in these areas related to entrepreneurship [25]. A systematic literature review uses an explicit and systematic process [26] and allows previous work to be synthesized within an integrated framework [27], allowing for a more impartial synthesis of research through a predefined research strategy that can be evaluated and replicated [28]. The review was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses PRISMA 2020 guidelines (Supplementary Materials). The review protocol was not registered in a public repository. The PRISMA approach [26] was used to carry out the systematic literature review, which aligns with previous systematic reviews in studies related to start-ups and entrepreneurship [29,30]. The PRISMA model used, illustrated in Figure 1, was adapted from [26], and it is based on three stages: identification, screening and inclusion.

The identification stage identified the databases, keywords and search queries best suited to the intended collection of articles. The Scopus and Web of Science databases were used for their comprehensiveness and quality assurance. The choice of keywords and search queries sought not only to include articles on the various types of BI, such as accelerators or university incubators, but also to specifically combine these types with the start-up selection phase. Therefore, the search query used was: (“selection” OR “screening”) AND (“business incubat*” OR “university incubat*” OR “science-parks” OR “accelerators” OR “technology cent*”). This search was conducted in June 2025 and obtained an initial total of 7232 records. Also at this stage, filters were applied in order to obtain only articles from relevant areas, as well as a filter for articles written exclusively in English. After applying these filters, the total number of records was 444.

The screening phase reduced the number of articles through two stages. In the first stage, by reading the various abstracts, the records corresponding to articles whose subject matter did not fit the topic under investigation were removed. In a second stage, in order to guarantee the quality of the final records, after reading the full texts, records relating to articles not published in journals were removed, except for those which were, according to the researcher’s criteria, highly relevant and of sufficient quality. The total number of records obtained at the end of this phase was 21.

The final stage of this process involved reading the full articles in order to identify new articles that were not retrieved in the previous stage. The literature review that each article presented allowed the identification of new articles that were highly relevant to the subject under study and that were not present in the initial list. This process was repeated for the new articles until the point where all the relevant articles that were already part of the final list were reached. This stage allowed us to add 10 relevant publications that did not emerge from our database search.

A total of 31 articles published in 27 journals between 1987 and 2025 were considered for review.

Table 1. Search protocol followed.

Keywords	Selection, screening, business incubators, university incubators, science-parks, accelerators and technology centers
Search fields	Scopus: Article title, Abstract, Keywords Web of Science: Topic (Article title, Abstract, Author keywords)
Databases	Web of Science and Scopus
Search keywords	(“selection” OR “screening”) AND (“business incubat*” OR “university incubat*” OR “science-parks” OR “accelerators” OR “technology cent*”)
Data range	No range applied
Search date	June 2025
Filters	Area: Scopus subject area (Business, Management and Accounting OR Decision Sciences OR Economics, Econometrics and Finance) Web of Science categories (Management, Business, Engineering industrial, Operations Research management sciences, Economics, Mathematics applied) Language: English

provides an overview of the search protocol followed in this stage of the systematic literature review.

3. Results

The detailed analysis of the articles identified in the systematic literature review allows a description of their contributions to the BI selection phase, and the results can be framed into three distinct categories. As summarized in

Table 2. Mapping of literature contributions in the dimensions of the proposed conceptual model.

Article	Selection Strategy Contribution	Selection Criteria Contribution	Selection Decision Contribution
[31]	Addresses the level of restrictiveness of selection standards advocating an intermediate level of restriction.	_	_
[14]	_	Identification of 6 critical factors for each of the first 2 questions.	Presents a model based on a 3-stage decision tree represented by 3 questions.
[32]	_	Identification of 3 categories of critical success factors.	_
[33]	_	Identification of 10 entry policy elements.	_
[16]	Addressed the level of concentration in the selection dimensions advocating a balanced selection process.	Framed the selection practices in 3 criteria-based dimensions.	_
[6]	Divides between (1) selection focused primarily on the idea and selection focused primarily on the entrepreneur or team and (2) flexibility or strictness in the process of applying the criteria.	_	_
[34]	_	Identification of 4 main components related to selection performance.	_
[35]	Considers the fit between the potential tenant and the incubator.	_	Presents a MCDM model using AHP.
[13]	_	Identification of 3 critical criteria-based dimensions in a total of 4.	_
[23]	_	_	Presents an interactive multi-objective incubatees selection model.
[36]	_	_	Presents a decision support system based on fuzzy rules.
[15]	_	Identification of 8 and 4 criteria for initial and later selection phases, respectively.	_
[20]	_	_	Presents the selection decision as a non-rational one due to the use of heuristics and biases by decision-makers.
[24]	_	_	Concludes that the selection decision is subject to various cognitive biases.
[37]	_	Identification of the most used and valued selection variable for accelerators.	_
[38]	_	Identification of main selection criteria for accelerators.	_
[5]	_	Identification of 12 critical criteria for commercial accelerators.	_
[21]	_	_	Presents a novel hesitant fuzzy linguistic decision-making method.
[22]	_	_	Presents a MCDM model that combines AHP and TOPSIS.
[39]	_	Considers 14 sub-criteria relating to 5 main criteria.	Presents an MCDM that integrates Dempster–Shafer theory into a probabilistic linguistic setting.
[40]	_	Identification of 5 main categories of criteria.	_
[41]	_	Identification of the most prominent entry into business incubators precursors among a total of 11 precursors.	_

, these results are aligned with three core dimensions of the proposed conceptual model presented in Section 4. This section is intentionally descriptive, with the aim of clearly presenting the different contributions of the studies analyzed before proceeding to a more integrative and analytical approach in the following section, where a conceptual model is proposed.

3.1. Business Incubators’ Strategic Vision and the Fit Between Applicant and BI

It is possible to identify contributions from the literature that focus on the theme of this phase of selection of start-ups by BI managers, more specifically, on the strategic positioning of BI in relation to this phase of selection and in relation to the fitting of incubatees to BI.

Kuratko & LaFollette [31] started by referring to the more restricted selection standards of publicly sponsored incubators derived from the criterion of job creation potential and the less restricted criteria of privately sponsored incubators that focused on profit potential. Kuratko & LaFollette [31] also pointed out that if, on the one hand, an unrestrictive selection process allowing any applicant to enter would cause problems, jeopardizing the survival of the incubator itself, on the other hand, an overly restrictive process could lead to the selection of only lower risk and older companies, which could hinder the process of growth and job creation. These findings underscore the importance of developing a clear rationale to guide the balance between more open and more selective selection strategies.

Somsuck & Teekasap [35] identify the need for suitability between the potential tenant and the host as one of the dimensions considered relevant for this selection process. This focus on entrepreneur characteristics is particularly relevant, as these have been shown to significantly influence the likelihood of selection for BI [42].

In terms of empirical comparison of selection practices, Aerts et al. [16] described the selection practices of European business incubators in 2003 and compared these results with those of American incubators in the 1980s. Based on the work of Lumpkin & Ireland [32], Aerts et al. [16] conducted a survey that framed the selection practices in three dimensions (financial factors, team and market) and evidenced that a greater concentration on one screening dimension is related to a higher failure rate and that the tenant survival rate is positively related to a more balanced selection process.

Emphasizing the need to place the selection phase in a strategic dimension and perspective, Bergek & Norrman [6], in an empirical investigation of 16 Swedish incubators, presented a framework for the selection phase that divides between selection focused primarily on the idea and selection focused primarily on the entrepreneur or team. It also presents a further division based on flexibility or strictness in the process of applying the criteria in the selection phase, thus identifying one approach based on the selection of a small number of start-ups with enormous potential (“picking-the-winners”) and another type of selection based on a more flexible selection criterion and which relies on the natural selection made by the market over time (“survival-of-the-fittest”). The combination of these dimensions thus results in a total of four selection strategies: survival-of-the-fittest and idea; survival-of-the-fittest and entrepreneur; picking-the-winners and idea; picking-the-winners and entrepreneur. These results highlight a critical choice inherent in the BI selection process, which will be transversal to the various stages of this selection process: rigor versus flexibility. This choice is associated with a compromise between structured objectivity (e.g., MCDM models) and adaptive pragmatism (e.g., heuristic adjustments).

3.2. Critical Elements and Variables in the Selection Phase

From the literature review, a series of articles were identified that present a set of criteria considered relevant to the selection process, acting as determinants of the success of candidates for the incubation program.

Following an approach based on structured models or sets of criteria, Merrifield [14] proposes, interconnected in a model based on a three-stage decision tree represented by three questions, the use of six critical factors for each of the first two questions: sales profit potential, political and social constraints, growth potential, competitor analysis, risk distribution and industry restructuring in the case of the first question and availability of capital, manufacturing, marketing and distribution competence, technical support, availability of components and materials and management in the case of the second question. Mian [33], in a study focused on the development of new technology-based companies and based on a sample that included three public universities and three private-university sponsored facilities, identified 10 entry policy elements: technology-based start-up, firms with high growth potential, strategic business plan developed, qualified management team, commercializable product/process/service, existing cash flow stream, manufacturing firm preference, ability to pay the rent, fit with the university resources/mission, investor’s commitment.

Some authors group criteria into broader categories—Lumpkin & Ireland [32] presented, in a survey conducted among incubator managers, three categories of critical success factors: financial ratios, personal characteristics of the management team and market factors, each containing several indicators. Also using an approach that tries to combine various criteria into main criteria categories, Fararishah et al. [13], in a study based on a survey questionnaire, analyzed the strategies used by Malaysian information and communication technology incubators in selecting potential incubatees. Wu & Liao [39] propose a multi-criteria decision-making model based on 14 sub-criteria relating to five fundamental criteria: social benefit, policy and matching, management ability, innovation ability and income ability. The survey questionnaire considered managerial characteristics, market characteristics, product characteristics and financial characteristics as the main items, and the empirical results indicated market characteristics, product characteristics and financial characteristics as the main factors. The dynamic and longitudinal nature of the selection process was demonstrated, notably through observable changes in selection criteria during the different selection phases. In a study based on a dataset of start-ups that applied to Southeast Asia’s first seed accelerator, Yin & Luo [15] analyzed the shift in the criteria used in the start-up selection decision between an initial selection phase among several start-ups and a final selection phase with few start-ups left. The results of the study showed that at an early stage of the selection process, eight criteria related to market attractiveness, product feasibility, product advantage and team competence were critical to the selection decision, while at a later stage, there was a shift to four criteria related to expected return and growth potential. Finally, while maintaining a categorical approach, Hackett & Dilts [34] in a study that collected data from 53 active incubators in the US, tested the selection performance construct through exploratory factor analysis, resulting in four components: star component (related to the potential to attract investment, exit options, product characteristics and intellectual protection), market (related to long-term market growth potential, market size and accessibility to its customers), differentiation (uniqueness of the product and competitive advantage over competitors’ products), and manager (related to previous experience).

Instead of approaching the selection process from the incubator’s perspective, Capatina et al. [41] invert the perspective by exploring the issue from the start-ups’ point of view. More specifically, using correlation-based analysis, machine learning techniques and fsQCA, it examines the causal relationships between 11 antecedent conditions and the degree to which these antecedent conditions affect entry into business incubators in Italy and Romania. The results present entrepreneurial team features, available financial resources, debts of potential incubated companies, and technological transfer from university/research centers as the most important precursors of entry into business incubators, with a difference between the two countries considered.

It is also possible to identify four studies that pinpoint relevant elements and variables for this selection phase, with the particularity of using case study approaches and/or focusing on a specific type of BI—accelerators. These accelerators differ significantly in terms of the length of the incubation period, providing another perspective on this topic. Despite this, the results are in line with the variables identified by studies that consider classic BIs. Mariño-Garrido et al. [37], in a case study based on an accelerator in Spain, considering an initial sample of 309 candidate projects, analyzed the most relevant variables for the selection of 15 of these projects. The extent of team consistency and speed of acceleration were the variables most used to assess an entrepreneurial project, while the most valued entrepreneurial skills were leadership and creativity.

Furthermore, qualitative approaches have been employed in studies focusing on accelerators. These approaches enhance the robustness of previous results and support: (1) the existence of categorizable selection criteria; and (2) a significant overlap in core evaluation criteria across studies. Through in-depth qualitative case studies of 10 accelerators in Turkey, Beyhan et al. [38] explored the selection process in accelerators. Through interviews, two main selection criteria emerged: the entrepreneurial opportunity and the team characteristics. Qualitatively investigating the selection criteria of nine European accelerators that consider social and/or environmental impact in comparison with the criteria used by accelerators focused on the economic dimension, Butz & Mrożewski [5] identified significant differences between these two types of accelerators. Customer affordability, product maturity, prior start-up expertise, feedback mechanism, sales and distribution were the criteria considered critical specifically for commercial accelerators. Also, with the aim of selecting start-ups to an accelerator, Mohammadi & Shafiee [40], using the Fuzzy Delphi method and aggregation of opinions of five experts grouped, an initial input of 43 criteria, applying the affinity diagram method, into five main categories of criteria: business idea, technical and operational criteria, financial feasibility criteria, market feasibility criteria and start-up team criteria.

3.3. The Final Selection Decision

There is also a final dimension of studies in the literature about the phase of selection of start-ups by BI managers that focus on proposing decision support models for this final phase of the selection process, which involves deciding which start-ups to include in BI and which to reject.

In addition to the model proposed by Merrifield [14], which uses a three-staged decision tree, there is a set of proposed models for this selection decision based on MCDM solutions. Oliveira et al. [22] propose a model for evaluating start-ups to be integrated into an incubator, by ranking them in order of selection priority. A combination of MCDM methodologies is used, namely the Analytic Hierarchy Process (AHP), Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and two normalization procedures. This model is applied to the evaluation of six business projects to be incubated. In another study using MCDM methodologies, Somsuck & Teekasap [35] applied an AHP methodology to determine the candidates with the highest probability of commercial success. In another approach that tries to deal with the multiple criteria to be taken into account, Wu & Liao [39] proposed an MCDM model based on 14 sub-criteria relating to five fundamental criteria. The model is validated in a case study of the selection of two SMEs from a total of five candidates for entry into a Chinese business incubator.

The introduction of fuzzy elements in MCDM models was followed in the studies of Lin et al. [21] and Arsenyan [36]. Fuzzy elements in MCDM help to resolve the uncertainties, imprecise data, and subjective judgments involved in these decisions. In order to overcome the disadvantages of MCDM, such as the instability of the decision results and the impossibility of coming up with suggestions for improvement for the non-optimal start-up companies, Lin et al. [21] propose a novel hesitant fuzzy linguistic decision-making method allowing all alternatives to be ranked and non-optimal alternatives to be improved. To deal with the hesitant attitudes of experts, the concept of hesitant fuzzy linguistic term sets is used. Based on the knowledge of an expert from a Turkish start-up ecosystem, Arsenyan [36] proposes a decision support system based on fuzzy rules, quantifying linguistic expressions through the use of three-level linguistic variables (high, medium, and low).

A different approach to this final decision, based on an interactive multi-objective model, is proposed by Wulung et al. [23]. This model incorporates incubator manager orientation, iterating over a financial orientation path and a social orientation path, dividing mainly between candidates with higher profitability prospects and candidates with higher worker absorption prospects.

Finally, Ahmad [20] studied five client selection processes from start to final decision in two business incubators with the aim of investigating the client selection process at incubators to understand the role of specific heuristics and biases in decision-making. The conclusions drawn from this study are that the selection decision, being made by human agents, is made with recourse to non-formal mechanisms such as heuristics and biases, tainting this process that one would expect to be totally rational. This conclusion is supported by Barlach & Plonski [24] who, through a quasi-experiment design involving directors or managers of Brazilian accelerators, concluded that the selection process is subject to various cognitive biases, including the Einstellung effect, characteristic of mental rigidity. These qualitative results, together with the introduction of fuzzy elements into quantitative methods, suggest that the literature has recently sought to incorporate the decision-making characteristics of human agents whose time constraints or cognitive limits are acknowledged.

4. Proposed Conceptual Model

The literature on the BI selection phase is fragmented and not aligned with the selection practices of BI managers, since they use unstructured processes in terms of selection criteria [19] and make decisions based on non-rational elements [20,24]. This decision-making based on heuristics or intuition is supported by the literature [43] and may even offer better decisions than the models proposed by the current state of the art [44].

The results obtained in the previous section support the division of the selection phase into different stages. Each of these stages is consistently defined by a series of studies, thus promoting the creation of a conceptual model that robustly captures the contributions of the literature. Despite this intra-stage validation, the need for a holistic view that integrates the different stages into a global model emerges from the literature review. The lack of a sequential or causal logic between stages is one of the problems identified in the results. Such clarification would be useful for BI practitioners, with implications, ultimately, for the quality of selection decisions. The need to expand research beyond a stage-centered view to include the influence that external factors may have on each stage is another point identified.

The results also identify the lack of integration of the selection phase into the overall incubation process. Whether from a longitudinal or retrospective perspective, using the outcomes of the incubation process as a way of measuring success or improving the selection phase would be an added value. The proposal of this conceptual model for the selection phase could be a catalyst for studies that integrate this selection phase into the overall incubation process.

The developed model aimed to gather and systematize the literature on this selection phase to not only map the state of the art but also to identify directions for future research on this subject. A comprehensive reflection on the various studies in the literature on the theme of the selection phase in BI and its intersection with the practices that are in fact followed by BI managers allows the development of a conceptual model that attempts to systematize this fragmented research topic, thus making it easier to understand which approaches and steps are followed during this selection process.

This conceptual model (see Figure 2) is divided into two perspectives: (1) the start-up perspective and (2) the BI perspective. The inclusion of the start-up perspective in the proposed model is intended to build on the fact that the literature presents an underexplored topic by not considering the selection process from the perspective of start-ups. Before we are faced with the problem of selecting start-ups to join a BI, start-ups also go through a process of selecting the BI to which they are going to apply. Therefore, any proposal for a model for this phase of selection of start-ups by BI should include this upstream perspective of the start-ups. In one of the rare studies from this perspective, Isabelle [45], through two surveys of managers and users of incubators and accelerators, identifies five key factors that start-up managers should take into account when choosing which incubator to apply for: stage of venture, fit with BI’s mission, selection and graduation policies, services provided, and network of partners. The proposed model, from this perspective, considers two fundamental phases that start-ups go through and have different motivations: motivations for entering an incubation program and motivations for applying to a specific BI.

Regarding the perspective of BI, derived from the analyzed literature, it includes three fundamental stages for the selection phase: selection strategy, selection criteria and selection decision, and identifies key issues in each stage. The various scientific works focused on the selection phase of BI that were presented in the previous chapter could, in a less granular view, be considered as all giving a direct answer to the problem of selecting firms to be incubated; however, a more detailed analysis reveals that the various studies provide answers to different questions related to the selection phase, thus allowing to frame these studies in three distinct stages of the selection phase. The proposal of a conceptual model for this selection phase is a novelty for the literature and attempts not only to organize and systematize the contributions made so far, but also to outline avenues for future research. Table 2 maps chronologically the various articles identified in the data collection process, placing them within the three dimensions of the proposed model.

4.1. Selection Strategy

This is the first element of the proposed conceptual model and builds upon the work of Aerts et al. [16], Bergek & Norrman [6], Kuratko & LaFollette [31] and Somsuck & Teekasap [35] and contributes to the selection process insofar as they provide indications for the alignment between the overall strategy and characteristics of the BI and its repercussion on the selection process, thus deserving the designation of selection strategy.

The typology and structure of the incubator, the level of intervention—from the laissez-faire to the strong intervention incubator [6]—the level of specialization in a particular industry [46], the funding sources and the objectives (which can range from a greater focus on regional development to maximizing financial return for investors) are all fundamental elements for characterizing the nature of BI and have a significant impact on BI’s strategy and consequently on the selection phase of BI. This BI nature emerges from the initial reason for BI’s existence and is shaped and evolves as the relationships of influence with the various stakeholders are established. It is also associated with the transposition of the perception that BI managers have of the nature of BI itself into practical day-to-day actions. It is also within this dimension that elements of BI’s mission can be incorporated, namely in terms of its role as a promoter of social and environmental sustainability, which these organizations should embrace by integrating sustainability principles into their strategic identity and selection logic.

These elements do not directly constitute objective criteria for assessing the potential of firms applying for the incubation program (although they may indirectly constitute exclusion criteria), but they do condition this process of defining criteria and their application. The concern with the fit between the incubator and the incubatee [35], a balanced level both in terms of restrictiveness of selection standards [31] and in terms of concentration in the criteria-based selection dimensions [16] and selection processes that can be focused on the idea or can be more focused on the entrepreneur or team [6] are the main contributions of the literature to this first dimension of this conceptual model.

4.2. Selection Criteria

Several studies [5,13,14,15,16,32,33,34,37,38,39,40,41] belong to the dimension most present in research on this topic, on the selection phase of BI—selection criteria. This set of articles focuses on identifying a set of criteria that can serve as determinants or predictors of the future success of start-ups applying for the incubation program. These contributions do not directly incorporate the strategic constraints imposed by the phase described above, nor do they include the upstream process of applying/considering the criteria identified in the decision-making process. Although several studies have proposed selection criteria, there are four main categories of criteria that stand out and are present in most studies: market-based criteria, product-based criteria, financial-based criteria, and manager-based criteria [13,32,34]. Within each of these main categories, various criteria are proposed, making this dimension of the model, called selection criteria, the most prominent.

In addition to the consensus that seems to exist on the fundamental categories of criteria for determining the success of start-ups before they enter the incubation process, there is also a consensus on the logic behind the consideration of each of these categories of criteria. Market-based criteria start from the premise that the external environment, namely the characteristics of the market in which the start-up operates (number of competitors, their respective market shares and their stability, growth rate, levels of failure), strongly influences the success of candidate start-ups. Similarly, the preponderance given to product-based criteria is due to the power that a product’s differentiating and innovative characteristics have in creating competitive advantages, attracting clients from the competition, or even creating new markets without competition (following the logic of the blue ocean strategy). The logic behind the consideration of financial-based criteria results from the importance attributed not only to the amount and management that the start-up makes of its current financial resources, but above all to the future expectation of return on investment measured in terms of cash flows. Finally, the manager-based criteria emphasize the role of the start-up manager in contributing positively to the success of the organization. The manager’s knowledge, leadership skills, personality and beliefs are some of the elements considered to influence the success of start-ups, thus justifying their consideration in this selection phase. Recent studies also highlight the emergence of sustainability-related criteria in the selection process, particularly in incubators and accelerators that incorporate a strategic focus on social innovation or environmental impact [5].

One of the reasons for the quantity and consensus in terms of criteria that this dimension shows is its proximity to other research topics where the identification of criteria and determinants of entrepreneurial success is fundamental. Topics such as business angels, venture capitalists, determinants of firm growth, determinants of firm performance, and start-up investments are all close to the topic of BI’s selection of start-ups, thus justifying the high level of development of the literature on this dimension.

4.3. Selection Decision

Finally, several studies [14,20,21,22,23,35,36,39] focus on a fundamental step in the process of selecting firms to join the incubator, which is the decision-making process, mechanism or methodology used to make the selection decision, incorporating the strategic indications of the BI—the first dimension of the model—and applying/considering the criteria used by the BI manager—the second dimension of the model. This third dimension, entitled selection decision, which also involves evaluating and assigning the relative weight of each criterion for decision-making, and its necessity derives from the nature of uncertainty and incomplete information that is associated with these complex decision-making processes. Given this complexity of alternatives, criteria and their respective weights, constraints and uncertainties, the emergence of mathematical models to aid this decision was natural and typically fits into a multi-criteria decision, clearly differentiating this third dimension from the previous two.

The literature on this dimension uses a series of methods based on mathematical models, ranging from MCDM models [22,35,39], MCDM with the inclusion of fuzzy elements [21,36], decision trees [14] to multi-objective mathematical models [23]. It is worth highlighting the existence of two studies [20,24] that consider the non-rationality of managers in these selection decisions and which, in our view, constitute the right path for future research in this area. The proposal of recent models that include fuzzy elements, or these studies by Ahmad [20] and Barlach [24] that highlight the not always totally rational decisions of BI managers, seem to indicate a bridge between the models proposed in the literature and the models actually used in practice. The importance of this dimension should not be underestimated since, ultimately, the entire selection phase (including the two dimensions presented above) depends on this final decision. In the end, a poor decision can completely counteract the contribution made by the other dimensions.

4.4. Schematization, Relationships, and Influences of the Proposed Model

The model presented in Figure 2 is not just a descriptive overview of the literature. It contains new elements that are associated with the vision of the selection phase as a sequenced and connected process involving these three dimensions, which are not independent of each other.

This model also includes two new interrelated concepts: BI nature and exclusion criteria. The concept of BI nature derives from the existential reason for BI, its objectives and structural characteristics. The existence of a BI that specializes and hosts start-ups from a single sector [46], or a BI that focuses on hosting start-ups with social or environmental priorities [38,47] are examples of elements that make up the nature of a BI and which have an impact on the strategic dimension of the selection phase. This nature of the BI is also related to the fundamental objectives of the BI itself, which can vary from maximizing profit for the BI itself or for the incubatees (example of private incubators), maximizing regional impact in economic terms (example of public funded BI), maximizing social/environmental gain, or maximizing technological impact, among others. Linked to this element of the nature of BI is the element of exclusion criteria. This involves a series of exclusion criteria (as opposed to the remaining criteria, which have a selection function) that result from the application of BI nature in the selection phase. A clear example of this link between BI nature and exclusion criteria is present in the case of a BI focused on a single sector. Logically, candidate start-ups that do not operate in this same sector will be excluded at an early stage of this selection phase because they do not fit in with the BI’s strategy. This BI nature element should not be confused with the alignment element between incubatee and incubator, since the latter acts in terms of aligning a set of more specific characteristics, already after a first exclusion phase derived from the BI nature element has been carried out. Even so, this element is linked to the exclusion criterion because it can constitute other reasons for excluding candidate start-ups.

It is also possible to see in the diagram the influence that the characteristics of the BI and the BI manager have on the three dimensions of the model. As previously explained, the selection strategy dimension is strongly influenced by the existential nature of BI, while the selection criteria and selection decision dimensions are more influenced on the characteristics of the BI manager. The selection criteria, although they can be influenced by the BI’s characteristics (namely in terms of exclusion criteria), are chosen and measured by the BI manager, placing him in a position of greater influence in this second dimension of the proposed model when compared to the influence of the BI characteristics. The selection decision dimension is the one where the BI manager’s influence is most noticeable and impactful. This is quite clear considering that, even when using rational decision models proposed by the literature, these require: (1) the BI manager to choose the specific decision model, (2) the BI manager to value (assign weights) the criteria considered and (3) the decision-maker (BI manager) to assign values to each alternative-criteria pair. The study by Ahmad [20] focuses precisely on the BI manager and the biases and irrationalities they can have on the decision-making process.

Finally, Figure 2 shows three new elements, categorized as external and past factors, which are linked to the selection phase. These new elements are named results of previous selection phases, stakeholders and economic situation and are intended to contribute to a gap identified in the literature, which is the impact that past experiences and the external environment have on the BI selection phases. Indeed, the experiences, mistakes and performance after the incubation process of each start-up that has gone through the BI selection phase are fundamental elements for a new selection phase. The BI manager’s perception of past decisions can lead him to modify the strategy, consider new criteria or change decision-making processes, thus justifying the link between this element and the selection phase as a whole. The variety of stakeholders, the power relationships with BI and the evolution of these relationships over BI’s lifetime are factors with great potential to influence the selection phase. In particular, significant changes in stakeholders with responsibility for financing the BI constitute a risk factor that may lead the BI to completely redesign its internal processes, including the selection phase. The element of the economic situation is another external influencing factor with a major impact on the selection phase. Changes in the level of public funding for BI, lack of access to capital, periods of recession, lack of access to skilled labor or high rates of start-up failures are some of the economic variables that, in addition to influencing the BI selection phase (adverse economic conditions can lead to a lower quantity and quality of candidates and a loss of hosting capacity), influence BIs as organizations.

The mapping of the most relevant studies on the selection phase of start-ups by BIs framed in the three dimensions of the proposed model presented in Table 2 allows for the identification of research in this area since 1987 with a more pronounced growth in research in recent years. With regard to the dimensions of the proposed model, the selection criteria dimension is the one with the highest number of contributions, with around half of the contributions, while the selection strategy dimension is the one with the least preponderance. There is also an upward trend in contributions to the selection decision dimension. This growth could be related to advances in the area of decision models, namely with the inclusion of fuzziness in these models, which are precisely approaches that are present in some of the articles presented here related to this dimension [21,36]. Finally, the existence of only four articles that contribute to more than one dimension at the same time supports the idea that there is a lack of studies that approach this selection phase in a comprehensive way and that integrate all the stages.

5. Discussion and Future Research

One of the main purposes of this study is to identify directions for future research on the selection phase of business incubators based on the proposed conceptual model. Therefore, sub-topics covering the three dimensions simultaneously are initially proposed and discussed, and subsequently, sub-topics relating to each of the three dimensions present in the proposed conceptual model are also proposed and discussed.

5.1. Business Incubators’ Selection Phase

From the analysis of Table 2, which summarizes the various contributions of the literature to this topic, it can be seen that there is no study/model that encompasses all three dimensions simultaneously. Even models that encompass two dimensions simultaneously are underrepresented [14,16,35,39]. Bearing in mind that none of the dimensions can be left out of the selection phase since BI managers, more or less directly, address all three dimensions during this selection phase, research that encompass all three dimensions would have great potential for BI, which would follow more robust, comprehensive selection processes that guarantee alignment with the incubator, with natural implications for BI performance. In addition to this lack of studies dealing with the selection phase as a whole, there has also been a shortage of contributions on the operational elements of this selection process, such as the process of communicating vacancies, the application process, the information required to assess candidates, the interviews and the process of validating the information initially provided or the process of operationalizing entry into the BI. These elements, although operational in nature, need to be clarified from the point of view of literature and would have important contributions for BI managers—helping to unravel the ‘black box’ of incubation process [34].

Another sub-topic that transcends the three dimensions considered relates to the impact of the selection phase on the performance of BIs. BIs offer a series of tangible and intangible services to incubatees, ranging from facilities and administrative services to access to funding and training [2], which are naturally expected to contribute to incubatees performance. The literature on the performance of incubated firms versus the performance of similar firms that have not been through incubation processes, although not unanimous—some studies even indicate negative effects [48,49], tends to identify a higher performance of firms that have been through incubation programs [50,51,52]. However, the literature does not yet allow for an understanding of the contribution and weight that each element of the incubation process has for this greater performance of the incubatees. More specifically, the impact of the selection phase on the final performance of the incubated firms and consequently of the incubator itself remains to be seen. Is the selection phase more or less important than the services provided by the incubator during the incubation period? Do BIs that put more emphasis on the selection process perform better? Empirical work on this sub-topic could make it possible to answer these questions and get a more precise picture of the importance of the selection phase for BI.

A growing number of studies were identified focusing on specific types of BIs, such as accelerators, university-based incubators, or technology-based-incubators, with a particular emphasis on accelerators [5,38,40,53]. However, the literature on BIs does not capture some of the particularities that accelerators and other specific BI typologies have in relation to the classic BI typology, such as the differences in terms of the incubation period or the life stage of the start-up they target. To tackle this problem, some researchers have focused their research on specific types of BI, thus creating sub-spaces of research within the topic of BI. Therefore, understanding the differences in terms of selection processes between the various types of BI would make a theoretical and practical contribution, clarifying something that the literature currently fails to answer. One of the most relevant explanatory questions would be to understand how the selection processes followed by the different types of BI capture and encompass the needs, objectives, and particularities of these specific BI typologies.

There is also room for future research, addressing one of the gaps identified in the literature, in the selection process from the perspective of start-ups. The motivations that lead start-up managers to be interested in and seek out incubation programs, as well as the motivations and criteria that lead them to apply to a particular BI over another, are points of interest to better understand this process upstream of the selection phase by BIs.

5.2. Selection Strategy

As discussed above, this dimension of the selection strategy encompasses the alignment between the BI’s strategy and characteristics with the selection phase. This assumption allows to identify three points for future research: (1) understand how the selection models used by BI managers integrate BI strategy, particularly in terms of objectives—which could be prioritizing regional economic development, return on private capital or even social/environmental impact [54]; (2) understanding the extent to which the various stakeholders and their power relationships with BI influence the selection strategy of BI—building upon theories that combine stakeholders and their power-related characteristics [55]; and (3) to study the consequent performance of BIs that opt for selection models with a strong emphasis on the alignment between BIs and incubatees in comparison with BIs that choose to give less importance to this factor, building on Kuratko’s [31] view that a standardized alignment can be a risk, if the specificities and stakes of the BI are not of the highest quality.

5.3. Selection Criteria

With regard to the dimension of the selection criteria, the literature is homogeneous and is based exclusively on identifying criteria that can serve as predictors of firms with a high level of growth and performance [37,38,41]. This search for criteria that predict future performance is not exclusive to BIs. In fact, several generic studies on small firm growth/performance determinants [56,57], as well as studies on the selection process of start-ups by venture capitalists [58], can be found in the literature with objectives very similar to those of BIs. Therefore, possible future research could focus on bringing together the criteria considered for each of these areas and checking which criteria are missing or neglected in BI research.

Although addressed in the literature, criteria related to factors external to the company, namely the characteristics of the industry in which they operate, remain little explored. Bearing in mind the risk and fragility that these start-ups face when up against solid competitors in the market, with economies of scale, larger client portfolios, easier access to finance and powerful relationships with clients and suppliers, the characteristics of the industry in which these start-ups operate can in no way be overlooked as a fundamental criterion for selecting firms to be incubated. Although it has been explored extensively, the work of Porter [59] can still provide good support for identifying new selection criteria that address this point.

One of the new aspects presented in the proposed model was the influence of the BI manager’s characteristics on the selection criteria and on the selection decision. Focusing on this dimension of the selection criteria, understanding this influence and its impact would be relevant in helping to understand the role of the BI manager for the incubator. Ultimately, it would be a question of whether the BI manager’s characteristics, such as knowledge, previous experience in the incubator or psychological traits—building on that establishes traits as predictors of leadership behavior [60]—lead to the attribution of different levels of importance in certain categories of criteria.

5.4. Selection Decision

An analysis of the various studies in this category of selection decisions shows that research has focused almost entirely on mathematical, rational, and systematic models. Despite the exhaustive literature review, only Ahmad’s [20] study recognizes the existence of obstacles in the application of these methods, meaning that in reality, this decision is made with non-total rational elements. Understanding the extent to which BI actually uses the models proposed in the literature or opts for more intuitive models would be a relevant starting point for studying this topic. Related to this, understanding the extent to which the models used by BI differ in terms of results from the models proposed by the literature would be important. In fact, it is not a given that the models proposed by the literature offer better results than the models used by BI managers, as there is evidence that, in some cases, “fast-and-frugal” heuristics often outperform complex models in real-world decisions [44].

Another topic for future research, which is related to the previous one and which would have significant implications, would be to understand the roots of this bridge between the models proposed in the literature, which provide answers with the quality derived from mathematical models, and the decision-making models actually used by BI managers, which are far from the maximum rationality they could use. The existence of literature explaining the introduction of intuition into managerial decisions may help to explain this phenomenon [43]. Is it due to a lack of knowledge of these models? A lack of ability to understand them? Because BI managers consider them of little benefit compared to the models used? Too time-consuming to apply? A lack of information about the candidate firms? Or, ultimately, is the literature formulating this selection problem incorrect since, in practice, this problem does not fit into an MCDM model? These are some of the possible hypotheses to explain this phenomenon, which, if tested, would make a strong contribution to research into this third dimension of the model.

6. Conclusions

The model developed sums up the findings that emerged from the literature by identifying three major dimensions associated with research into the selection phase of BI: selection strategy, selection criteria and selection decision. These three dimensions provide answers to different questions relating to this phase of selection of start-ups by BI. The selection strategy dimension is based on the alignment between the overall strategy and characteristics of the BI and its repercussion on the selection process. It is in this dimension that the level of intervention, the level of specialization, the structure, and the objectives of the BI are considered during the selection phase. In turn, selection criteria focus precisely on the criteria considered to select the start-ups that will be part of the incubation program, and the main categories of criteria identified were market-based criteria, product-based criteria, financial-based criteria, and manager-based criteria. Finally, the dimension relating to the selection decision concerns the process of assigning values and weights to each criterion considered and the actual decision on which candidates are selected and which are rejected. The literature on this dimension of the model is quite uniform, proposing a series of rational decision models based mainly on MCDM models.

In addition to its implications for the performance of incubators and start-ups, the results of this study also highlight the potential of the selection phase as a tool for promoting sustainable development. This concern should be embedded in the strategic identity of BIs by incorporating criteria into the selection process that value start-ups aligned with sustainability goals.

The relevance of the proposed model stems not only from the process of collecting articles and reviewing literature on this specific topic, which was according to a PRISMA approach in a systematic literature review, but above all from the consistency and clarity with which the various contributions to the literature are grouped into one of the model’s three dimensions. These three dimensions are distinct from each other and relate to different stages of this selection phase, thus justifying the formation of these clusters and the fragmentation in the literature, which was pointed out as one of the elements in support of the relevance of this study.

Having presented the model, the mapping of the studies on this topic, considering the model’s three dimensions, showed that selection criteria had the greatest preponderance in terms of research, with the selection strategy being the dimension with the lowest number of publications. An upward trend in research contributions to the selection decision dimension was also shown. Overall, the existence of 22 relevant articles on this topic indicates that there is still considerable ground for research.

The proposition of ideas for future research was one of the main objectives of this work, which not only presented a series of ideas focused on the selection phase of BI start-ups in a generic way but also presented ideas for future research focused on the three dimensions of the model separately. The lack of generic and practically oriented models/frameworks for the selection process that cover the various dimensions of the model, the lack of information on the operational processes followed in this selection phase or the need to explain the differences in selection processes between the various types of BI are gaps identified for this selection phase in general and which would benefit from future research. The perspective of start-ups selecting BIs to apply to must also be considered. With regard to specific research proposals for each of the model’s dimensions, understanding the extent to which the various stakeholders and their power relations with BI influence the selection strategy, the influence of the BI manager’s characteristics on the selection criteria or understanding the extent to which BI actually uses the decision models proposed in the literature instead of opting for more intuitive models are some of the proposals for future research within the framework of this model.

The fact that, after applying the PRISMA protocol, there were only 31 articles on this topic to analyze can be seen as a limitation of this model. Although this number of articles was sufficient to create the proposed model, the descriptive analysis that follows would benefit from more research on this topic, consolidating the results and conclusions drawn. However, it is precisely the small number of studies in this area of research that justifies the relevance of this investigation in trying to conceptualize this area of research. Other possible limitations of this study include the absence of segmentation by type of BI, specific economic sector, or regional context. This methodological choice was intentional, given the scoping nature of the research, aimed to ensure that the proposed model retained broad applicability across different sectors, contexts and BI types.

The systematization of the literature, structured in the three proposed dimensions, has theoretical implications by responding to the fragmentation that existed in the literature on the subject of the BI selection phase. Additionally, it provides a clear framework that guides future research directions identified by the proposed model. Moreover, the mapping of approaches proposed by the literature in terms of criteria to be considered in selection processes and decision models to be used in the final selection phase constitutes a relevant database for BI managers. This set of resources has practical implications for BIs and their managers by helping to increase the effectiveness of these selection processes, which contributes to greater efficiency in allocating BI resources exclusively to those start-ups with the greatest potential for financial and economic growth. Ultimately, these practical implications benefit the start-up managers and investors, as well as the government and society.