A Bibliometric Analysis of Studies on ‘Start-up Success’ Covering the Period 1981-2019

Despite being received increasing attention from academic scholars, there have yet any review study on the topic of start-up success. This work fulfills this research gap by investigating 1554 start-up success documents collected from Scopus dataset between 1981 and 2019. Using bibliometric analysis, we reveal that the topic of start-up success only receives more attention from academic scholars since 2011 onwards. Regarding geographical distribution, the US, Germany, and the UK are the three countries contributing the highest number of start-up success related documents. Besides, it’s revealed that 305 (or 19.6%) start-up success documents were published in the top 20 journals. The co-author analysis found that the research groups of start-up success are still small and dispersed and there was a lack of continuity in the research. The science mapping identified six main topics of start-up success, including: (1) Business in General, (2) Start-up Ecosystem, (iii) Academic Start-up, (iv) Drivers of Start-up Success, (v) Resources for start-up, and (vi) Start-up Model. The study’s findings provide implications for stakeholders, academic


INTRODUCTION
Start-ups are paramount to the development of any economy. [1][2] Previous scholars identified various roles of start-ups in the economy, including driving force for modern economic development, [3][4][5][6] source of technological creativity [7][8] and engines for innovation. Start-ups are supposed to grow more rapidly, [9][10] and deal more flexibly with difficulties than conventional businesses. [9] Initiating a start-up is rarely painless. [11] Some start-ups succeed, grow quickly and exit the start-up phase, while many others fail or languish as small firms. [12][13] During the start-up step, businesses are often highly vulnerable [14] since they must face various challenges such as selecting co-founders, selecting investors, determining capital contribution within the founding team. [15] They are also confronted by cultural barriers, barriers to market entry, barriers to access to financial support, barriers to experience. [16][17] The shortage of resources such as financial capital, human capital, and social capital [18][19] may also distrust a start-up's growth opportunities. Besides, contextual factors also affect the success opportunities of start-ups. [20] There have been several previous authors investigating the topic of start-up success. For instance, Gyimah et al. [21] Hormiga and Batista; [22] Meshram and Rawani; [23] and Spiegel et al. [24] tried to introduce conceptualizations and Ahmad and Hoffmann; [25] Kessler et al. [26] Maurya; [27] Murphy et al. [28] Rompho; [29] Tehseen and Ramayah; [30] and Duchesneau and Gartner [31] endeavour to measure of start-up success. Parallelly, Abimbola and Agboola; [32] Gelderen et al.; [33] Santisteban and Mauricio; [34] and Song et al. [35] examined the various critical factors of start-up success. In addition, some extant studies aimed to look at the characteristics of founder that influence start-up success, such as psychological characteristics, the entrepreneur's personality, gender, the entrepreneur's educational level, previous start-up experience. [36][37][38][39][40][41][42][43] Regarding the scope of the study, most previous start-up success research focused on a single country such as Singapore, [44] Nigeria, [32] India [45] or the US. [46] Some approaches of industry field range as fresh juice distributors industry. [31] Thus, there is lacking an overview analysis of start-up success at a global scale. Hence, this study aims to fill this gap by conducting a bibliometric analysis on start-up success with query was conducted in September 2020. No limitation was set on searching for the earliest "start-up success"-related publications. However, concerning the latest publications, we allowed Scopus to search to the end of 2019. Regarding the languages of searched publications, following [57] we only focused on English while ignoring other languages, including Chinese, French, Spanish, Bahasa Indonesia, etc. In view of this, we admit from the outset that one of the drawbacks of our research is this ignorance.
Eventually, the following keywords were used for the search query: TITLE-ABS-KEY ( success* AND ( "start up*" OR "startup*" OR startup* ) ) AND ( LIMIT-TO ( DOCTYPE , "ar" ) OR LIMIT-TO ( DOCTYPE , "cp" ) OR LIMIT-TO ( DOCTYPE , "ch" ) OR LIMIT-TO ( DOCTYPE , "bk" ) ) AND ( EXCLUDE ( PUBYEAR , 2021 ) OR EXCLUDE ( PUBYEAR , 2020 ) ) AND ( LIMIT-TO ( LANGUAGE , "English" ) ) AND ( LIMIT-TO ( SUBJAREA , "BUSI" ) OR LIMIT-TO ( SUBJAREA , "SOCI" ) OR LIMIT-TO ( SUBJAREA , "ECON" ) OR LIMIT-TO ( SUBJAREA , "DECI" ) OR LIMIT-TO ( SUBJAREA , "ARTS" ) OR LIMIT-TO ( SUBJAREA , "PSYC" ) OR LIMIT-TO ( SUBJAREA , "MULT" ) ) In this study, the authors used the literature search guiding principles outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). [60] PRISMA is followed by many scholars to conduct systematic evaluation, such as education, [57] tourism. [49] In the field of entrepreneurship, scholars use PRISMA to make systematic assessments as. [61][62] PRISMA allows authors to make all steps in the process of quest and screening clear ( Figure 1) [47] two co-authors scanned the titles and abstracts of articles to identify their relevance with the research questions. This step was undertaken between September and October 2020. At the final step, the comparison of the two co-authors yielded a 98% agreement on the inclusion/exclusion of articles. If certain publications receive mixed opinion between the two co-authors, the two co-authors will discuss together until the agreement is reached. In this final step the authors removed 440 non-conforming documents and 11 duplicate documents. data obtained Scopus database. Bibliometric analysis is widely used in order to draw a comprehensive picture in particular research topic such as education, [47] entrepreneurship, [48] tourism [49] and social Media, [50] machine learning. [51] Bibliometrics analysis helps to capture various features from the extant literature on start-up success, including number of related articles over longitudinal perspective, geographical distribution of the authors, co-author collaboration pattern, major sources of publication, major authors, research groups or topics.
Specifically, this study intends to address the following research questions (RQ): RQ1: What is the overall volume, growth pattern, geographical distribution of publications on start-up success?
RQ2: What are the most important outlets (i.e., journals, books, book series or conferences), authors, research groups and publications on start-up success?
RQ3: What are the most important topics in the start-up success literature?

METHODOLOGY
This review used bibliometric methods to examine trends and patterns in the scientific database relating to start-up success published in the early 1980s. The bibliometric approach builds bibliometric networks based on knowledge downloaded from bibliographic databases (e.g., Web of Science or Scopus). [52] Bibliometric analysis technique review aims to understand the trends and systemic structure of the knowledge base in different disciplines. [53] Hence, differing from the traditional review method, the bibliometric analysis examines bibliographic meta-data, which seek to integrate substantive findings within a field of study. [52] Search Criteria and Identification of Sources There are two most common scientific databases, including Scopus and the Web of Science, which are often selected by previous authors for bibliometric review. [54] In this particular case, however, Scopus is more suitable for the reasons below: First, according to Hallinger and Nguyen, [55] Scopus was selected because the collection of documents for inclusion in its index uses a consistent standard. Second, Scopus has more coverage than Web of Science. [56][57][58] Moreover, in Social Sciences and Humanities, Scopus is more widely used than Web of Science. [59] The review's topical emphasis was delimited to "start-up success"; however, variations of the keyword "start-up*" were also used for search, such as "start-up*" and "startup*"-the search for documents encompassed journal articles, books, book chapters, and conference proceedings. The search Eventually, we obtained 1554 publications for final analysis ( Figure 1).

Data Analysis
Bibliographic data (authors, titles affiliations, citations, etc.) related to the 1554 start-up success documents were exported from Scopus into a master Excel file. Excel was used to perform descriptive statistical studies to chart the landscape of startup success scholarship. (e.g., growth pattern, geographical distribution, types of research papers). VOSviewer software, a bibliometric software package used for science mapping, was used to analyze data in this master Excel file. [63] Using VOSviewer software, visual maps are extracted to illustrate citation analysis and co-citation analysis. The VOSviewer software package has been widely used in published reviews of research fields such as social sciences, [64][65] business and management, [66] medicine [67] and education. [68][69] Biblioshiny package was also used to perform extraction of related Figures.

RESULTS
In this section, we represent the results of the bibliometric analysis of start-up success, which correspond with our three research questions.

Total volume, growth pattern and geographical distribution of startup success literature
In this sub-section, studies on start-up success are considered in three aspects: total volume, growth pattern and geographical locations.
As mentioned above, we obtained 1554 start-up success documents for final analysis. These documents were comprised of 1159 journal articles (74.6%), 58 books (3.7%), 121 book chapters (7.8%) and 216 conference papers (13.9%). These 1554 documents were published in 769 outlets (i.e., journals, books and conferences). Figure 2 presents documents on startup success over the years, with 1981 marking the first time that a study on start-up success was published. [70] Longitudinal analysis revealed that start-up success knowledge base might be divided into three periods, namely Incepting period, Accumulating period, and Accelerating period      The colour of the nodes represents the experiences of the respective countries in start-up success literature. The purple ones (e.g., the US, the UK, Canada, Japan, Israel or Belgium) indicate that co-authors from the respective countries had the most experiences in studies of start-up success (prior to 2012). Meanwhile, the green ones (e.g., Germany, Spain, Thailand, Malaysia, Nigeria) show that co-authors from the respective countries seem to start their attention on start-up success between 2012 and 2015. Last, it appears that co-authors from India, Brazil, Portugal, Slovakia and some other yellows are the newbie in this research topic. Interestingly, co-authors from Germany are not among the most experienced scholars in start-up success literature despite that they have become the second most productive scholar community (behind the US) on this topic at the present time. The first start-up success document authored by a scholar from Germany was in Heuer. [71] The widths of the lines connecting different nodes illustrate the co-author patterns between scholars from respective countries.
The wider a line is, the more documents are co-published by authors from the two countries. As shown in Figure 4, the two countries groups with the most co-publications on startup success (15 co-published documents) are the US-the UK and the US-Germany). They are followed by the US-Canada) and the US-Spain with 9 co-published documents.  Of the total, our database reveals that there are 3120 authors who have published at least one document on start-up success.

The most influencing outlets, authors, and documents on start-up success
Despite the large number, few authors have a substantial  influence is considered not only by the greatest number of published documents but also by the higher h-index of their published documents. As shown in Table 2, it may conclude that Gartner WB is the most important scholar of this topic, regardless of criteria, number of publications or total citations. Since the first publication on start-up success in 1990, [31] this author published seven documents on start-up success and received 1201 citations in total. Other influencing authors in this topic include Reynolds PD (6 documents,   colour. However, there are several new publications on the topic of start-up success (light colored buttons) including single-author and multi-author papers, such as (Jackson P.; Richter N.; and Schildhauer T.); (Kee Dmh.; Rahman Na.); Mets T.; Jabeen F.; and Steinert M ( Figure 5).
Next, we used the number of citations and the number of citations per year to identify the most important documents on start-up success (Table 3). Apparently, it appears some documents are authored by top scholars as found in Table 2.
For instance, Gartner WB, who is ranked number 1 in Table 2 co-authors three documents which are listed in Table 3. [31,[72][73] Other authors who are simultaneously found in Table 2 and have documents listed in Table 3 include Reynolds PD, Wright M and Frese M. These authors only have one document for each listed in Table 3. [72,[74][75] Nevertheless, it appears 15 documents listed in Table 3 not being co-authored by top scholars in Table 2. Especially, the five most cited documents [76][77][78][79][80] were all co-authored by less prolific scholars who are not ranked in

Topical trends in the SS knowledge base
To explore the most important topics in start-up success literature, we investigated keywords of start-up success documents. First, following van Eck and Waltman; [63] Zupic and Čater, [81] co-keyword analysis was conducted to identify the key themes in start-up success literature. Figure 6 represents our co-keyword analysis. Specifically, 57 keywords corresponding with 57 nodes with at least 9-time occurrences were shown in Figure 6. Similar to Figure 5, the size of each node reflects the number of occurrences of the respective keyword in our studied documents, while the widths of the lines connecting different nodes represent the number of co-occurrences of the respective keywords in the same start-up success documents. As shown in Figure 6, six overlapping clusters representing six topical themes and featured by six colours are identified from the co-keyword analysis: (1) Business in General, (2) Start-up Ecosystem, (iii) Academic Start-up, (iv) Drivers of Start-up Success, (v) Resources for start-up, and (vi) Start-up Model. Each cluster would be comprised of some respective keywords (Table 4). Apart from cluster identification, VOSViewer also supports researchers to identify the recency of topical themes. Temporal co-word analysis reveals the time period when particular topics were at the height of their popularity. [52] As shown in Figure 7, the nodes with purple colour represent the "traditional" topics (i.e., prior to 2010) in start-up success literature e.g., Entrepreneurialism, [81] high tech, [82] growth. [83] Meanwhile, the nodes with yellow colour represent the "recent" topics (i.e., 2015 onward) in start-up success literature e.g., Entrepreneurial ecosystem, [84][85] Start-up performance, [86] Open innovation; [87] and the nodes with green colour represent the topics first appeared between 2010 and 2015.
The combination of occurrence frequency and average publication year of keywords suggests interest of scholars in each period. In the early period (dark colored nodes), scholarly interest centered on" traditional and general economic issues, such as: high tech (9) (2009), small firm (9) (2002), growth (17) (2010), or entrepreneurialism (25) (2008)). Next, the issues of concern focused on issues related to start-up (innovation

DISCUSSION AND CONCLUSION
Despite having received increasing attention from researchers, there has been little known on the overall picture of start-up success literature. To address this research void, we adopted bibliometric analysis to explore the extant literature on start- Our study identified 1554 documents on start-up success worldwide from Scopus database. The first start-up success document was published in 1981. [69] Nevertheless, it appeared that during the period of 1981-1992 (named as Incepting period), start-up success seemed to not receive significant interests from the academic community as we found only 59 start-up success documents (3.80% of the total start-up success publications) published in this period. Following the Incepting period, our study identified the period of 1993-2010, during which 496 start-up documents (31.92 % of the total publications) were published. In this period, although start-up success appeared to attract more attention from scholars than the previous one, the absolute output of startup success literature was still limited, and the trajectory of a number of publications per year was still unstable. Thus, this period was named as Accumulating period. Only after 2011, the number of start-up success documents published annually has raised gradually. The year of 2019 marked the year with the highest number of start-up success documents (194 documents). Thus, the period 2011-2019 was named as accelerating period. Overall, the growth trajectory of start-up success studies since 1981 reflects the overall evolvement of the start-up in the actual business environment. It is apparently that start-up has only emerged in recent decades, especially over the past ten years as a global phenomenon. Thus, the accelerating growth of number of start-up success documents in the past ten years, as found in this study is understandable.
Regarding geographical distributions of start-up success publications, our study revealed that the developed countries of which the three most important Figures are the US, Germany and the UK contributed the majority of publications on startup success. This finding is plausible as the current statistics show that the developed world has been recognized as the hub of start-up worldwide: Europe leads in terms of constituents in the Emerging Ecosystems list with 38 start-up ecosystems.
Eight of these are from Eastern Europe. It is followed by North America with 32 start-up ecosystems and Asia-Pacific is third with 22 ecosystems. [91] Regarding the sources of publications, our study explored that start-up success documents were published in different outlets (journals, books, conferences) with different foci, including startup-related journals (e.g., Journal of Business Venturing, Journal of Business Venturing, International Entrepreneurship and Management Journal), technology-related (Technovation, Journal of Technology Transfer), management/business in general (e.g., Journal of Small Business and Enterprise Development, Harvard Business Review, Journal of Business Research ), and interdisciplinary (e.g., Regional Studies, Sustainability). While the appearance of start-up success documents on startup-related and management/business in general journals is obvious, their appearance on technologyrelated journals reflects the nature of start-up as technologybased firms rather than regular medium and small firms. [82,[92][93] At the same time, the finding that some start-up success studies were published in interdisciplinary-oriented journals reflects the nature of the current trend on interdisciplinary research in social sciences.
One of the strengths of bibliometric analysis pertains to its capacity to explore the key authors, research groups of the studied topic. In this project, we revealed top scholars and their research groups as the main hubs of start-up success knowledge. As represented above, our study revealed that despite receiving increasing interests from the academic community, start-up success's knowledge base is not comprised of truly productive and active scholars and research groups. Specifically, the most productive scholar only published seven start-up success documents, whereas most high-profile research groups gather around three or four authors. These Figures are much lower than the respective ones of research topics which were initiated in the same period (i.e., the early 1980s). [94][95][96] The sample shows that 3120 scholars from 79 countries around the globe have published 1554 documents on the topic of startup success. The topic of startup success has attracted many scholars to research, affirming worldwide interest in SS but also reveals a noticeable geographical imbalance in this knowledge base. There is a lack of continuity in research. Several prominent research groups in the past have not had recent publications. It is uncommon for new scholars to join established research groups, according to the Scopus data.
Last, our study adopted co-word analysis to identify the key themes of start-up success literature. These include (i) business in general, (ii) start-up ecosystem, (iii) academic start-up, (iv) drivers of start-up success, (v) resources for start-up, and (vi) start-up model from longitudinal dimension. The key themes (ii) start-up ecosystem, (v) resources for start-up, and (vi) start-up model have successively published studies in the accelerating period, while the remaining topics have no published studies recently. This study also identified the most recent "hot topics" of start-up success, such as Entrepreneurial ecosystem, Entrepreneurial finance, Start-up performance, Sustainability, Entrepreneurship education Accelerator; along with the more traditional ones Entrepreneurialism, High tech, Technology transfer, University, Internet, Nascent entrepreneur. These findings would be necessary references for future scholars who want to undertake empirical research on start-up success.
This study may provide implications for stakeholders. First, as start-up success has received increasing attention from academic scholars in recent years and is expected to receive more attention in upcoming years; thus, scholars may use this bibliometric analysis as key reference for their further studies on start-up success. Specifically, thanks to this study, scholars on start-up success have already known the key hubs of startup success studies as well as the collaborating pattern among these key hubs. Besides, this study also identifies for future scholars on start-up success the key outlets for referencing and submitting their works on start-up success. Furthermore, future scholars who want to undertake studies on start-up success may consult from the finding of six topical trends on start-up success as unveiled in this study. Second, not only academic scholars, policymakers, start-up owners, entrepreneurs and practitioners may also use this study as knowledge base for their policy-making and decision-making activities. Last, this study may also be used as material in business course at higher education levels, including undergraduate and graduate education.

Limitations and Suggestion for Further Research
This study has several limitations, as many others do. [97] First, despite the advantages of bibliometric analysis, we should be aware that it only works with metadata information and does not cover the content of start-up success studies. Future attempt to review the current status of start-up success literature may employ different approach such as content analysis [98] to get insights into this knowledge base. Second, this study encompasses all studies, irrespective of their sources of studies. Thus, it may provide a worldwide picture of start-up success literature but lack of specific niche which corresponds with different regions or countries. Future scholars who would like to shed further light on start-up success literature may follow the approach of Phillip Hallinger and his colleagues, who conducted a series of bibliometric analysis on educational leadership and management with different contexts such as Europe [99] Asia; Africa and Latin America; [100] Arab societies. [69]