This study aims to analyse the literature on the digital transformation of family businesses and the impact of artificial intelligence on this process, highlighting key areas of interest and future perspectives.
A bibliometric analysis is performed to explore the interconnection between variables and the relationships between authors, countries and journals in this research area. The Scopus database was used as of March 2024, and the data analysis was carried out with Bibliometrix for result analysis and VOSviewer for scientific mapping.
The analysis confirms the increasing relevance of the topic, with a high number of articles in 2023. Prominent journals are identified, and authors are mainly from China and Europe. Keywords “family business” and “family firms” are strongly linked, showing a connection to artificial intelligence and digital transformation. Family businesses are embracing the digital era, and research must respond accordingly.
This pioneering study offers a novel contribution, as no prior bibliometric analysis has addressed this topic. It lays the groundwork for future research, identifying emerging themes with significant future potential.
1. Introduction
Family businesses, as a specific subset of companies, represent the most widespread form of business organization worldwide (Hernández-Perlines et al., 2021), so it is not surprising that research on them has proliferated exponentially in recent decades (Chrisman et al., 2024). However, Upadhyay et al. (2023) note that it is necessary to delve deeper into the specific challenges of integrating disruptive technological innovations, such as artificial intelligence (AI), into family businesses, as previous research on the subject is scarce and fragmented. Furthermore, according to a survey conducted by Deloitte (2019) among 791 executives of family businesses, the adoption of such technologies in the workplace is perceived as the most significant factor in remaining competitive in the market.
Today, the business landscape is undeniably being transformed by the increasing integration of AI into various processes and sectors (Atienza-Barba et al., 2024). In the specific case of family businesses, AI presents unique challenges and opportunities for enhancing their competitive advantages, which are related to the origin of social capital (Soluk, 2022), intergenerational succession issues (Wang and Li, 2023), and the peculiarities of governance and decision-making (Ulrich et al., 2023; Upadhyay et al., 2023). However, it is precisely the defining characteristics of family businesses—specifically the importance of family structure and how its interactions and values permeate the corporate culture (Upadhyay et al., 2023)—that are shaping their strategy, leading them to view AI integration as a low priority (Ulrich et al., 2023), with some even being sceptical of its potential (Lannon et al., 2024).
The desire to preserve the family legacy for future generations often drives this more conservative stance (Radu-Lefebvre et al., 2024). However, as Szymanska et al. (2019) argue, the intention to safeguard the past can sometimes conflict with efforts to secure the future through innovation. In such cases, Kidwell et al. (2018) suggest that family businesses may end up passing down not values but rather bad habits across generations. On the other hand, when the founder serves as a role model, the family business adopts an entrepreneurial orientation, which necessitates a mindset open to change (Cruz and Nordqvist, 2010). Additionally, fostering the ability to think originally and promote innovation can be particularly valuable for the continuity of the next generation (Astrachan et al., 2020). In this context, AI, as a disruptive technological innovation, emerges as one of the primary tools for achieving competitive advantages and successfully leveraging them against other companies (Wang and Li, 2023). At the same time, the use of AI raises questions about social responsibility and ethics (Dwivedi et al., 2021), which may have particular significance for family businesses, as AI could even be seen as a threat to the family’s position within the business structure (Ulrich et al., 2023).
The development of AI is so recent that we lack a systematic understanding of its integration into family businesses (Lannon et al., 2024). Consequently, understanding its influence on both the present and future poses a challenge for advancing both theory and practice. The novelty and fragmentation of AI research (Atienza-Barba et al., 2024) exacerbate these blind spots. In fact, this fragmentation arises from the multidisciplinary approach to AI research, which has led to the parallel development of three distinct bodies of literature, each examining AI from different perspectives: technical, social, and economic. These separate streams of literature have not yet been integrated into a unified perspective on AI, despite previous attempts to synthesize current research (Dwivedi et al., 2021; He and Liu, 2024). Although some works implicitly address “what AI is” and “what it is used for” in family businesses, their narrative reviews do not identify “how to integrate it” nor do they examine the various AI contexts at the national or business-type level.
This article maps and integrates previous research on AI adoption within the context of family businesses. Our review contributes to the literature on family businesses by offering a comprehensive overview of current knowledge on AI and identifying areas where further research is needed. Given that AI is a source of competitive advantage, this integrated understanding has both theoretical and practical implications, especially regarding the survival and growth of family businesses over time (Lannon et al., 2024). These are matters of significant economic, social, and political importance, given that family businesses contribute between 70% and 90% of global employment and Gross Domestic Product (GDP), according to Deloitte estimates (2019).
Additionally, other contributions arise from addressing the main research questions, which are as follows:
What are the most relevant studies on AI in the context of family businesses?
In which countries and institutions do the most influential researchers in the field of AI in family businesses work?
In what research networks do the main authors on AI in family businesses participate?
Which scientific journals produce the most knowledge about AI in family businesses?
What relevant research topics stand out in the field of AI in the context of family businesses?
To answer these questions, the next section describes the process undertaken to identify, select, and code articles on AI, before presenting the main results of the study. Based on the findings, we suggest theoretical perspectives and research gaps to guide future research on AI in family businesses.
2. Methodology
To ensure a rigorous and transparent approach in identifying and selecting relevant studies, we applied the PRISMA protocol (Preferred Reporting Items for Systematic Reviews and Meta-Analyses). PRISMA is a widely recognized standard in academic research that systematizes and documents each stage of the review process. This ensures that the bibliometric analysis meets the criteria of comprehensiveness and replicability, thereby contributing to the robustness and validity of the findings presented (Page et al., 2021). Figure 1 illustrates the phases followed.
2.1 Phase 1: Identification and exclusion criteria
For the systematic analysis, data were retrieved from the Scopus database, one of the leading global academic information sources that provide comprehensive access to peer-reviewed literature, abstracts, and citations, ensuring the thoroughness of the search and representativeness of the selected studies (de Moya-Anegón et al., 2007). A search was conducted using the terms (TITLE-ABS-KEY (“artificial intelligence” “family business”) OR TITLE-ABS-KEY (“artificial intelligence” “family firms”) OR TITLE-ABS-KEY (“digital” “family business”) OR TITLE-ABS-KEY (“digital” “family firms”)), which yielded 161 records. The inclusion and exclusion criteria were as follows: (1) no time limit was applied; (2) only articles and conference papers were selected to ensure the quality of the research; (3) no restrictions were imposed on fields of study, due to the multidisciplinary nature of the subject; and (4) for language, only English was included for compatibility with the subsequent analysis. As a result, of the 161 records initially found, 119 met all these criteria and proceeded to the next phase of the analysis.
2.2 Phase 2: Data pre-processing
In this stage, the 119 documents were manually reviewed to eliminate false positives, as the data volume made this feasible. This false positive analysis identified 10 records that were not related to the subject matter. Therefore, after pre-processing, 109 valid records were identified. These were analysed using Bibliometrix and VOS viewer, as proposed by Donthu et al. (2021). Two different software tools were used to analyse the data in a comprehensive and multi-faceted manner, minimizing potential bias from relying on a single software.
2.3 Phase 3: Bibliometric analysis
Following He and Liu (2024), the bibliometric analysis was carried out using two approaches. The first approach was a performance analysis, where the scientific production and its impact on the subject were examined. This included analysing the number of documents and their annual citations, relevant authors and sources, and the most cited articles. All of these analyses were performed using Bibliometrix (version R-3.6.1). The second approach was a science mapping analysis, which graphically mapped the evolution of this subject. This involved conducting a co-occurrence network of author keywords, a co-citation analysis of sources and authors, and a co-authorship analysis of authors and countries. All of these analyses were carried out using VOSviewer (version 1.6.20), except for the keyword co-occurrence analysis, which was done using Bibliometrix. Additionally, an analysis was included on the most recent family business articles that feature the keyword “artificial intelligence.”
3. Results
3.1 Annual scientific production
In Scopus, the first article recorded on this topic dates back to 1960, titled “Bibliography on Simulation, Gaming, Artificial Intelligence and Allied Topics” (Shubik, 1960). In this century, artificial intelligence has been expanding across all fields, both scientific and economic. In the economic sphere, the number of publications on family businesses and artificial intelligence has increased exponentially, although they were scarce until 2017, corroborating the results of previous research (Atienza-Barba et al., 2024).
Figure 2 shows how, since 2017, the number of articles related to artificial intelligence and family businesses has been increasing, following a linear trend of y = 5.7857x − 9.8571, R2 = 0.7574, represented by the blue line (the coefficient of determination, R2, indicates how much of the variability in the data is explained by the linear model. With a value of 0.7574, it shows that 75.74% of the variability in the number of articles is explained by this linear trend, indicating steady and predictable growth over time). The exponential equation y = 1.1565eˆ0.4918x, R2 = 0.9669, represented by the green line, has a higher R2 of 0.9669, meaning that 96.69% of the variability in the data is explained by this exponential trend, indicating a much stronger fit than the linear model. The exponential equation suggests that the number of articles is not just increasing but doing so at an accelerating rate. This implies that, as time passes, the growth in the number of articles accelerates, rather than remaining constant. In fact, 2023 saw 47 publications, doubling the number from 2022. If the trend continues, we can expect even more rapid growth in the number of publications in this field in the coming years.
The number of citations is variable but is expected to follow an upward trend due to the increasing number of articles. While 2021 had the highest number of citations, with 479, an exponential trend line was constructed: y = 16.968eˆ0.4187x, R2 = 0.1578, represented by the purple line (the R2 of 0.1578 indicates that only 15.78% of the variability in the number of citations is explained by the exponential model, possibly due to non-linear patterns in the data). A linear trend line was also constructed: y = 38x + 6.5714, R2 = 0.2792, represented by the orange line (the linear equation suggests that the number of citations increases steadily over time at a rate of 38 citations per year, plus an initial constant of 6.5714 citations. This linear trend indicates a steady growth in the number of citations, although the R2 suggests that only 27.92% of the variability in the number of citations is explained by the linear model). Moreover, by March 2024, ten articles had already been published, garnering a total of six citations, reflecting the current interest in the topic.
3.2 Most relevant authors
Following Álvarez-García et al. (2023), the top five most relevant authors on artificial intelligence in family businesses have been identified based on the number of articles published (Table 1). Leading the list is J. Soluk, with five articles. While he has accumulated a total of 406 citations, only 328 (80.79%) pertain to the topic under analysis. Notably, he is the only author in the top five not affiliated with a European institution. Additionally, most of his work in this area is co-authored with N.H. Kammerlander in four documents and with A. De Massis in 2 documents, both of whom also rank in the top five.
Author analysis: The five most frequent authors in the area
| Ranking | Author | Documents (Year) | Affiliation | h-Index | Total citations | TFCI | ||
|---|---|---|---|---|---|---|---|---|
| FF | DT | AI | ||||||
| 1 | Soluk, J. | De Groote et al. (2023), Soluk (2022), Soluk et al. (2021 a, b), Soluk and Kammerlander (2021) | Stanford Engineering, Stanford, United States | 7 | 406 | 1.32 | – | – |
| 2 | Kammerlander, N.H. | De Groote et al. (2023), Soluk et al. (2021 a, b), Soluk and Kammerlander (2021) | WHU – Otto Beisheim School of Management, Vallendar, Germany | 28 | 3.318 | 1.32 | – | – |
| 3 | Bargoni, A. | Bargoni et al. (2023a, b, c) | Università degli Studi di Torino, Turin, Italy | 6 | 72 | – | – | – |
| 4 | Vrontis, D. | Bargoni et al. (2023a, b, c), Chatterjee et al. (2023), Chaudhuri et al. (2023) | University of Nicosia, Nicosia, Cyprus | 53 | 9.848 | 1.32 | 1.69 | 1.33 |
| 5 | De Massis, A. | Soluk et al. (2021a, b) | Free University of Bozen-Bolzano, Bolzano, Italy | 62 | 11.973 | 1.32 | 1.69 | – |
Note(s): Topic Field-Weighted Citation Impact (TCI); Socioemotional Wealth, Family Firms, Innovation (FF); Digital transformation, Strategic alignment; COBIT, Business Model Innovation, Innovation, Enterprise Architecture (DT); Human Resource Information Systems, E-HRM, Artificial Intelligence, Interpretive research, Hermeneutics (AI)
Source(s): Authors’ own creation
In second place is A. De Massis, who has the highest h-index (62) and citation count (11,973) among the top five authors. The h-index measures a researcher’s productivity (number of articles) and impact (number of citations), with an h-index of 62 indicating that De Massis has at least 62 publications in Scopus, each of which has been cited at least 62 times. Additionally, De Massis has a Topic Field-Weighted Citation Impact (TFCI) of 1.32 in Socioemotional Wealth, Family Firms, Innovation (FF), and 1.69 in Business Model Innovation, Innovation, Digital Transformation (DT). TFCI is a metric used to evaluate the impact of an article’s citations relative to the average in its specific field. A TFCI greater than 1, as seen here with values of 1.32 and 1.69, means that De Massis’ articles have been cited 32 and 69% more than the field average, indicating a significantly higher influence.
In third place is A. Bargoni, who has the lowest h-index (6) and total citation count (72) among the top five. In 2023, Bargoni collaborated with D. Vrontis, who holds the second-highest h-index (53) and boasts a TFCI greater than 1 in the topics of FF (1.32) and DT (1.69). Moreover, Vrontis is the only author with a TFCI greater than 1 (1.33) in the topic of Human Resource Information Systems, E-HRM, Artificial Intelligence, Interpretive Research, and Hermeneutics (AI). In summary, Vrontis is more frequently cited than the average in the fields of family firms, digital transformation, and artificial intelligence.
3.3 Most relevant journals
Following Pérez-Romero et al. (2023), the journals that have published the most articles on this topic are listed in Table 2. The Journal of Family Business Management (Q1 in Scopus) ranks first with 15 publications. Notably, one of the articles from this journal, by Rashid and Ratten (2020), has received 24 citations. Four out of the five journals are in the first quartile (Q1), with the exception of Case Journal, which is Q4.
Source analysis: The five most frequent journals in the area
| Ranking | Journal | Highest percentile (HP) | CiteScore (2022) | CiteScoreTracker 2023 | Documents | Topics |
|---|---|---|---|---|---|---|
| 1 | Journal of Family Business Management | 86% | 4.6 | 5.5 | 15 | Economics, Econometrics and Finance Business, Management and Accounting |
| 2 | International Journal of Entrepreneurial Behaviour and Research | 94% | 9.9 | 10.2 | 4 | Business, Management and Accounting |
| 3 | Journal of Business Research | 96% | 16 | 20.2 | 4 | Business, Management and Accounting |
| 4 | Sustainability | 87% | 5.8 | 6.8 | 3 | Social Sciences Computer Science Environmental Science; Energy |
| 5 | Case Journal | 9% | 0.2 | 0.2 | 3 | Business, Management and Accounting Social Sciences |
Source(s): Authors’ own creation
However, the journal with the highest CiteScore is the Journal of Business Research, with a score of 16, and projections suggest it will reach 20.2 by 2023 (as of April 5, 2024). CiteScore is a metric created by Scopus to evaluate the quality and impact of academic journals based on the number of citations their articles receive. A high CiteScore generally indicates higher quality, as is the case with the journals where articles on this subject are published.
3.4 Most globally cited documents
According to Kumar et al. (2024), the most cited document in this category (Table 3) is Soluk and Kammerlander (2021), with 135 citations and an impact factor of 14.07 (>1), indicating it is a key reference in this field, averaging 33.75 citations per year. All the articles in this ranking are from 2021, except for one from 2019, which has an impact score of 9.27 with 116 citations. It is important to note that this impact score is calculated based on the number of citations, publication year, document type, and the disciplines associated with the journal.
Document analysis: The five most globally cited documents
| Ranking | Title | Author (year) | Source | Total citation | Citation per year | Impact citation |
|---|---|---|---|---|---|---|
| 1 | Digital transformation in family-owned Mittelstand firms: A dynamic capabilities perspective | Soluk and Kammerlander (2021) | European Journal of Information Systems | 135 | 33.75 | 14.07 |
| 2 | Investors’ choices between cash and voting rights: Evidence from dual-class equity crowdfunding | Cumming et al. (2019) | Research Policy | 116 | 19.33 | 9.27 |
| 3 | Family Influence and Digital Business Model Innovation: The Enabling Role of Dynamic Capabilities | Soluk et al. (2021b) | Entrepreneurship: Theory and Practice | 112 | 28 | 10.42 |
| 4 | A dynamic panel study on digitalization and firm’s agility: What drives agility in advanced economies 2009–2018 | Škare and Soriano (2021) | Technological Forecasting and Social Change | 88 | 22 | 7.63 |
| 5 | Exogenous shocks and the adaptive capacity of family firms: exploring behavioural changes and digital technologies in the COVID-19 pandemic | Soluk et al. (2021a) | R and D Management | 59 | 14.75 | 26.91 |
Source(s): Authors’ own creation
Notably, the article by Soluk et al. (2021a) stands out for having the highest citation impact (26.91). Among the top five most cited articles, three are authored by J. Soluk, confirming his prolific contribution to the field (Table 1). In terms of sources, these articles have been published in different journals, but all belong to the Q1 category in the WoS Core Collection. This highlights both the importance and quality of research on this topic.
3.5 Co-occurrence network of author keywords
According to Jiménez-Islas et al. (2024), the co-occurrence analysis of keywords (Figure 3) reveals seven clusters of related terms. Cluster I shows a strong connection between the keywords family business and family firms, which are synonymous, along with digital storage and decision making. This cluster is linked to Clusters II and III, relating to digital transformation and artificial intelligence. Cluster II connects the keywords artificial intelligence and absorptive capacity, both of which are related to family business, even though they belong to different clusters. Cluster III focuses on digital transformation, which links both family business and family firms and is further connected to Cluster V because of its association with the digital domain.
Cluster IV links rural area with innovation, though this connection is isolated from the other thematic areas along with Clusters VI and VII. Cluster V associates digital technologies with competition, reflecting how constantly evolving digital technologies must remain competitive. Cluster VI connects sustainable development with economic and social effects, underscoring the relationship between sustainable objectives and socio-economic factors. Finally, Cluster VII links family structure with firm ownership and industrial enterprise, as the structure of a family firm parallels that of any other type of enterprise.
3.6 Authors Co-citation analysis
The authors’ co-citation analysis (Figure 4) was conducted following Álvarez-García et al. (2023) using VOSviewer, with a minimum of 20 citations. This analysis identifies how authors in the field have cited each other. The analysis reveals 60 authors divided into four distinct clusters.
- (1)
Cluster I represents authors focused on digital transformation in family firms. This cluster includes 20 authors with a total of 1,208 citations. The most cited authors are A. de Massis (262 citations), N. H. Kammerlander (125 citations), J. Soluk (54 citations), I. Miroshnychenko (24 citations), and I. Naldi (23 citations).
- (2)
Cluster II centres on socioemotional wealth in family firms under various circumstances, encompassing 15 authors with a total of 500 citations. Key contributors include D. Miller (76 citations), L. R. Gomez-Mejia (67 citations), and I. Le Breton-Miller (43 citations).
- (3)
Cluster III focuses on the study of behaviour towards the adoption of technology and artificial intelligence in family businesses during times of crisis. This cluster comprises 15 authors with a total of 519 citations. Prominent figures in this cluster are S. Kraus (67 citations), S. Chatterjee (39 citations), D. Vrontis (38 citations), and Y. K. Dwivedi (23 citations).
- (4)
Cluster IV addresses family business and its effective organization, representing 10 authors with a total of 306 citations. Notable authors in this cluster include M. Nordqvist (61 citations), F. Chirico (48 citations), and G. Campopiano (25 citations). (25).
3.7 Co-citation sources analysis
This analysis examines the sources where documents related to the topic are published, based on co-citation. VOSviewer was used to create this connection map, which features 25 nodes grouped into three clusters, with each node having at least 30 citations and presenting 299 links between them (Figure 5).
- (1)
Cluster I: “Business, International Management and Accounting” includes 12 sources, each with 24 links and a total of 1,096 citations. Notable high-impact journals in this cluster include Family Business Review (96th percentile, 255 citations), Entrepreneurship Theory and Practice (99th percentile, 215 citations), and Strategic Management Journal (93rd percentile, 102 citations). These journals have a strong interconnection (link strength of 1,528). Furthermore, Journal of Family Business Strategy (Cluster III) is more strongly connected to Family Business Review (link strength of 907) than to Entrepreneurship Theory and Practice (link strength of 672).
- (2)
Cluster II: “Management of Technology, Management Information Systems and Innovation” encompasses seven sources, each with 24 links and a total of 374 citations. Prominent journals in this cluster include Journal of Small Business Management (96th percentile, 92 citations), Journal of Business Venturing (99th percentile, 49 citations), and MIS Quarterly (98th percentile, 45 citations).
- (3)
Cluster III: “Strategy, Marketing and Management” represents six sources, each with 24 links except for Sustainability (23 links), totalling 542 citations. Journals in this cluster, including Journal of Business Research (96th percentile, 151 citations), Journal of Family Business Strategy (93rd percentile, 147 citations), and Journal of Family Business Management (86th percentile, 106 citations), exhibit high impact as indicated by their percentile rankings and citation numbers.
3.8 Co-authorship author analysis
The co-authorship analysis identifies collaborations between authors based on their joint publications. Figure 6 illustrates nine authors grouped into clusters according to their collaborations, with a minimum of two documents per author related to the topic as the inclusion criterion.
- (1)
Cluster I (Yellow): This cluster includes four authors who have published articles on the topic in 2023. Key authors are D. Vrontis and A. Bargoni, who each have five articles (Bargoni et al., 2023a, b, c; Chatterjee et al., 2023; Chaudhuri et al., 2023). The most cited article is by Chaudhuri et al. (2023), with 23 citations. Authors in this cluster are from different countries, indicating international and recent collaborations: D. Vrontis (Cyprus), A. Bargoni (Italy), S. Chatterjee (India), and R. Chaudhuri (France).
- (2)
Cluster II (Green): This cluster contains two authors who have published articles in 2021 and 2023 (González-López et al., 2021; Zapata-Cantu et al., 2023). Both authors are affiliated with the University of Extremadura (Spain), reflecting a collaboration within the same institution.
- (3)
Cluster III (Blue): This cluster features three authors with a total of five articles, although one document is a solo work by Soluk (2022). Articles from 2021 include Soluk et al. (2021a, b), and Soluk and Kammerlander (2021), with an additional publication in 2023 (De Groote et al., 2023). This cluster also represents international collaboration: J. Soluk (United States), N.H. Kammerlander (Germany), and A. de Massis (Italy).
Additionally, there are authors with occasional collaborations in two or more articles on the topic:
- (1)
N. Lattanzi (Italy) with Amato et al. (2022) and Patuelli et al. (2022).
- (2)
V. Pereira (France) with Jayawardena et al. (2023) and Pereira et al. (2024).
- (3)
S. Sawang (United Kingdom) with Sawang and Kivits (2023) and Tiwasing et al. (2023). Notably, Sawang’s article, “Revolutionizing family businesses with artificial intelligence: a perspective article,” discusses the promising future of using Theory of Mind AI to aid decision-making in family businesses (Sawang and Kivits, 2023).
3.9 Co-authorship country analysis
Based on Safta et al. (2021), a chronological co-authorship analysis (Figure 7) visualizes the countries contributing to the field, with a minimum of five documents and 30 citations per author.
Pioneers and Recent Contributors: The earliest articles from 2019 come from the United States, considered a pioneer with six articles and 170 citations. Spain followed in 2020 with 12 articles and 161 citations. In 2021 and 2022, the United Kingdom, Italy, and Germany contributed significantly with a total of 1,292 citations across 45 documents. In 2023, France and India contributed with 12 documents and 111 citations.
Country Collaborations: Notable collaborations include:
- (1)
United Kingdom (17 collaborations) leads in co-authorship, followed by Italy (10), Germany (8), India (8), United States (7), China (6), and France (6).
- (2)
Spain has only 2 collaborative articles with China out of the 12 published on this topic.
- (3)
United Kingdom and Italy have collaborated with all countries depicted in Figure 7 except Spain, indicating sustained collaborations.
- (4)
More recent collaborations (between India and France) are limited to interactions with each other and with Anglo-Saxon countries and Italy.
The extensive and diverse country collaborations indicate that the topic is of global interest, demonstrating its relevance both in Europe and in developing countries like India.
3.10 Recent articles with AI as an author keyword
Table 4 presents the five most recent articles on family businesses that include “artificial intelligence” as an author keyword. These studies are mostly theoretical, except for Upadhyay et al. (2023), which propose a structural equation modelling approach. Notably, three of the articles are from 2023 and two are from 2024. The article with the highest citations is Upadhyay et al. (2023), with 26 citations. The authors of these articles are affiliated with diverse institutions across countries such as the United Kingdom, Ireland, United Arab Emirates, Jordan, China, and India.
Keyword artificial intelligence analysis: The five most recent articles with “artificial intelligence” as an author keyword
| Author (Year) | Articles | Author keywords | Main contributions |
|---|---|---|---|
| Rahman and Zhu (2024) | Detecting accounting fraud in family firms: Evidence from machine learning approaches | Accounting fraud detection; Artificial intelligence; Family firms; Imbalanced ensemble learning; Machine learning | Utilizes machine learning algorithms for the first time to predict fraud in family firms |
| Lannon et al. (2024) | Generation AI and family business: a perspective article | Artificial intelligence; Education; Successors | Determines that to capitalize on AI, it is essential to study the role of successors in family businesses |
| Upadhyay et al. (2023) | The influence of digital entrepreneurship and entrepreneurial orientation on intention of family businesses to adopt artificial intelligence: examining the mediating role of business innovativeness | Artificial intelligence; Business innovativeness; Digital entrepreneurship; Entrepreneurial orientation; Family business; Technology adoption | Proposes a model to determine the factors influencing the adoption of artificial intelligence by family firms |
| Gilani et al. (2023) | Savior or distraction for survival: Examining the applicability of machine learning for rural family farms in the United Arab Emirates | Artificial intelligence; Family businesses; Farms; Innovation; Machine learning; Rural; UAE | Studies factors affecting the adoption of machine learning in family-owned agricultural enterprises |
| Sawang and Kivits (2023) | Revolutionizing family businesses with artificial intelligence: a perspective article | Artificial intelligence; Digital transformation; Emotional AI; Family business; Innovation adoption | Proposes a future line of research on using Theory of Mind AI to assist family businesses in decision-making and reducing conflicts |
Source(s): Authors’ own creation
As common factors in these articles, we identified a lack of a theoretical framework for the research (Lannon et al., 2024; Gilani et al., 2023). Thus, currently, there are still very few quantitative empirical contributions (Upadhyay et al., 2023), with most articles being theoretical (Lannon et al., 2024) or qualitative (Gilani et al., 2023), which seems to align with a novel research topic. The scarce research agrees that, despite being a source of competitive advantage (Lannon et al., 2024; Rahman and Zhu, 2024), the integration of AI poses a challenge for family businesses due to their corporate culture and governance compared to non-family businesses (Upadhyay et al., 2023; Gilani et al., 2023).
4. Conclusions
This study provides a detailed analysis of previous research on the use of AI in family businesses, based on SCOPUS data. Given the novelty of the topic and the strategic relevance of AI for SMEs and family firms (Buer et al., 2021; Ulrich et al., 2023), authors like Chaudhuri et al. (2023) and Chrisman et al. (2024) call for further research to enrich the debate and structure the existing knowledge. In response, the primary goal of this work has been to contribute to creating a specific body of knowledge for family businesses that integrates different explanatory and justificatory approaches for a better understanding and application of AI in these businesses through bibliometric and network analyses.
The results show an exponential growth trend in the number of published documents, indicating a marked increase in the popularity and interest in the topic. However, citations are concentrated in 2021, possibly due to the impact of the health crisis and the relevance of technologies related to digital transformation across various fields, including family businesses, which temporarily focused the attention of researchers from various knowledge domains. In this regard, the seminal study on AI in the context of family businesses (RQ1) is by Soluk and Kammerlander (2021) (Table 3). These authors analyse how small and medium-sized family businesses with limited resources face the digital transformation process, identifying a veto power of patriarchs based on paternalistic leadership, which hinders, even more than the scarcity of resources, the adoption of disruptive technologies such as AI. In the last year, the most relevant article is by Upadhyay et al. (2023), which propose a model to determine the factors influencing the adoption of AI by family businesses, with a notable emphasis on entrepreneurial innovation capacity in the intention to use AI in family firms.
Furthermore, J. Soluk is the most prolific author in the topic (RQ2), although not the one with the most international collaborations or the highest h-index. In this regard, A. De Massis has the highest h-index (62), with D. Vrontis (53) also standing out. Regarding countries with the highest level of international collaboration (RQ3), although several European countries lead the ranking (United Kingdom, Italy, and Germany), India and China also stand out, corroborating that this is a globally relevant topic, with an emerging research network among countries from both continents, including the United States.
In relation to the sources generating the most knowledge about AI in the context of family businesses (RQ4), Family Business Review stands out for the number of citations (255), although it does not have the highest number of documents published on the topic. The Journal of Family Business Management leads the ranking with 15 documents, although the publication of this special issue may contribute to an increase in citations. Lastly, based on the findings, key research topics in the field of AI in the context of family businesses (RQ5) have been identified, which are grouped into four clusters: the use of AI as a tool in the digital transformation process of family businesses, socioemotional wealth in family businesses and the effects of technologies and AI, behaviour towards AI adoption in times of crisis, and the effects of AI on the effective organization of family businesses.
4.1 Theoretical and practical recommendations
Despite the growing importance of scientific production on family businesses, it is observed that literature has not yet reached a consensus on the economic theory best suited to the specific characteristics of family businesses. Various theories, such as systems theory, agency theory, management theory, resource-based view, stakeholder theory, or socioemotional wealth perspective, have been used to explain the behaviour of these businesses (Chrisman et al., 2024), revealing that the research corpus on family businesses is dispersed. Although there have been some attempts to systematize it, such as those by Araya-Castillo et al. (2022) or Baltazar et al. (2023), further research is needed to address the challenge of identifying which theoretical paradigm predominantly supports these businesses. This gap persists despite attempts by Sawang and Kivits (2023) to propose Theory of Mind AI to assist family businesses in decision-making through AI. It is important to remember that family businesses involve two subsystems (family and business), so it is advisable to ensure a balance between the objectives of both to guarantee business survival. However, these subsystems can involve individuals with different cultures and values. Given that these are principles guiding people’s actions (Schwartz, 1994); we propose a paradigm shift from balance-based approaches to a more holistic integration of ethical principles, encompassing not only the technical aspects of AI systems but also the sociotechnical contexts in which these systems operate. This would ensure that AI technologies not only advance in sophistication but do so with a commitment to social welfare and ethical integrity, following the approach of Manjarrés et al. (2021) and the Sustainable Development Goals (Astobiza et al., 2021).
Practically, recommendations focus on the fact that family businesses, like any other organization, are required to innovate, continuously improving their models and processes to survive in a competitive and global market. However, these businesses lag in the use and integration of AI (Ulrich et al., 2023) due to various reasons (Soluk et al., 2021a, b), including the shortage of skilled labour or resistance to change from senior family members, who may view AI as a threat to the family’s position in the business structure. As noted by Stommel et al. (2024), it is important to consider that decision-making processes in family businesses must balance business growth with the risk of losing family control. In this sense, public policies should address these challenges, for example, by designing effective strategies to promote an internal attribution mindset, in terms of Attribution Theory (Schwartz, 1994), or through training programs to improve skills and perception of AI as a source of competitive advantage. Moreover, family business managers should seek opportunities to gradually increase their organization’s AI competence, both through research projects and participation in knowledge exchange networks. In any case, we suggest that AI integration in family businesses be gradual, starting with, for example, integrating AI into CRM systems, as suggested by Chaudhuri et al. (2023), to gradually familiarize with AI’s capabilities.
4.2 Limitations and future research
Regarding the limitations of this work, it is important to acknowledge that it was based solely on SCOPUS. Although it is a database with extensive coverage and recognized by academia, future research should replicate the study in other databases, such as WoS or PubMED, including publications in other languages, to offer a more comprehensive view of this research field, consistent with what was proposed by Atienza-Barba et al. (2024). Given the novelty of the topic and the expected growth rate of publications, it would be advisable to repeat this study periodically to analyse the future evolution of the intellectual structure and emerging trends in this research area.
Secondly, to calculate the impact of citations weighted by topics (TFCI), three thematic groups have been identified: the first includes Socioemotional Wealth, Family Business, and Innovation; the second includes Digital Transformation, Strategic Alignment, COBIT, Business Model Innovation, Innovation, and Enterprise Architecture; and the third includes Human Resource Information Systems, Electronic Human Resource Management, Artificial Intelligence, Interpretive Research, and Hermeneutics. These thematic groups make it difficult to accurately assess the real impact of the main themes, AI and Family Business, so future research should disaggregate the proportion of impact attributed to each theme, identifying the primary and secondary themes.
Thirdly, this study focused on the generic term AI, but it encompasses a range of technologies (Berente et al., 2021), with chatbots being particularly popular (Zhang et al., 2023), and it is a term in constant evolution, which may have led to the exclusion of some relevant articles. Future research could include terms related to chatbots or other AI-based technologies, such as virtual assistants, to capture any relevant articles that may have been excluded in this study.
Finally, it should be noted that this is a theoretical and descriptive study aimed at identifying the intellectual structure of AI in the context of family businesses as an emerging and novel topic, facilitating the detection of potential future research gaps. In this regard, based on the findings presented here, future research should delve deeper into the process of integrating AI into these types of businesses and the effects that variables such as business size (Buer et al., 2021) or socioemotional wealth (Hernández-Perlines et al., 2021) have on the success of this process. We propose a research agenda for adopting digital transformation in family businesses by means of (1) developing a theoretical framework and (2) increasing the empirical contributions focusing particularly in the factors inherent to family businesses hindering the integration of AI.
Funding: This work was supported by University of Castilla-La Mancha (UCLM), Spain, and the European Regional Development Fund (ERDF) under Grant 2022-GRIN-34373. This publication has been made possible thanks to funding granted by the Consejería de Economía, Ciencia y Agenda Digital de la Junta de Extremadura and by the European Regional Development Fund of the European Union through the reference grant GR21161.
