Current dynamic and volatile environments require organizations to develop functional dynamic capabilities to be competitive (Barney, 1991; Penrose, 1959; Wernerfelt, 1984). In this context, the use of knowledge, information, and technologies becomes relevant as means to reduce uncertainty and successfully overcome the demands of the environment (Chuan et al., 2017; Gray, 2006; Hassan & Raziq, 2019; Nonaka & Takeuchi, 1995; Väyrynen et al., 2017). According to Mundy (2016) , management cannot always establish the optimal combination of tacit and explicit knowledge for value generation during decision-making. In this scenario, KM arises as an answer to the question: What is the most efficient way to manage knowledge in turbulent environments?

On the other hand, it is more feasible for large companies to design robust KM systems than smaller ones (Hutchinson & Quintas, 2008; Rodríguez-Ponce et al., 2010; Xie et al., 2020). The structure of SMEs is usually very flexible because resource scarcity forces their owners and other workers to perform different functions to solve everyday requirements. In this sense, Hall (1979) , Pratten (1991) , and Velandia et al. (2016) state that in SMEs, discretion exists to act differently from programmed; workers become versatile depending on circumstances. This situation generates an environment rich in new solutions giving place to new uses and configurations of organizational resources and capabilities. However, the same act of living for everyday solutions makes that organizational learning processes take place in a spontaneous, natural, more informal, and less planned way (Xie et al., 2020), leaving aside formal KM processes, which makes their study difficult despite the relevance of SMEs in the economy.

In this sense, SMEs focus on knowledge absorption, knowledge use, and innovation development. However, their informality makes these processes a natural, tacit, and short-term reaction (Tokman, 1992; Velandia et al., 2016). According to Mcadam and Reid (2001) , SMEs are less advanced, with a mechanical approach to knowledge and a lack of investment in KM approaches and systems, making it challenging to identify the effects of KM practices in these organizations. From the above, the question arises: What variables are affected by KM practices in the context of SMEs?

When trying to answer this question from the empirical approach of SMEs, it is common to find owners of SMEs, who are usually also managers, narrating the feats they perform in the day-to-day of their positions. Knowledge usually remains tacitly in the memory of individuals, making it difficult to objectively measure the variables related to KM practices, their effects on the organization, and the mediating and moderating variables. Then arises another question: What variables have researchers used when addressing KM in the empirical context of SMEs? Which suggests systematically analyzing the existing literature.

To answer this question, an exhaustive search in the Scopus database was carried out, when is confirmed that the literature on KM in SMEs is fragmented and unrelated, with few comparative studies between countries (Figure 1) (Chaithanapat & Rakthin, 2021; Massaro et al., 2016). The descriptors "Knowledge management" and "SMEs" were used for the search without a time limit. The outcome was 304 documents, of which ten correspond to documentary reviews, which are far from identifying the variables affected by KM, thus leaving a knowledge gap.

Note: own elaboration from VOSviewer

Figure 1: Relationship between titles

In this order of ideas, it would be beneficial to characterize the effects of KM practices in SMEs through available literature. The database was refined to 276 documents under exclusion/inclusion criteria. Subsequently, dependent, mediating, and moderating variables related to KM practices in SMEs were identified. This paper presents the theory used for understanding KM in SMEs, subsequently used methodology, and finally, the results. It is relevant to highlight the results suggest a tendency to study the dimensions of KM, such as dynamic technological capability, human capital, and its effect on financial and innovative performance, among others.

The results show the diversity of affected factors for KM in SMEs through 122 open categories defined as dependent variables. Additionally reveals 19 moderating variables and 33 mediators. KM affects organizational capabilities, improves processes, encourages the company to reconfigure itself into a learning organization, influences personnel behavior, drives innovation, and promotes social development. These results contribute to the existing literature for characterizing variables and rearranging the discussion on the subject. In this sense, the study represents an advance because it offers different results from those of the previous reviews and has practical utility because it presents new, concrete, orderly, and valuable knowledge to support decision-making by managers so that they focus on more precise and consistent organizational variables with the nature of SMEs.

2. Knowledge management

From the resources and capabilities theory (Barney, 1991), the organization is understood as a repository of resources (Penrose, 1959; Wernerfelt, 1984), from whose reconfiguration it is expected that the company will develop competitive advantages manifested in capabilities. When these capabilities help respond to the environment's requirement, it is said that the organization has developed dynamic capabilities, which are rare and difficult to imitate because they occur in a particular context putting the company in a privileged competitive position (Teece et al., 1997). In this scenario, knowledge is considered relevant because it constitutes a source of new resources used to build new capabilities. In this order of ideas, knowledge affect organizational performance positively (Bueno, 1998; Escobar et al., 2019; Hock-Doepgen et al., 2021; Kmieciak & Michna, 2018).

It is necessary to differentiate organizational learning from KM. Organizational learning is the natural and spontaneous action of learning from everyday activities by acquisition, internalization, exploitation, and transfer of knowledge (Su & Daspit, 2022), and KM is the action of taking responsibility for learning (Velandia et al., 2022). In this sense, KM is defined in literature as a process, capability, method, or tool to manage knowledge (Bueno, 1998; Davenport & Prusak, 1998; Nonaka & Takeuchi, 1995; Sveiby, 1997; Tarí & Garcia, 2009) which implies the technological component and measurement of knowledge (Barney, 1991; López-Zapata et al., 2019; Probst et al., 2001).

There is a vast literature that presents different variables that are affected by KM. One of the most developed perspectives studies the impacts of KM on financial performance. In this line, it is highlighted the studies of Salojärvi et al. (2005) , Potjanajaruwit (2021) , Khraishi et al. (2022) , and Permatasari et al. (2022) . In addition, the alternative studies of Alegre et al. (2011) , Horvat and Bobek (2017) , Lee and Rosdi (2017) , Alshanty et al. (2019) , Khan et al. (2019) , and Bhatti et al. (2020) , demonstrate the impact of KM on organizational structure, culture, learning, innovation, knowledge absorption, and adaptation capabilities. Agostini et al. (2020) showed that the core topics that consistently link KM, innovation, and networks are performance, absorptive capacity, strategic alliances, and collaboration between organizations. Additionally, the authors reveal KM's human and social side as an emerging topic.

According to the above, are identified two currents for studying KM. The codification approach understands KM as a structure that facilitates tacit knowledge to become models, policies, or formals routines (Explicit Knowledge) (Nonaka & Takeuchi, 1995; Probst et al., 2001), giving place to intellectual capital (Donate & Sánchez de Pablo, 2015; Sveiby, 1997). However, knowledge is built by persons; hence in the human approach, KM is principally understood as a process based on intuition and experience that generates tacit and not very measurable knowledge, which gives place to risk of leakage (Ditillo, 2012; Nonaka & Takeuchi, 1995) but with the capability of leverage the competitive position (Davenport & Prusak, 1998; Scuotto et al., 2020; Tomasini, 2001; Willis & Tucker, 2001).

Therefore, in an organizational context, the existence of KM informal practices is possible, principally in the context of SMEs. Because they are usually faced with resource shortages, they apply their efforts in cheaper and short-time results (Velandia et al., 2016), nevertheless, in current is feasible for SMEs to access big data and implement KM using affordable technological applications to achieve a competitive advantage (Wang & Wang, 2020).

3. Methodology

Qualitative, descriptive, and documentary research was conducted (Guirao-Goris et al., 2008; Peña & Pirela, 2007). The dependent, mediating, and moderating variables used in KM studies in SMEs were revealed through systematic analysis. With the specialized tool for delimiting topics, "Browse by topic" of Science Direct, several search descriptors were used to define the best related to the issue of interest through the keywords declared by the different authors. This action made it possible to identify "Knowledge management" and "SMEs" as the ones that yielded the most consistent results with the study´s objective.

The first articles were downloaded and analyzed in Scopus in June 2020, which were analyzed and categorized in initial exploration work. Subsequently, in May 2022, an update of the download was carried out, which yielded 1592 documents. Both downloads were delimited to "Business, Management and Accounting" and "Economics, Econometrics, and Finance.” This work resulted in 304 papers defined as the database to work with. Table 2 shows the types of documents that make up the initial sample. Authors such as Durst stand out for the number of publications, with seven papers. Adam, Heaving, Soto-Acosta, and Wong with four documents each, thus becoming the most representative. However, among the most cited highlight Riege, Desouza, Awazu, Durst, Bruns, Edvardsson, Alegre, Sengupta, Lapiedra, Mcadam, and Reid, to mention only those with more than 200 citations.

These documents were carefully analyzed, and 28 were eliminated, using as exclusion criteria: duplicate documents, reviews, letters from editors, and unrelated topics. This operation resulted in 276 documents that were rigorously analyzed with the help of analysis perches and analysis logs. In a rigorous categorization process, the variables used by the researchers were identified, giving rise to 122 open categories for the dependent variables, 33 mediators, and 19 moderators, which were grouped according to their common characteristics to make emerge general concepts for achieving the research´s objective (Packer, 2013; Rodríguez et al., 1999). Scopus was used because it integrates about 95% of the articles that enter other databases such as Web of Science (WOS), Science Direct, and others (Cañedo et al., 2010; Velandia & Escobar, 2019).

All selected articles are written in English, and from 2006 publications had a significant rise, remaining relatively constant until 2016; an upward trend was observed, showing that they gradually gained more space in high-impact journals, becoming an essential source of knowledge (Figure 2). 100% of the studies are published in journals whose publishers are from Asia, Australia, Europe, and the United States and correspond to studies that used populations from the same territories. Latin America is at a disadvantage in this regard

Note: Data taken from Scopus.

Figure 2: Behavior of publications

4. Dependent variables in KM studies in SMEs

Once the categories were closed, ten dependent variables were defined that synthesize the 122 open categories, which, considering the approaches of Penrose (1959) ; Wernerfelt (1984) ; Barney (1991) ; Nonaka and Takeuchi (1995) ; Teece et al. (1997) ; Sveiby (1997) ; Bueno (1998) ; Probst et al. (2001) ; Gray (2006) ; Donate and Sánchez de Pablo, (2015) ; Väyrynen et al. (2017) ; Chuan et al. (2017) ; Kmieciak and Michna (2018) ; Escobar et al. (2019) ; Hassan and Raziq (2019) , and Hock-Doepgen et al. (2021) condense critical factors of competitive success. The emerging categories are summarized in Table 3; their behavior is described individually, and the analysis perches are presented.

The "Performance" variable refers to the aspects researchers used to measure sustainable, economic, business, productive, exporting, and innovative performance, among others (Bueno, 1998; Escobar et al., 2019; Hock-Doepgen et al., 2021; Kmieciak & Michna, 2018). The most cited authors are Alegre et al. (2011) , with 223, and Soto-Acosta et al. (2018) , with 127 citations. In 2020 there was a peak in the number of publications that far exceeded previous years. Likewise, the "innovation" variable refers to aspects such as innovative work behavior, relocated innovations, frugal innovation, green innovation, and efficiency in the innovation model, among others (Alegre et al., 2011; Alshanty et al., 2019; Chuan et al., 2017; Hock-Doepgen et al., 2021; Khraishi et al., 2022). Along these lines, highlight the studies of Martinez-Conesa et al. (2017) with 178 citations.

On the other hand, in the "Reconfiguration in learning organizations" variable, topics such as transfer, creation, integration, capture, and knowledge management, among others, are considered representative of the organizational decision to become a learning organization (Davenport & Prusak, 1998; Hall, 1979; Pratten, 1991; López-Zapata et al., 2019; McAdam & Reid, 2001; Mundy, 2016; Su et al., 2022; Tarí & García, 2009; Velandia et al., 2022; Xie et al., 2020). The most influential studies are those of Wong and Aspinwall (2005) , with 331 citations. Regarding the "Improvement of processes" variable, it is observed that the most significant number of publications was in 2018 (Davenport & Prusak, 1998; Riege, 2005; Saratchandra & Shrestha, 2022; Scuotto et al., 2020; Soto-Acosta et al., 2018; Tomasini, 2001; Willis & Tucker, 2001). In this line, Basly (2007) stands out with 80 citations. This area involves elements such as internationalization strategies, misalignment between knowledge and KM systems, access to international purchases, and empowerment of the organization.

There are some aspects related to the company´s culture, information assimilation, knowledge absorption, communication, technologies (ICT), employee commitment, and empowerment of individuals, among others, that can be grouped into a category that is related to the "Behavioral" (Alegre et al., 2011; Alshanty et al., 2019; Bhatti et al., 2020; Desouza & Awazu, 2006; Horvat & Bobek, 2017; Khan et al., 2019; Lee and Rosdi, 2017; Nguyen & Mohamed, 2011; Wong & Aspinwall, 2005). In this line, the study of Grace (2009) stands out. This publication has 150 citations. The years 2017 and 2018 have the highest number of publications.

There are factors related to organizational transformation by the demands of the environment, for example, preparation for internationalization, knowledge of internationalization, strategic renewal, and adoption of electronic commerce in the supply chain, among others, giving place to the “Organizational transformation” category (Castagna et al., 2020; Chaithanapat & Rakthin, 2021; Grace, 2009; Riege, 2005; Wang & Wang, 2020). The number of publications in this line is not so abundant, revealing a line available. The most representative study is that of Villar et al. (2014) . The paper has 149 citations.

The "Social development" category has the lowest number of publications (9 articles), which address aspects such as economic growth, quality of life in the regions, sustainability, growth of women entrepreneurs, and strategic entrepreneurship, among others related to the human and social side of KM (Agostini et al., 2020; Petrov et al., 2020). This line of work does not present publications between 2009 and 2015 and has only one published article per year. It is deduced that these little visible lines constitute an opportunity to formulate news studies in SMEs from a different approach, which can be innovative and profitable. The study of Singh and Kumar (2020) is highlighted with 23 citations.

From resources and capabilities current (Barney, 1991; Ditillo, 2012; Durst & Wilhelm, 2012; Hardwig, 2020; Hutchinson & Quintas, 2008; Martinez-Conesa et al., 2017; Penrose, 1959; Pratten, 1991; Teece et al., 1997; Weaven et al., 2021; Wernerfelt, 1984), the "Organizational capabilities" category has been defined, made up of the dimensions: dynamic capabilities, survival, and growth, technological innovation capability, brand capabilities, absorption capability, and innovation capability, among others. In this line, highlight the study of Yao et al. (2020) with 24 citations.

Studies located in the "Management" category are made up of decision-making, sense of opportunity, owner of the SMEs, and risk management (Chaithanapat & Rakthin, 2021; Desouza & Awazu, 2006; Khraishi et al., 2022; Permatasari et al., 2022; Potjanajaruwit, 2021; Salojärvi et al., 2005; Velandia et al., 2016). The low production in this category suggests that researchers have paid little attention to aspects concerns related to managers, which opens the door to a fertile perspective of study with the potential to improve the conditions of SMEs in general significantly. The study by Wee and Chua (2013) is the only one cited and has 118 citations.

Finally, the category named “Convergence” is made up of knowledge exchange, transparent interoperability, platforms that induce networks and conversations, and a stable and long-term business network strategy (Castagna et al., 2020; Saratchandra & Shrestha, 2022; Valkokari & Helander, 2007; Wang & Wang, 2020). The little scientific production in this line demonstrates the need for studies on how ICTs contribute to the consolidation of common information banks as a strategy to increase available information for SMEs. This perspective opens questions such as: Does publicizing KM information generate competitive benefits or disadvantages by revealing sensitive organizational information? Are the companies that use these platforms ethical when sharing and using the information? Likewise, this group's most representative study (37 citations) comprises Soto-Acosta et al. (2014) .

Below are the analysis perches that contain the complete relationships between the closed and open categories linked by the previous theories. Table 4 includes the most widely represented approach in the literature, in which researchers study the effect of KM on organizational performance dimensions. The high production in this category can be explained by the various options to measure performance, going beyond the financial sphere by including qualitative indicators.

Table 5 refers to the effects of KM on the actions and performance of innovation. As SMEs are companies that naturally stimulate creativity (Hutchinson & Quintas, 2008; Rodríguez-Ponce et al., 2010), this approach is especially relevant when revealing categories for performance such as: “Frugal innovation,” “Green innovation,” or “Delocalized innovations.” On the other hand, the closed category “Actions” refers to the actions of innovation, such as flexibility, efficiency, or implementation.

Thanks to their typically organic structure, where human resources become relevant, SMEs have a high capability for reconfiguration, adaptation, and evolution. In this sense, the study of the effect of KM on the ability to become true learning companies is approached from "Learning processes,” "Personnel,” "Management," and "Knowledge" (Table 6).

On the other hand, the effects of KM on process improvement are addressed from an ex-ante approach when it refers to its activity, emphasizing the presence of elements such as: "Participation in public tenders" or "Access to international purchases,” for instance. And an ex-post approach when referring to results such as "Value creation" or "Business excellence," among other elements of Table 7.

Table 8 shows the studies that address the effect of KM on behavior focus on individuals and organizations. The former refers to variables such as "Knowledge governance acceptance" and "Commitment of individuals.” The latter refers to aspects of the organizational culture that materialize because of KM, which helps educate staff, reduce barriers to managing knowledge, and reduce the probability of leakage.

Table 9 presents the variables used to analyze KM's effect on the company's capabilities. They give rise to the emerging category "Organizational capabilities," which combines all the elements the company uses to respond to the environment and acquire competitive advantages.

Table 10 brings together the effects that KM has on the emerging categories of "Organizational transformation," "Social development," "Convergence," and "Management" of SMEs. The first refers to the use of flexibility and the ability to absorb technologies and use them to create competitive advantages, thus, allowing export activities and administrative reconfigurations, among others.

The second refers to elements of general interest; that aim at social benefit or the development of the economy. Under the shadow of this category are variables such as "Growth of women entrepreneurs," "Quality of life in the regions," or "Sustainability," which make the emerging category "Social development" take on a shade of justice and emancipation.

In the “Convergence” category are grouped the variables refer to actions and elements aimed at sharing knowledge that is affected by KM activities in SMEs. The implementation of "Platforms that induce networks and conversations" is highlighted, which seeks to improve competitiveness for a group of companies by constituting a storage site for shared knowledge and creating channels that increase the capability of absorption, transfer, and transformation from tacit knowledge to explicit and socially available knowledge.

Finally, the emerging category "Management" groups those variables affected by the KM related to business administration. KM affects how managers manage risks, interact with employees, make decisions, and see opportunities. This category is compliant for flexibility, ease of reconfiguration, commitment-based HR practices, and informal knowledge management of SMEs.