In most developing economies, the informal sector is growing remarkably, claiming a significant share of economic activity. Likewise, in Zimbabwe, the sector accounts for an estimated 60.6% of the total gross domestic product (GDP) and 90% of the national labour force (Medina & Schneider 2018). Although most of the businesses are largely unregistered and some operating in illegal premises, there is significant evidence of entrepreneurship and innovation flourishing in the informal sector of Zimbabwe. According to Muzurura (2019), Zimbabwe’s entrepreneurial ecosystem remains very fertile, owing to the high literacy levels in the country and the high levels of resilience and ability to identify opportunities, incubate new startups and take up calculated risks. The government has responded by embracing informality and making it a policy priority and top of the inclusive growth and sustainability agenda (Mujeyi & Sadomba 2019). Although commendable efforts have been made by the government in this regard, cumbersome bureaucracy, stringent regulations and corruption rentals have slowed down the pace of entrepreneurship growth in Zimbabwe. Muzurura (2019) states that corruption is amongst the impediments to entrepreneurship in Zimbabwe, which tends to have a sanding instead of greasing ‘effect on the wheels of entrepreneurship growth’.

Social capital, defined as a set of socio-structural resources derived from social relationships that enable individuals, groups or teams, coordinates action to desired outcomes (Carrillo Álvarez & Riera Romaní 2017) has increased in understanding social and business relations. According to Suseno (2018), the underlying idea behind social capital is that actors’ relationships and ties with others matter, as they help provide resources or benefits to create value and positive outcomes for the organisation. Social capital has helped explain firm growth, success and resilience, sparking overwhelming interest amongst practitioners and scholars. In this sense, social interaction, social presence and relationship quality will positively impact an organisation’s wealth by reducing transaction costs and enabling collective actions (Andriani & Christoforou 2016). Building on the work of Kreijns and Kirschner (2002), social capital can be measured in terms of sociability and social presence of the entrepreneur and their level of online social interaction.

According to Kreijns et al. (2007), sociability relates to the extent to which an individual or an environment facilitates the emergence of a sound social space characterised by trust, a strong sense of belonging and good working relationships. Henriksson (2019) expands on the above definition by stating that sociability relates to the informal interaction between individuals without the influence of any subject matter external to the current situation. It is the kind of interaction that foregoes the creation of a group to maintain the situation open to all forms of interaction. In this context, sociability can be said to relate primarily to fitting in and adapting oneself to the situation and foregoing one’s personal motivations for being present (Henriksson 2019). Evidently, the type and superiority of connections between stakeholders will influence the business’s nature and quality performance outcomes (McLeod & MacDonell 2011). Sociability is thus a critical element in enhancing relationship building and collaboration amongst entrepreneurs. In the current world of the COVID-19 pandemic, sociability has been identified as a critical factor in shaping disease susceptibility and how fast actors respond to its spread (Borgonovi & Andrieu 2020). The development of new technologies has presented entrepreneurs and firms with a greater opportunity to create emotional connections with stakeholders and safeguard consumer loyalty (Olavarría-Jaraba et al. 2018).

Social presence has been described as the feeling of being socially present even though one is not physically in a particular space (Kim, Song & Luo 2016). The operationalisations of social presence have been expanded to include social interaction, immediacy, intimacy, emotion and connectedness. Scholars like Short, Williams and Christie (1976:65) have conceptualised social presence as the ‘degree of salience of the other person in the interaction and the consequent salience of the interpersonal relationships’. Thus, the degree of social presence relates to the subjective experience of presence by the natural person, both in terms of thoughts and emotions (Lowenthal & Snelson 2017). The capacity to identify social interactions is embedded in humans and is manifested in the primary stages of development, affecting both cognition and behaviour (Choi & Luo 2015; Hamlin et al. 2011; Tatone, Geraci & Csibra 2015). There is a growing consensus that social interactions play a vital role in influencing human behaviour (Crawford & Harris 2018). The informal connection amongst individuals plays a significant role in promoting engagement and social presence. It is believed that significant social interactions transversely across a cognitive space are made possible by the existence of trust (Hoffmann & Söllner 2014; Ojansivu & Alajoutsijärvi 2015; Park & Lee 2014), which enables the parties to make a clear understanding of each other’s perceptions (Grabher & Ibert 2014; Mueller 2015). Furthermore, McLeod and MacDonell (2011) argue that the prerequisite to understanding the explanatory organisational and individual parameters like social interactions management may not be overemphasised.

In recent years, online social media interaction, through social networking sites (SNSs), has evolved to become an integral part of people’s daily lives worldwide, altering the way people communicate and socially interact (Moretta & Buodo 2018). The general increase and expansion in information and communication technologies have significantly improved the way individuals and businesses interact, team up and network (De Choudhury et al. 2010). These SNSs facilitate the culmination and continuance of social interactions online, and they have become entrenched ways in which people communicate today (Khalis & Mikami 2018). Firms have also started using social media platforms like Facebook and Twitter to smoothen the advancement towards social interaction amongst customers and employees beyond the context of corporate emails and memos (Burrus 2010). Social interaction refers to the process of behaving and conversing with one another, predominantly in an informal way (Choi & Kim 2004). Firms that can organise resources such as information and social influence potentially benefit these resources depending on the firm’s ability to mobilise these resources in its favour (Herrero 2015).

Relationship quality relates to the strength of a relationship between parties, including the satisfaction, trust, assurance and social benefit between the parties involved (Japutra, Molinillo & Wang 2018). Relationship quality has been theorised as a higher-order construct consisting of several distinct dimensions, including trust and satisfaction, which are considered an ‘emotional state that occurs in response to an evaluation of these interaction experiences’ (Huntley 2006; Liu, Guo & Lee 2011; Su, Swanson, & Chen 2016). Relationship quality is considered a factor in developing loyal consumers, employees and suppliers (Walsh et al. 2010; Lo et al. 2018) and tourists’ environment-related conduct (He et al. 2018). Empirical studies show that relationship quality is valuable in predicting customer behaviour (Hyun 2010; Jin et al. 2013) and is central to organisational decision-making (Morgan & Hunt 1994). Other scholars have indicated that relational quality is developed by enhancing stakeholder experience (Hyun 2010). Whilst the central view is that business relationships are determined by legislative limits and measures such as partnerships and alliances, hands-on management of relationships is of tactical and strategic importance (Jelodar, Yiu & Wilkinson 2016). In addition, as stated by Feeney and Fitzgerald (2018), lack of emotional attachment, insecurity and connected conflict behaviours wear down relationship quality. Therefore, at the core of relationship quality is recognising and fulfilling the needs of parties in a relationship, and relationship quality theory suggests that satisfied stakeholders reward with continuing value over time (He et al. 2018).

The notion of ‘entrepreneurial intensity’ (Morris, Kuratko & Covin 2010) was developed to evaluate the overall altitude of entrepreneurship amongst firms and individuals, and it captures the degree and frequency of entrepreneurship. It has been defined by Morris and Sexton (1996) as the amalgamation of the frequency of entrepreneurial activities and the degree of entrepreneurship, measured in terms of innovation, risk taking and proactivity. There are three dimensions of entrepreneurship in EI: innovativeness, risk taking and proactiveness (Covin & Slevin 1989; Miles & Arnold 1991). The innovativeness dimension entails the tactical pursuit of creative and novel solutions to problems and needs within a particular context (Prakash, Jain & Chauhan 2015). The risk-taking subconstruct entails the entrepreneur’s readiness to commit a considerable amount of resources to the opportunities that arise although such opportunities have a reasonable chance of failure (Morris 2015). The risks taken by the entrepreneur are, however, calculated and manageable. The proactiveness dimension relates to the considerable determination, flexibility and capacity to look into the future (Prakash et al. 2015). Frequency in EI is understood as the number of entrepreneurial events that an entrepreneur undertakes. The term EI, therefore, refers to the variable nature of entrepreneurship within an individual. The EI notion is derived from the postulation that entrepreneurial conduct varies in its innovativeness, proactiveness and risk-taking characteristics (Prakash et al. 2015). Market irregularities provide the potential for entrepreneurs to create monetary returns by introducing new goods and services, markets, processes and raw materials previously unavailable (Kirzner 1979; Shane 2003). According to Shu et al. (2018), opportunity recognition is generally modeled as the outcome of a wide-ranging search that is targeted in a particular direction where an opportunity is to be discovered, and by occupying a central position, an entrepreneur is more likely to make sense of ambiguous situations and validate the information from different sources.

This study’s theoretical underpinnings draw inspiration from the SET, an influential theoretical paradigm that promotes the idea of two-sided relationships and reciprocal resources exchanging behaviours (Miles 2012). The SET has been understood as ‘a general sociological theory concerned with understanding the exchange of resources between individuals and groups in an interaction situation’ (Ap 1992:668). According to the SET, an entrepreneur’s ability to establish high-quality business and social relationships are critical. It ensures that they benefit when the other party reciprocates the relationship by providing favours, helping the entrepreneur in market sensing and developing a cognitive structure for selecting and developing opportunities (Shu, Ren & Zheng 2018). Within SET, parties are engaged in an exchange relationship because of the anticipated benefits expected where economic (product, service and knowledge) and social (friendship and reputation) resources are likely to be exchanged amongst different actors (Miles 2012).

Based on the SET, this study suggests that social capabilities derived from business sociability, the entrepreneur’s sociability, social presence and online social interactions will influence the firm’s relationship quality whilst influencing EI. The underlying proposition is that to increase the entrepreneurs’ EI, they must improve their social capabilities. The direct and indirect relationships between social capabilities and EI are represented in Figure 1.

There is proof in the existing literature that there is a relationship between social and psychological dimensions and organisational decision-support systems. These dimensions aid in the decision-making processes (Liberatore & Quijano-Sanchez 2017; Quijano-Sanchez, Díaz-Agudo & Recio-García 2014), hence the above pictorial representation.

The improvement in social capabilities achieved by the sustainable growth of interactions and relationship management is believed to enable the identification, organisation and resolution of socio-economic social challenges (Lee 2018). The recent development of smartphone technology has enabled instant interactions with friends and business partners, thereby potentially allowing improved benefits of social interactions (Dwyer, Kushlev & Dunn, 2018). It has been empirically proven that social interactions and high-quality relationships are significant in many facets. O’Sullivan and Walker (2018) and Gonzalez et al. (2018) came to a similar conclusion that employment opportunities in developed economies are significantly found through informal contacts with colleagues, family or job service providers. Empirical research has proved that strong social relationships are valuable for individuals in providing social support (Zellweger et al. 2018) and comfort (Dunn 2018; Joly & Connolly 2019), connections (Whelan, Hingston & Thompson 2019), in supporting a positive mind-set and regularity in behaviour (Farrow & Yuan 2011). In essence, the development network and social relationships are key factors influencing firms’ and entrepreneurs’ efficient performance (Prashantham et al. 2019; Smith, Smith & Shaw 2017). Based on the arguments and empirical evidence, this study predicts that entrepreneurs who nurture their social capabilities are associated with high EI levels. Hence:

This study on the relationship between social capabilities and EI adopted a positivistic paradigm and a quantitative approach. The positivist paradigm choice was influenced by the need to objectively and hence quantitatively analyse data to confirm the proposed relationships. According to Burns and Grove (1993), a quantitative approach is a formalised, objective process of describing and testing relationships amongst variables of interest. In terms of the research design, this study adopted a cross-sectional survey design to confirm the relationship’s direction and magnitude without experimental manipulation.

Based on the conceptual model, a survey questionnaire was designed to measure the proposed constructs that contribute to EI. To ensure both validity and reliability of the measurement instrument, the items used to operationalise each construct were adapted from the previous literature. The questionnaire’s measurement items were subjected to intensive review by three research experts to ensure reasonable consistencies, simplicity and contextual significance before distribution. Based on the reviewers’ comments, the items were slightly modified. Furthermore, a pilot test was conducted involving 13 SMEs owners and managers, and also, based on their observations and suggestions, the items were modified accordingly. The questionnaire had four sections: the first section (Section A) looked at the respondent’s biographic data, Section B solicited information on the Social Interaction construct and its three subconstructs, namely sociability, social presence and online social interaction. Sections C and D looked at relationship quality and EI, respectively. All construct items were assessed using a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree).

The population of interest in this study was SMEs, and to make the group manageable and feasible, the sampling frame was confined to those SMEs’ owners and managers who are part of an online social group like WhatsApp or Facebook group. Theoretically, the population to be studied should comprise all SMEs in Zimbabwe but access to information, cost and time constraints, and privacy concerns make it difficult to include all SMEs. Therefore, a non-probabilistic sampling procedure (convenience sampling) was used to collect the data. An online questionnaire constructed using Google Docs was posted on Masvingo-centred Facebook and WhatsApp groups from 15 October 2018 to 19 December 2018. To eliminate repeated responses, responses with duplicate IP addresses were deleted, and this was done with the help of information systems experts and following the recommendations of Zhao et al. (2016). Of the 523 questionnaires retrieved in this web-based survey, 312 questionnaires were valid, giving an estimated response rate of 59.66%. The demographic characteristics of the participants are presented in Table 1.

The study results indicate that most of the respondents were aged between 26 and 50 years (40%), followed by those below 25 with 33%. The general thrust towards entrepreneurship in Zimbabwe could explain why there is a sizable number of entrepreneurs below 50. In addition, most of the respondents are fairly educated. Seventy per cent of the respondents have an undergraduate degree although a small number of them have moved on to acquire a postgraduate qualification (10%). The businesses that these respondents represent have generally low levels of turnover. Most of the firms are below $9000 in annual turnover, with only a small number (29%) scoring above $9000. This could be a result of the fact that SMEs are generally undercapitalised.

In this study, data analysis was conducted using the co-variance based structural equation modelling technique (SEM). Structural equation modelling is a statistical and multivariate technique that uses pictorial representations to hypothesise complex relationships between independent constructs and the dependent variables. It allows the empirical assessment of hypothesis with or without latent variable structures, using statistical measures such as coefficient of determination R² (Mostafa & Roorda 2019). This technique is growing in recognition, and several studies have adopted this technique across different disciplines demonstrate (Hair, Babin & Krey 2017; Ringle, Sarstedt & Straub 2012). The self-developed questionnaire was validated through confirmatory factor analysis (CFA), whilst the interrelationships amongst all variables were determined by using path analysis.

This article followed all ethical standards for a research without direct contact with human or animal subjects.

A CFA was conducted using AMOS 25.0 to validate the measurement items for both reliability and validity. Reliability was examined using Cronbach’s alpha (α), composite reliability (CR) and average variance extracted (AVE) for each construct. All constructs had Cronbach α values greater than 0.7, which is regarded as the minimum value for confirming reliability in social studies (Taber 2018). The results presented in Table 2 indicate that Cronbach α ranged between 0.741 and 0.927, indicating the acceptable reliability of constructs. A CFA was applied to assess the constructs’ construct validity with AMOS 25.0. According to Campbell and Fiske (1959), there are two key construct validity dimensions: convergent and discriminant validity. In assessing convergent validity, three criteria recommended by Fornell and Larcker (1981) were applied, namely all factor loadings should exceed 0.7, CR should exceed 0.7 and the AVE for each construct should exceed 0.5. All three conditions for convergent validity were met, as presented in Table 2.

According to Gefen and Straub (2005:92), ‘discriminant validity is shown when each measurement item correlates weakly with all other constructs except for the one to which it is theoretically associated’. Discriminant validity was assessed following the recommendations of Fornell and Larcker (1981). Remarkably, the Fornell–Larcker criterion is regarded as a benchmark approach in assessing discriminant validity in covariance-based SEM (Shah & Goldstein 2006; Shook et al. 2004). In this study, as indicated in Table 3, the square root of the AVEs exceeded all other cross-correlations between the constructs in the model, indicating acceptable discriminant validity.

The goodness of fit indicates how well the model fits the data acceptably, and in this study, this was evaluated using the recommendations of Hu and Bentler (1999). The χ²/df was lower than 2 (1.951), whilst the root mean residual (RMR) and the root mean square error of approximation (RMSEA) were both below 0.08 (0.056 and 0.059, respectively). Acceptable model fit is indicated by an RMSEA value of 0.06 or less (Hu & Bentler 1999). The comparative fit index (CFI) for the model was 0.964, and the acceptable model fit is indicated by a CFI value of 0.90 or greater (Hu & Bentler 1999). Overall the model fitted well with the data, allowing for path analysis and hypothesis testing.

A structural model path analysis was conducted to evaluate the hypothesised relationships amongst the constructs of interest. In line with the criteria suggested by Hu and Bentler (1999), the findings suggest good model fit (χ²/df = 1.950, RMR = 0.072, RMSEA = 0.066, GFI = 0.925, AGFI = 0.889, NFI = 0.945, RFI = 0.936, IFI = 0.961, TLI = 0.955, CFI = 0.964).

According to the results stated in Table 4, Sociability has a positive and significant relationship with relationship quality and so is social presence with relationship quality (β = 0.586 and β = 0.624 respectively) at p < 0.001 level of significance. This implies that both H1 and H2 are fully supported. The other hypothesis H3 proposed is a positive and significant relationship between online social interactions and relationship quality. The results indicate that the β coefficient for the path was 0.798, p < 0.001, which others support for H3. Sociability and online social interactions both had results supporting their respective hypotheses (β = 0.456, p > 0.001 and β = 0.542, p < 0.001), implying that both were found to be positively influencing EI and upholding H4 and H5. Like all the other hypotheses, H6 was supported at β = 0.340, p > 0.001, implying a positive and significant relationship between relationship quality and EI.