The significance of SMEs to the economic growth and development of both developed and developing countries is recognised across the world (Nasr & Rostom 2013). SMEs require little capital to set up and have the potential to create jobs for the burgeoning number of unemployed youths in most developing countries (Gill & Biger 2012; Sonobe, Akoten & Otsuka 2011). Manufacturing SMEs boost economic activities in the areas in which they operate, because they produce goods that meet the immediate needs of locals, thus serving a market that the large enterprises have ignored (Chege et al. 2016; Sonobe et al. 2011). In Kenya, manufacturing SMEs account for about 80% of the number of firms in the manufacturing sector and employ over 80% of the workforce in this sector (Kenya National Bureau of Statistics 2016). To further emphasise the importance of SMEs to development, the Kenyan government recognised the sector in its ‘Kenya vision 2030’ strategic plan for its role in the country’s industrialisation agenda (Government of Kenya 2007).

There are case studies of countries across the globe such as China (Chen 2006), Malaysia (Karikomi 1998) and India (Das 2008) that have industrialised by focusing on small and medium-sized manufacturers. This implies that countries in sub-Saharan Africa such as Kenya can also facilitate industrialisation by focusing on the development of manufacturing SMEs. However, as highlighted earlier, SMEs face resource challenges, making it difficult to own capabilities that can enable efficient and effective operations (Bowen et al. 2009). Some of the capabilities required by SMEs, especially in logistics, can easily be accessed through outsourcing (Murphy et al. 2012). It is expected that SMEs are more likely to focus on the quality of goods and distribute widely at lower costs, resulting in improved firm performance, once they contract expert logistics service providers (LSPs) (Soinio, Tanskanen & Finne 2012).

Logistics outsourcing refers to the transfer of all or part of the logistics functions to be performed on behalf of the firm by third-party logistics service providers (Lieb & Randall 1996; König & Spinler 2016; Van Laarhoven, Berglund & Peters 2000). Pratap (2014) argued that logistics outsourcing can best be explained by the RBV and TCE theories. The RBV theory holds that an enterprise can acquire resources and capabilities through outsourcing to meet its customers’ needs (Wong & Karia 2010). Similarly, the TCE theory highlights that logistics outsourcing provides an avenue for conducting business at lower transaction costs when compared to in-house operations, thus improving firm performance (Bolumole, Frankel & Naslund 2007; Pratap 2014). The RBV and TCE theories have been applied extensively in logistics outsourcing research (Liu et al. 2015; Pratap 2014), depicting their relevance to guiding development of predictive models in logistics management. Past research has identified the increased importance of logistics across the globe in determining overall firm performance, as supply chains become complex (König & Spinler 2016), thus reinforcing the importance of this study.

The growing significance of logistics among firms can further be revealed by the increased spending and its central role in improving customer service (Langley & Capgemini 2016). As logistics performance becomes more significant, firms are expected to focus on their core business areas, leaving logistics to be outsourced to expert LSPs (König & Spinler 2016; Rahman & Wu 2011). Using LSPs implies that the firms (in this case manufacturing SMEs) would access the logistics capabilities they lack in-house, as they focus on their core manufacturing activities to improve performance (König & Spinler 2016; Langley & Capgemini 2016; Murphy et al. 2012). Thus, it is hypothesised that:

Logistics outsourcing is also adopted to reduce logistics costs and as a long-term strategy to increase customer satisfaction and improve overall enterprise performance (Fawcett, Magnan & McCarter 2008; Lee, Lin & Cheng 2013). Logistics activities that are commonly outsourced can be classified into operational (transportation, fleet management, clearing and forwarding), information processing (logistics information system, procurement and order management, product track and trace), and strategic and value-adding services (inventory, warehousing and packaging management) (Langley & Capgemini 2016; Liu et al. 2015; Solakivi et al. 2011).

Firm performance refers to how well a firm achieves its overall goals, both financial and non-financial (Kasie & Belay 2013; Quang et al. 2016). Financial measures such as return on assets (ROA) and profitability are objective as they make use of actual figures, whereas the non-financial measures, such as customer satisfaction, use perception and are generally subjective (Tseng & Liao 2015; Yang, Marlow & Lu 2009). Measuring firm performance is not an easy task as one should select the most appropriate measures for the industry, period (long-term and short-term) and firm size to achieve the desired results (Kasie & Belay 2013; Quang et al. 2016). As such, because SMEs lack clear management structures, the specific performance metrics selected, should be growth based and should focus on financial liquidity and customer satisfaction (Marchand & Raymond 2008; Raymond et al. 2013). Therefore, this study measured performance of manufacturing SMEs through growth in sales, profits, ROA, return on capital employed (ROCE), earnings before interest and tax (EBIT), number of employees and customer satisfaction (Solakivi et al. 2011; Tseng & Liao 2015).

High-performing SMEs are likely to deliver goods to meet local demand, create employment and improve living standards of communities where they operate (Chege et al. 2016; Nasr & Rostom 2013). Thus, improving the performance of manufacturing SMEs will go a long way to improving not only livelihoods but also economic growth. This study highlights that logistics presents the potential to improve the performance of SMEs if managed diligently (Waugh & Luke 2011). Logistics outsourcing has emerged as one such strategy that can be used to realise efficient and effective logistics operations, as the expertise of third-party logistics (3PLs) can be utilised (Waugh & Luke 2011). Although the common view is that logistics outsourcing leads to improved enterprise performance (Fawcett et al. 2008; Langley & Capgemini 2017), the actual relationship between logistics outsourcing and firm performance has not been established precisely in past research. Some past research has observed that logistics outsourcing improved firm performance (Lee et al. 2013; Parashkevova 2007), whereas Cho et al.’s (2008) study reported a negative relationship between logistics outsourcing and firm performance and Solakivi et al.’s (2011) study found no relationship with firm performance among Finnish SMEs. Therefore, this study sought to empirically investigate the relationship between logistics outsourcing and performance of small and medium-sized manufacturing enterprises in Nairobi.

Logistics outsourcing allows the SMEs to access capabilities of expert LSPs to achieve high logistics performance in terms of reduced cycle times, reduction of wasteful operations, increased flexibility, delivery timeliness and smooth operations in upstream and downstream activities (Green, Whitten & Inman 2008; Zailani et al. 2017). Green et al. (2008) argued that increased logistics performance positively influences firm performance. As such, it is hypothesised as follows:

Logistics outsourcing performance refers to the extent to which LSPs meet a firm’s expectations (Wagner & Franklin 2008). Logistics outsourcing processes that take into consideration communication, trust levels, cooperation and innovation are likely to result in high logistics outsourcing performance (Deepen et al. 2008; Križman & Ogorelc 2010; Oshri, Kotlersky & Gerbasi 2015). In addition, Leuschner et al. (2014) highlight that logistics outsourcing performance positively affects logistics performance and consequently firm performance. Therefore, it is hypothesised as follows:

The relationship is investigated directly and indirectly through logistics outsourcing performance and logistics performance as mediating variables. The measures of logistics outsourcing performance in this study focus on the quality of communication and cooperation between SMEs and 3PLs (Deepen et al. 2008), trust levels (Huo, Ye & Zhao 2015) and innovative capabilities of the 3PLs (Oshri et al. 2015). Logistics performance is measured by operational efficiency and customer service indicators (Zailani et al. 2017). Finally, it is hypothesised as follows:

These relationships are conceptualised, as illustrated in Figure 1.

The model (Figure 1) was used to test six hypotheses using structural equation modelling. Additionally, mediation hypotheses were tested to measure the indirect effect of logistics outsourcing (LO) on firm performance (FP) through logistics outsourcing performance (LOP) and logistics performance (LP). Thus, the study tested the following hypotheses:

The uniqueness of the SME context from industry to industry and even country to country calls for combined methods to better address the research questions, as claimed by Saunders, Lewis and Thornhill (2016). This study therefore adopted a pragmatist’s philosophical orientation, which allows for a mixture of qualitative and quantitative research approaches in a single study (Takkashori & Teddlie 2010), that is, a mixed methods research design (MMD). MMD refers to the use of both qualitative and quantitative approaches to address a research problem (Creswell & Clark 2007; Jogulu & Pansiri 2011). The use of this design helped to answer the ‘why’ and ‘how’ research questions that required qualitative data and the ‘what’ question which required quantitative data. The quantitative approach helped to test the application of RBV theory to logistics outsourcing. The results from the qualitative approach enriched the quantitative findings and helped advance logistics management literature regarding performance of SMEs. The results from the two approaches were interpreted together to provide a comprehensive response to the research questions as opposed to using either of the approaches (Saunders et al. 2016). Convergent parallel design variant of MMD was used in this study, whereby quantitative and qualitative data were collected and analysed concurrently (Jogulu & Pansiri 2011). The qualitative approach was used to collect data that helped interpret the relationships tested between logistics outsourcing and firm performance. These two approaches were thus able to complement each other. Upon analysis of the data, the findings were interpreted together, whereby results from the qualitative approach helped explain the relationship with logistics outsourcing, hence addressing the research problem adequately (Creswell & Clark 2011). Specifically, qualitative data helped explain why SMEs outsource logistics services as well as the process they follow to outsource. This provided insights of the kind of statistical relationships that could be expected between logistics outsourcing and performance of the SMEs. The combination of qualitative and quantitative approaches overcame the shortcomings such as the bias of relying on a single approach, by providing triangulated data to study the research problem (Creswell & Clark 2011). The sampling techniques, instrument development and data collection are discussed in the following sections starting with the qualitative approach.

Purposive sampling was used to select five manufacturing SMEs, which participated in the interviews. It was used to select the manufacturing SMEs to provide rich data (Creswell 2014) regarding logistics outsourcing and firm performance. The enterprises selected represented the five main manufacturing categories in Kenya, as identified by Chege et al. (2016). Collecting data from multiple SMEs helped cross-checking to ensure that credible data were obtained.

Qualitative data were collected through face-to-face interviews using a semi-structured interview guide at the manufacturing SMEs’ site. Conducting the interviews on site helped observe that actual manufacturing was taking place. The interview guide was piloted with two logistics managers. The piloting helped to adjust some questions which were not clear and drop others that were repetitive. In-depth interviews were conducted with directors or owners or managers in charge of logistics in their respective manufacturing SMEs. One interview was conducted in each of the SMEs. The selected respondents were those with all the relevant information regarding the enterprises’ logistics management practices. The five interviews were sufficient as there was no additional information obtained after the fifth interview. The five interviews were sufficient, considering the claim by Creswell and Clark (2007) and Jogulu and Pansiri (2011) that four to five case studies are appropriate to meet the requirements of a mixed methods design. Three interviews were recorded using an audio recorder, whereas two were captured in interview notes.

The qualitative data were analysed through thematic analysis to answer the why and how research questions (1 and 2, respectively). Thematic analysis involved searching for themes across the interview data, as described by Saunders et al. (2016). The qualitative data collected through in-depth interviews were transcribed and coded. Themes related to the research questions 1 and 2 were identified. The research process followed in the qualitative approach is outlined in Figure 2.

The units of analysis comprised manufacturing SMEs in Nairobi. The study population encompassed 406 manufacturing SMEs in Nairobi as of 01 March 2015 as per the Nairobi City County (NCC) Licensing Department. Only SMEs engaged in manufacturing within NCC were counted as units of analysis. SMEs within the manufacturing sector that were engaged in trading or were service orientated were excluded. As such, actual manufacturing was observed during data collection, otherwise the enterprise was excluded. All other manufacturing firms were included in the study, regardless of the use of 3PL services. This inclusivity allowed for the determination of the extent of logistics outsourcing within the SMEs.

Quantitative data were collected using a seven-point Likert-type survey questionnaire with end-points defined to help respondents understand the scale. Although there is no preferred number of points that specific rating scales should have, Krosnick and Presser (2010) argue that a lengthy scale (e.g. 7-point scale) may increase data validity compared to a 5-point scale. In addition to this, the 7-point scale was considered appropriate for this study because of the successful application of the same scale length in past studies (Cho et al. 2008; Liu et al. 2015). The questionnaire collected data on enterprise demographics, logistics outsourcing, logistics outsourcing performance, logistics performance and firm performance. Logistics outsourcing was measured by six items, which were self-developed in line with Langley and Capgemini (2016) and Solakivi et al. (2011). Logistics outsourcing performance was measured by 15 items comprising communication, trust, cooperation and innovation indicators (Deepen et al. 2008; Križman & Ogorelc 2010). Logistics performance was measured using five items adapted from Green et al. (2008). Similarly, firm performance was also measured by five items (Solakivi et al. 2011; Yang et al. 2009). The questionnaire was piloted among ten manufacturing SME managers involved in the day-to-day management of logistics in their enterprises. The piloting exercise led to rewording of some items to make them clear before the data collection.

The drop and pick later method was used to collect quantitative data from SME managers in charge of logistics operations. In some cases, the questionnaires were filled on the spot. In instances of collect later, three attempts (spaced one week apart) were made to collect the filled questionnaire. Any questionnaires that were not collected within the three attempts were counted as not returned. The method was appropriate as only the physical and postal addresses of the manufacturing SMEs were available at the NCC Licensing Department. Although some respondents promised to email or send the filled questionnaires through postal services, none of these were received through post or email.

The quantitative data were analysed using partial least squares structural equation modelling (PLS-SEM) to investigate the relationship between logistics outsourcing and performance of manufacturing SMEs. The study selected the PLS-SEM technique because of its predictive performance (Davcik 2014; Richter et al. 2016). Prior to SEM, the data were tested on suitability for exploratory factor analysis (EFA) using the Kaiser–Meyer–Olkin and Bartlett tests (Pallant 2010). EFA was performed, and it identified the latent constructs scales. The scales’ reliability was tested and found satisfactory in preparation for confirmatory factor analysis using Smartpls 3 (Ringle, Wende & Becker 2015).

The direct relationship between logistics outsourcing and firm performance was positive, as predicted, but not statistically significant. This finding provided an empirical validation of the results reported in the studies by Chatzoglou and Sarigiannidis (2009) and Hsiao et al. (2011) studies, which found no statistically significant relationship between logistics outsourcing and performance. Similarly, it partly supported the Cho et al.’s (2008) study, which found no statistically significant effect of logistics outsourcing on firm performance. This result suggests that logistics outsourcing benefits relating to performance of manufacturing SMEs are not straightforward as earlier expected, but might be dependent on factors other than just outsourcing, as also argued by Solakivi et al. (2011). The finding might also imply that the manufacturing SMEs in Nairobi outsourced their logistics for reasons other than just to improve their performance. Some of the reasons were identified as a lack of in-house capabilities, to reduce logistics costs, an aversion to risk and a desire to free management time to focus on core activities. Finally, logistics outsourcing could be used as a long-term restructuring strategy by the SMEs and not just to achieve financial and customer satisfaction goals, because the majority of the SMEs had <5 years’ experience with logistics outsourcing. Thus, it is likely that this strategy would yield performance-related benefits into the future (Zailani et al. 2017).

Logistics outsourcing performance as a mediating variable provides a deep understanding and a clear explanation of how logistics outsourcing influences firm performance (Baron & Kenny 1986; MacKinnon & Fairchild 2009). This study revealed that LO has a statistically significant positive indirect effect on FP through LOP as a mediating variable. As highlighted earlier, communication, trust, cooperation and innovation were identified as measures of logistics outsourcing performance. Thus, manufacturing SMEs practicing logistics outsourcing should focus on the quality of communication, building high trust levels, nurturing close cooperation with 3PLs and aiding 3PLs to enhance their innovative capabilities to improve performance. The positive indirect effect of logistics outsourcing on firm performance validates the results in the Hsiao et al.’s (2011) study. The indirect effect also validates partly the finding in the Solakivi, Töyli and Ojala (2013) study, which argued that upon logistics outsourcing, cooperation between SMEs and 3PLs results in customer satisfaction. Similarly, the result also supported the findings of the Lin, Pekkarinen and Ma (2015) study which highlighted close cooperation between manufacturer and 3PL as an important factor through which logistics outsourcing can positively influence enterprise performance. This is because cooperation promotes good communication, the building of trust and innovation, thus making it easier for the enterprise to understand how the services offered by a 3PL can influence its performance and make suggestions on any modifications required to achieve the expected results (Lin et al. 2015). This finding could imply further that the maintenance of good communication, high trust levels, close cooperation and 3PL innovation reduce contract management costs and other relationship costs, making it possible for the SMEs to reap the benefits of logistics outsourcing, such as reduced logistics operations and fixed costs.

Some of the manufacturing SMEs in this study had a formalised process of logistics outsourcing, although not designed to result in improved performance. In addition, it is established in this study that LO has an indirect influence on FP via LOP. To achieve tangible benefits to SMEs based on these findings, there is need to develop a logistics outsourcing model to guide the LO process among SMEs to achieve improved performance. Hitherto, SMEs do acquire expert LSPs’ capabilities to reduce costs and risks and to focus on core business in line with the RBV and TCE theories. Therefore, applying the findings of this study, a logistics outsourcing model is deduced along the critical path LO–LOP–LP–FP (Figure 3). The selected path has the highest statistically significant path coefficients, depicting a strong relationship (see Figure 3).

SMEs expecting to apply the proposed logistics outsourcing model are expected to do it through a rigorous process. Steps antecedent to the application of the proposed logistics model include the following: (1) a thorough information search to identify the reasons for outsourcing; (2) select the right activity to outsource to achieve objective in (1); (3) the right LSP that will better perform the activity selected in (2). Upon selecting an LSP, the proposed model is applied by (4), negotiating a favourable contract with the selected LSP. The contract should promote LOP attributes (i.e. quality communication, building high trust levels, nurturing a close cooperative relationship with the LSPs and enhancing innovative capabilities of the selected LSPs). High LOP would ensure that (5) LP goals such as cost reduction, risk reduction, meeting of delivery times, the provision of quality goods and the seamless flow of goods are achieved. The achievement of LP goals could lead to better (6) FP in terms of improved profitability, customer satisfaction, return on capital employed, ROA, and increased sales and market share. Thus, manufacturing SMEs could focus on the path LO–LOP–LP–FP when practicing LO to improve performance, as illustrated in Figure 2. This path was selected because it had the highest path coefficients, signifying the strongest significant relationship paths in the model. It requires that upon logistics outsourcing, SMEs should focus on logistics outsourcing performance to improve logistics performance. Improved logistics performance might result in improved firm performance (Zailani et al. 2017). The deduced process upon applying the logistics outsourcing model is summarised, as illustrated in Table 5.