Logistics Network Features Supporting Coordination by 3PL

1. Introduction

Logistics operators, or logistic service providers, are among the most common and critical components of logistics networks and supply chains. Third-party logistics (3PLs) have been the subject of academic discourse for quite some time, with the very concept of 3PL being deeply analyzed by numerous authors for years (e.g. Sheffi (1990), Selviaridis and Spring (2007) and Marasco (2008)) an there is a huge social demand for this kind of enterprises (Yan et al., 2003). The challenge of selecting the right logistics service provider dates back to older publications (e.g. McGinnis et al. (1995)), through slightly newer ones (e.g. Vaidyanathan (2005)), to contemporary works (e.g. Aguezzoul (2014)). Interestingly, some authors even debated in the 1990s whether such a form of enterprise in supply chains made sense at all (e.g. Berglund et al. (1999)). It’s evident that, despite the passage of time, the trend of analyzing entities like 3PL operators remains highly relevant, attracting interest not just among logistics industry practitioners but also scholars examining various facets of relations and flows in logistics networks and supply chains. This enduring and persistent trend results from the dynamic changes taking place in logistics and global supply chains. 3PL operators play a pivotal role in today’s business environment (Halldórsson and Skjøtt-Larsen, 2004), allowing companies to focus on their core activities by outsourcing various logistic processes. Such enterprises provide services like transportation, warehousing, inventory management and coordination of activities within the supply chain (Wang et al., 2019). This subsequently leads to more efficient operations, cost optimization and increased agility in response to fluctuating market conditions. Researchers continue to focus their investigations in this domain for several key reasons. Primarily, technological evolution (Fu et al., 2021) and shifting customer preferences (Premkumar et al., 2021) influence how firms collaborate within supply chains. Analyzing 3PL operators enables understanding of how they adapt to these changes and influence the entire logistic ecosystem. Additionally, the globalization of the economy results in more intricate supply networks (Guerrero et al., 2023), which in turn leads to increased management complexity, the need for rapid disruption response and the provision of adequate service levels to end customers. The ongoing interest in researching 3PL operators confirms that, in the face of advancing globalization, technological innovation and consumer preference shifts, analyzing relations and flows in logistics networks and supply chains remains a crucial research domain. The market key changes like e-commerce or omnichannels force the demand for new business models (Jintana et al., 2021), the same situation occurs in the 3PL.

Simultaneously, parallel to the analysis of 3PL logistics operators, scholars endeavor to deeply understand and describe the networks themselves, comprising various organizations. Studies focus on diverse aspects of organizational networks, encompassing knowledge transfer (Marchiori and Franco, 2020), data flow (Belik and Knudsen, 2023) and enhancing communication between businesses (Ioannidis et al., 2019). In examining organizational networks, scholars aim to capture how information and knowledge are relayed among network participants. They also consider how communication improvements can contribute to heightened efficacy within these networks. A frequently broached research topic is the mechanism of network coordination (Kumar et al., 2023), affecting the ability of individual enterprises to collectively manage operations network-wide. Today’s business environment poses myriad challenges for organizations, and organizational networks emerge as a means to address these challenges (Krejci, 2015). Recently, literature hints at linking logistics operators performing services within networks to provide coordination (e.g. Kmiecik (2022a, 2022b, 2023a) and Kramarz (2023)). Authors also specify which mechanisms are most crucial and how evaluation should proceed when choosing a logistics operator aiming to implement logistic coordination solutions (Kramarz and Kmiecik, 2024). However, there remains a gap in deeper analysis concerning the environment in which logistic operators are situated and the influence of this environment on the operator’s actions fundamental to coordination. Thus, in this paper, the author posits the following research question:

RQ.1: What characteristics of the logistics network support the utilization of logistic coordination mechanisms by a 3PL enterprise?

The studies conducted in this paper focus on revealing a literature gap that the answer to the above question will be able to fill. The general procedure related to the conducted research is presented in Fig. 1. It assumes the two main parts — literature review and empirical research. Empirical research focused on finding the research gap though bibliometric analysis using VOSviewer software and examination of chosen research papers connected with integration and coordination roles of 3PL, coordination mechanisms and 3PL activities in the reality of network and logistics coordination. Empirical research focuses on case of 10 logistics plants (LPs) (with 69 different distribution networks) located in two European countries. Conducted examination of 69 networks in which 3PL logistic operators’ function, and evaluating both the networks themselves and the operators serving these customers.

Conducted case study and calculated correlations, as a final result, allow to fully answer for created research question characteristics of the logistics network support the utilization of logistic coordination mechanisms by a 3PL enterprise.

2. Theoretical Background

2.1. Literature gap

The literature gap analysis was conducted in several steps. First and foremost, the SCOPUS database was searched, which is one of the most popular databases containing articles and other academic works. The SCOPUS database was searched for sources that included in the title, author keywords, publisher keywords or abstract at least one of the following phrases: network coordination, network governance; logistics coordination, logistics network, 3PL, logistic service providers, network features. The choice of following keyword for SCOPUS searcher based on author’s experience connected with the topic of logistics coordination. Author assumed that it is possible to use more synonymous terms, but the results wouldn’t be too different. There were 13,139 records in the SCOPUS database that matched the number of sources at the time of the analysis (i.e. August 2023). The first step taken was to narrow down the subject area to: business, management and accounting and social sciences, as these two disciplines were most thematically relevant to the topic discussed in this paper. This reduced the number of sources to 4,657. From these sources, only positions classified as: article (3,326 items), conference paper (720 items), book chapter (374 items) and book (56 items) were selected. This gave a total of 4,476 items that were analyzed in the context of the associations of keywords assigned to them by the authors and publishers. The association analysis was performed using the VOSviewer tool. The visualizations were created using the associations strength method, with layout settings as $\mathrm{\text{attraction}}=2$ and $\mathrm{\text{repulsion}}=0$, and clustering settings with $\mathrm{\text{resolution}}=1$ and minimal cluster $\mathrm{\text{size}}=1$, merging the small clusters.

As depicted in the constructed map, three distinct areas can be identified in the analyzed articles that are most strongly interconnected. The first area is logistics, which is associated with, among other things, the creation and management of networks. The second is supply chain management, which has a strong connection to outsourcing and 3PL. The third area pertains to network governance, which isn’t conventionally associated with logistics or the supply chain. Initially, the themes of publications related to network coordination were examined. In Figs. A.1–A.4 were illustrated the main connections of clusters related to particular keywords. It is closely aligning with clusters associated with network governance and sustainability, but also with decision-making, which is nearest to the cluster related to logistics. Conversely, the keyword associated with “decision making” is closely linked to logistics and network coordination itself. This suggests that network coordination is not typically considered from a logistics perspective. Intriguing connections are formed in studies related to network analysis. Such studies are directed both towards the logistics area and the area of network coordination. However, works related to network analysis and network coordination are not linked with logistics, and vice versa. The cluster associated with the outsourcing of logistic services and 3PL is strongly linked with supply chain management and logistics topics. However, there are no strong associations related to network coordination and network analysis itself. Associations related to network coordination, identified through keyword analysis from sources in the SCOPUS database, revealed that network coordination is not frequently thematically linked with concepts related to logistics, 3PL or supply chain management. Despite this, there are individual publications, also indexed in SCOPUS, that connect to this topic or revolve very closely around it (Table 1).

Coordination in the logistics networks, within the scope of this paper, will be considered from the perspective of logistic outsourcing companies engaged in coordinating activities related to flow management. Publications touching upon logistics or logistic outsourcing usually give limited attention to the mechanisms driving logistic networks. On the other hand, studies delving into network-related issues often overlook or only vaguely touch upon the aspect of managing flows. Currently, a trend is emerging that considers issues related to organizational networks and attempts to transfer mechanisms typically deepened by the activities of logistic companies providing services on behalf of other links in supply chains. The concept of coordination in the context of 3PL is shaping up, which assumes the role of coordinator to be taken on by the logistic service provider. As mentioned, this concept is still in its formative phase, and subsequent sections of this paper aim to shed light on the role of contemporary logistic service providers and to embed this role into the current concept. The goal will be to comprehensively highlight the research gap related to correlating certain characteristic features of networks with the ability of an operator to assume the coordinator’s function.

2.2. Integration and coordination role of 3PL

Currently, many scholarly articles continue to explore the issue of 3PL from the perspective of appropriate selection, choice or the creation of evaluation criteria for the decision-making process associated with engaging a logistics service provider. Such viewpoints are considered, among others, in works by Yayla et al. (2015), Aguezzoul and Pires (2016), Gürcan et al. (2016), Singh et al. (2018), Bianchini (2018), Qureshi (2022) and Kahraman et al. (2022). This approach may generate fundamental collaborative issues related to treating 3PL enterprises solely within contractual boundaries that operate with limited access to crucial information and typically without greater understanding with their clients. This risk was noted by Huo et al. (2015), and the voices of researchers suggesting that companies can gain more when their relationships with 3PLs are nurtured are growing stronger. For instance, Darko and Vlachos (2022) emphasize the necessity of fostering valuable relationships with 3PLs, demonstrating its significance based on multiple case studies. 3PLs play a crucial role in the integration and coordination processes within modern supply chains. Ever since the studies conducted by Mortensen and Lemoine (2008), the potential of 3PLs for effective integration with manufacturers has been recognized. Researchers highlight that 3PLs act as integrators of transport operations (Tyan et al., 2003) and transport–warehousing operations (Gürler et al., 2014), establishing consistent and seamless logistic processes. The evolution of the 3PL role as integrators is also evident in the fields of production and distribution. Studies by Jung et al. (2005), Fu (2014) and Noroozi et al. (2018) show that 3PLs can actively integrate processes in these areas, resulting in increased efficiency and flexibility of the supply chain. An especially important aspect is the influence of 3PL on customer satisfaction (Sheikh and Rana, 2011; Wu et al., 2023a). Through their actions, 3PLs contribute to creating positive customer experiences. Not only in operational areas but also in social and environmental matters, 3PLs play an active role. Research by Liu et al. (2020) indicates that 3PLs support ecological goals through sustainable logistics practices, while Hassanzadeh et al. (2022) highlight their role in promoting sustainable development.

As key participants in the supply chain, 3PLs not only act as integrators but also as orchestrators, which is evident in the works of Zacharia et al. (2011) and Mir et al. (2021). These researchers underscore the ability of 3PLs to coordinate and manage entire supply chains, acting as catalysts for efficiency. In terms of coordination, 3PLs play a significant role in reverse flows (Weraikat et al., 2016) and in the competitive retail market (Jiang et al., 2014). In some cases, 3PLs can even act as sub-coordinators (Jiang et al., 2019). By supporting the main supply chain participants, 3PLs create an additional layer of coordination, enhancing process efficiency. They are capable of forming coherent and organized mechanisms in high-complexity areas. Particularly vital is the role of 3PLs in information management and information flow within the supply chain. Studies by Pinna et al. (2010) and Carrus and Pinna (2011) confirm that 3PLs can effectively coordinate information processes, contributing to better synchronization of actions in the supply chain. The proposition of 3PLs utilizing coordination mechanisms, i.e. advanced tools enabling full coordination in networks and supply chains, is relatively rare (discussed in works like Multaharju and Hallikas (2015), Kmiecik (2022a, 2022b, 2023a), Kramarz and Kmiecik (2022) and Wu et al. (2023b)). However, this issue is of utmost importance if one truly wishes to perceive 3PLs as “full-fledged” coordinators of supply chains and networks, in line with management and business science concepts.

3PL enterprises are a significant part of supply chains (Qureshi, 2022; Shanker et al., 2022). While research on other entities, focusing on coordination aspects, is relatively extensive, research specifically on network coordination in the context of 3PL offers ample opportunities for new studies. For instance, researchers like Mogos et al. (2022) present the concept of operationalizing production network coordination, which involves coordination of production activities. They observe that such an approach can reduce disruptions in the supply chain. Conversely, Fritz and Hausen (2009) and Liao et al. (2010) noted that supplier network coordination can provide increased flexibility across the entire supply chain. However, the analysis of network coordination from a logistics operator perspective is typically either conducted at the local user-provider interface (discussed in Heide and John (1992), Nickerson et al. (2001), Griffith and Myers (2005), Zhao et al. (2006) and Wang et al. (2020)), or from the perspective of aligning 3PL activities and adapting these businesses to ensure a higher level of integration in networks (Wang et al., 2020). Some authors closely associate supply chain coordination with the operations of 3PL. Studies conducted by Lv et al. (2011) demonstrate the positive effect of enterprise intervention in supply chain operations regarding coordination, extending the research by Ouyang et al. (2007) analyzing lead time in supply chains. However, these works do not address the particular issues connected with the main elements of coordination.

Researchers occasionally attempt to describe specific coordination issues in networks where 3PL companies are of significant importance. The most commonly discussed mechanism, given the nature of 3PL operations providing contract logistics services, is the contractual mechanism. A cost-based contractual mechanism is based on mutual cost-sharing with fixed compensation contract as the primary network coordination mechanism. Occasionally, authors also discuss the application of modern technologies as a coordination mechanism in 3PL networks (Zhang et al., 2023). An approach that fully represents the network coordination concept, considering 3PL, and termed it as logistical coordination, includes mechanisms such as demand forecasting, inventory management, transportation planning and resource planning (Kmiecik, 2022a; Kramarz, 2023). Some of these mechanisms, such as inventory management (Kmiecik, 2022b) and transportation planning (Kmiecik, 2023a), have been empirically tested. In the work of Kramarz and Kmiecik (2024), logistical coordination mechanisms are combined with network coordination mechanisms (presented in Czakon (2009, 2018)), thereby offering a broad spectrum of mechanisms. This work also assessed the weight of each mechanism, enabling companies considering 3PL selection to evaluate them in terms of network coordination capabilities. The mentioned work lists the following mechanism elements and assigns them the respective weights: organizational integration (0.1344), forecasting flows in the network (0.1312), control structures and systems resulting from the management style (0.1216), network participants resources management from the logistics operator level (0.1148), trust (0.0780), demand management (0.0697), bureaucratic allocation of resources (0.0640), price (0.0560), formal relationships (0.0462), logistical information management from the logistics operator level (0.0410), bilateral collateral (0.0378), social standards (0.0260), significant information exchange (0.0247), transport organization and extraordinary transport (0.0246) and human resources and infrastructure management in the network (0.0246). Thus, a tool for evaluating operators and detailing the most crucial mechanisms and their forms was proposed. However, it was observed that a significant research gap exists regarding which network characteristics favor the implementation of this concept (Fig. 2).

Addressing the identified research gap will allow for the expansion of theories related to network coordination and logistical coordination by incorporating the characteristic features of networks that promote logistical coordination. To achieve this, a research procedure will be used, the results of which are presented in this paper.

2.3. Proposition of 3PL coordination elements assessing

The crucial step in the analysis is the evaluation of operations carried out in individual networks by 3PL. General information about the assessment of operations conducted by 3PL in specific networks is presented in Table 2. Importantly, for the analysis, due to the accessibility of research data which was conditioned by collecting reliable data over the last three months, four logistics coordination elements were selected for examination.

If forecasts are created within inventory management by 3PL, the errors and forecasting accuracy can be used for XYZ classification. Ex post errors in forecasting vary, but the most commonly mentioned ones (including in Satchell and Hwang (2001) and Ostertagova and Ostertag (2012)) for XYZ analysis and inventory management are mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE). In the case of conducted research, the MAPE was chosen. MAPE could be described in the following equation :

where $y_i$ is real value in i-period, $y_i^{\ast }$ is forecast value in i-period and n is the number of observations.

Forecasting errors generated in individual networks will be classified into five groups. Intervals for these classes will be established using a method based on quartiles (Bland, 2015; Luo et al., 2018). Appendix B presents generalized fragment of the R script used to create the mentioned class intervals. In the subsequent analysis of 3PL, the classification also relies on the same method of division into five classes, where the classes are constructed by calculating quartiles. The classes are then assessed on a scale from 1 to 5, where 1 always indicates a negative evaluation, and 5 indicates a positive one. In the case of the assessment of forecasting errors, the established classes are as follows: 1 — Very Bad Accuracy; 2 — Bad Accuracy; 3 — Average Accuracy; 4 — Good Accuracy; 5 — Very Good Accuracy. In the case of XYZ classification, selecting appropriate values that suggest an stock keeping unit (SKU) is placed in the right class is also an individual matter. Despite certain authors providing guidelines (such as Al-dulaime and Emar (2020) or Herlambang and Parung (2021), who suggest the following classes for specific MAPE error ranges: X for results less than 35%, Y for results less than 60% and Z for the rest, treatment of forecast errors as small or large depends on many factors related to business strategy, SKU type and the operational environment of the company. However, advanced forecasting systems nowadays offer the possibility of minimizing errors and increasing forecast accuracy. In the case of this study author used the common requirements in the particular networks to establish the proper ranges of MAPE to particular XYZ classes. When 3PL doesn’t provide demand forecasts to particular network then the XYZ classes where established based on coefficient of variation (CV) indicator. It is also a common approach to establish classes in XYZ (Suryaputri et al., 2022). The CV is calculated using the following formula (Brown, 1998) :

where S is standard deviation and $\bar{x}$ is average value.

CV describes the degree of data variability; therefore, in the context of inventory management, it provides information about the variability of demand for a particular product. In the literature, attempts have been made to categorize XYZ groups based on CV values. One of the classifications, proposed by Kaczorowska et al. (2019), has been presented in Table 3.

Certainly, it’s important to be aware that when determining CV values, the ranges should be established based on a good understanding of the assortment being managed in inventory management. Various factors influence these groups, and rigid categorization may not be suitable in this context. For example, SKUs on the warehouse shelves in the fast-moving consumer goods (FMCG) industry will be structured differently compared to industries where rotation and frequency of issues don’t reach such levels (e.g. construction industry). Another mentioned method of classification involves categorization using ex post errors generated for individual SKUs. The classification of the network will be based on the analysis of the share of Group X in the total assortment found in the warehouse managed by the operator. The classification in this case will also be based on the creation of five classes using quartiles in the following manner: 1 — Very High Variability; 2 — High Variability; 3 — Average Variability; 4 — Low Variability; 5 — Very Low Variability.

The analysis of inventory levels will be based on data from the warehouse management system (WMS) from the last three months. For each network, the average inventory level in the occupied pallet spaces in the warehouse will be provided, converted to SKU based on the last three months. The calculation will be conducted based on the following formula :

where st is stock level, $\overline{\text{pp}}$ is average number of busy locations (EUR1 size pallet places) based on last three months and sk is sum of SKU (based on last three months).

The level of inventory plays a key role in effective demand management, especially in the context of warehouses serviced by 3PL providers. Understanding its significance is crucial because it directly affects product availability (Song et al., 2021), operational costs and the overall efficiency of the supply chain and network (Perdana and Arief, 2021). It is often believed that maintaining the right level of stock in the warehouse affects product availability, reduces lead time, optimizes costs, improves decision-making processes and minimizes the impact of random demand fluctuations (Derhami et al., 2021; Afentoulis and Zikopoulos, 2021; Dhaigude and Mohan, 2023). The classification was based on the following scale and also relied on quartiles: 1 — Very High Stock; 2 — High Stock; 3 — Average Stock; 4 — Low Stock; 5 — Very Low Stock.

Delays in loading or shipping orders can have a significant impact on transportation organization in the entire supply chain perspective (Ali, 2019; Fałdziński et al., 2021; Kanike, 2023). These delays can have cascading consequences that permeate different stages of the delivery process, leading to disruptions, cost increases and reduced customer service quality. Delays caused by 3PLs can disrupt the entire delivery schedule (Lorenc and Kuźnar, 2021), affect transportation costs, impact customer satisfaction (Li et al., 2016) and reduce the flexibility (Zitzmann and Karl, 2018) of the supply chain or network. In the case studied, delays will be calculated for the last three months using WMS data. Each of the discussed factors was rated on a five-point scale, where: 1 — Very Big Delays; 2 — Big Delays; 3 — Average Delays; 4 — Small Delays; 5 — Very Small Delays. Delays were calculated in the following way :

where del is average delays in days caused by 3PL warehousing actions based on last three months, pl is planned shipping date, ac is actual date of shipping and n is number of observations.

Technological assessment in individual LPs and networks will be based on expert opinions, which are provided by high-level managers with extensive knowledge related to current technological developments and insights into operations within specific networks in LPs. Technological advancements and the ongoing evolution of automation and digitization are the primary drivers for changes in logistics (Gafert et al., 2021). While traditionally logistics service providers were not strongly linked with innovations or the use of cutting-edge technologies, the landscape has significantly shifted. Assessing the level of technology is frequently carried out using expert interviews (Choi et al., 2007; Lee and Kim, 2020; Yulherniwati and Ikhsan, 2020). This approach is often considered a good methodology or a component of a methodology for evaluating logistics providers (Huang and Zhao, 2012; Kwon et al., 2022). In the technological assessment, experts were asked: “The service recipient utilizes most of the implementable technologies that can enhance the functioning of information flow, which the logistics operator can offer to them.” Responses were constructed using a Likert scale (Albaum, 1997) and were correlated with the following values: 1 — Strongly Disagree; 2 — Disagree; 3 — Neutral; 4 — Agree; 5 — Strongly Agree.

Lastly, the evaluation of 3PL underwent a weighted assessment. All factors were evaluated based on the weights provided by Kramarz and Kmiecik (2024): Forecasting flows in the network — 0.1312; demand management — 0.0697; organization of transportation — 0.0246; logistical information management — 0.0462. These authors propose a tool for evaluating logistics operators in the context of their performance in terms of logistics coordination, applying the mentioned weights. It’s important to note that these weights do not sum up to 1, as the authors also list other factors related to logistics coordination, including aspects connected with network coordination, which are not considered in this paper.

3. Methods

3.1. Chosen LPs

The aforementioned analysis of 69 networks will be based on nine LPs located in two European Union (EU) countries (Table 4). Six LPs located in Poland and three in the Czech Republic will be analyzed. The total number of customers studied is 69, and the total number of products flowing through all the networks, based on the last three months, is 766,275 SKU.

The analysis will consist of four main steps: evaluation of the service provider (3PL), assessment of the service recipient, assessment of the level of cooperation within the network and calculation of correlations.

3.2. Procedure of service recipient’s assessment

The next component of the analysis concerns the service recipients. A general overview of these service recipients is presented in Table C.1. The majority of the service recipients were manufacturing companies, accounting for 65.2%. Other notable categories included retailers, wholesalers, manufacturer & wholesalers and manufacturer & retailers (Fig. 3).

The number of SKU was assessed similarly to the evaluation of most factors for the logistics operator, that is, by creating five intervals based on quartiles. The remaining factors were evaluated using the previously mentioned five-point Likert scale (Albaum, 1997). For the analysis, three factors were selected (Table 5), which were evaluated by managers collaborating with service recipients from the logistics operator’s side.

All the indicated factors shape the overall perspective on the service recipient and the challenges in servicing them. The emphasis on the significance of SKU diversity in logistical flows is highlighted by authors such as Chandra and Kumar (2001) and Jiang et al. (2020). On the other hand, warehousing susceptibility is identified as a critical factor in logistics service (Walaszczyk and Szymonik, 2020). Specific requirements of service recipients are also factors that condition the flows in networks and supply chains (Vakulich, 2021; Karcz and Ślusarczyk, 2021). An assessment of such a set of criteria will provide insights regarding the service recipients themselves. The next evaluation element is the collaboration assessment.

3.3. Procedure of collaboration level assessment

The factors were also assessed using the previously mentioned five-point Likert scale (Albaum, 1997), with the same assumptions as before. The evaluated items are presented in Table 6. Assessments were made arbitrarily based on the opinion obtained from a manager responsible for maintaining contacts with the service recipient designated by the logistics operator, and from information sourced from the individual responsible for contact with the 3PL enterprise on the service recipient’s side.

The issue of information exchange and flow is frequently addressed as a critical factor for achieving the right level of cooperation (e.g. in Macdonald (1992), Sagun et al. (2009) and Anthony Jr (2021)). The clarity of information is similarly discussed, as evidenced by studies such as Arciniegas et al. (2013). Some authors believe that coordination can be achieved without the need for exchanging strategic information (Zoghlami et al., 2016), which is a fluid topic since, in many situations, information that some businesses might consider strategic must be provided. For instance, in demand forecasting, retail companies should provide information on promotional campaigns, which they might deem strategic. This can lead to significant opportunism among companies. However, modern technologies like blockchain enable data encoding and leveraging the full potential of data, e.g. by the logistics operator’s forecasting systems without directly sharing confidential and strategic data. The use of blockchain technology in 3PL companies is the subject of many academic publications (e.g. Shostak et al. (2019), Zhang and Chen (2020), Varriale et al. (2023), Görçün et al. (2023), Zhang et al. (2023) and Zhang and Liu (2023)). Flexibility is also frequently cited as a desirable characteristic in supply chain operations, reflecting a high level of business cooperation (Wang and Wei, 2007; Angkiriwang et al., 2014; Scholten and Schilder, 2015). The last factor analyzed is the mutual satisfaction of network participants, as emphasized by Erik Eriksson (2010) and Blaževska-Stoilkovska et al. (2015); it’s also a crucial factor in nurturing cooperation in supply chains. Subsequently, a procedure was developed to measure the correlation between the measured evaluation results.

3.4. Procedure of correlations calculation

Correlations were calculated to identify relationships between individual criteria for the functioning of logistic networks and the mechanisms of logistic coordination, as well as the actual operations of logistic operators. R software was employed for the correlation analysis, using three widely used methods for calculating correlations: Pearson correlation, Spearman rank correlation and Kendall’s tau correlation (Chok, 2010). The calculation method for these correlations is presented in Table D.1. The analysis was conducted by examining correlations using the Pearson linear correlation coefficient, Spearman’s rank correlation and Kendall’s tau correlation between the specified factors, with a statistical significance level set for results with a p-$\mathrm{\text{value}}<0.05$. Adopting a p-value of this level is common in scientific articles (Genovese et al., 2006; Goodman, 2008). In statistics, the p-value is a measure that determines the statistical significance of test results (Andrade, 2019). When calculating correlations, the p-value indicates whether the detected correlation between variables is statistically significant or could be a result of chance. A p-$\mathrm{\text{value}}<0.05$ means that there is less than a 5% chance that the observed correlation was obtained randomly in the sample. In other words, if the p-value is less than 0.05, the result can be considered statistically significant, suggesting a genuine relationship between the variables being studied. If the p-value is greater than or equal to 0.05, it means there is not enough evidence to reject the null hypothesis, which states that there is no correlation between variables. This does not automatically mean that there is no correlation, but it suggests that we do not have enough certainty to conclude that one exists. All steps of the analysis procedure were implemented on the data collected by the author.

3.5. Data description

The data for the analysis were collected by the author of the publication. The primary source of data was the WMS, or in cases where a forecasting tool was operational within the network, data was sourced from both the WMS and the forecasting tool. The general characteristics of the data are presented in Table 7.

The data used in this paper have been deliberately anonymized, and any information regarding the know-how or specifics of the company operations that aren’t publicly known have been omitted. The data was statistically processed by the author. For simpler calculations and chart visualizations, the author utilized spreadsheet software. Google Forms was used to gather responses from the selected experts. Statistical calculations related to determining intervals for various data sets, as well as the Pearson, Spearman and Kendall correlation, were carried out using custom scripts written by the author in RStudio. Meanwhile, Statistica was used for calculating the Pearson correlation between specific factors under evaluation and for constructing scatterplots.

4. Results

4.1. Assessment of 3PL operations

The first element studied was the forecasting flows in the network. As mentioned earlier, this aspect was examined in two ways, depending on whether the logistics operator in a given network had a forecasting tool and had data regarding the forecast verification accumulated over a minimum of three months. In the analyzed cases, 22 networks (31.88%) were serviced by an operator who had a forecasting tool (Fig. 4).

A relatively small percentage of networks with an operator possessing a forecasting tool is due to the fact that such solutions are not yet widely adopted in 3PL enterprises (Kmiecik, 2021, 2022a; 2023a; Kmiecik and Wolny, 2022). The MAPE analysis was conducted by calculating this ex post forecast error for the last three months. Ranges for individual results were created based on quartiles, and their span and frequency are presented in Table 8.

Overall, it can be stated that the forecasts were made with high accuracy. The majority of the networks fell within the range characterized by good accuracy (i.e. with a forecasting error between 1.08% and 6.49%). Importantly, forecasts were generated collectively for warehouse releases in a unit associated with predicting the number of pallets to be issued. In the good accuracy group, there were nine networks, constituting 40.90% of all networks in which a forecasting tool operated. Figure 5 shows a detailed breakdown of MAPE values for individual networks.

As can be seen in the figure, in the vast majority of cases, networks achieved similar results related to the magnitude of errors generated by the forecasting tool. Networks that generated significant errors are networks 68, 09 and 07. Interestingly, the networks with the worst results (09 and 07) are situations where the operator serves a manufacturer & wholesaler. For networks where no forecasting tool is implemented, the XYZ classification based on the analysis of assortment variability was used for evaluation. Lower assortment variability provides a perspective for better functioning of forecasting tools (Hall and Tacon, 2010; Errasti et al., 2010; Baryshnikova et al., 2021). The network classification using the XYZ analysis is presented in Table 9.

The largest number of the examined networks was characterized by high or very high variability in terms of SKU issues (a total of 28 networks, i.e. 59.57% of networks examined using XYZ analysis). However, a significant proportion of networks also showed low or very low variability (a total of 12 networks, i.e. 25.53% of networks examined using XYZ analysis). In these networks, there is a very high share of products from the X group in the total number of SKUs, which is a very good indicator for the effective implementation of forecasting tools in the future. The aggregate assessment for the forecasting flows in the network factor is presented in Table 10.

The evaluation scale was assigned based on the rating given to the networks based on the MAPE value or variability in issues examined using XYZ analysis. A very good rating (5) was received by 10.14% of the networks, a good rating (4) by 23.19% of the networks, an average rating (3) by 14.49% of the networks, a poor rating (2) by 30.43% of the networks and a very poor rating (1) by 21.74% of the networks. As for the evaluation of the factor in relation to the entire logistic plant, the best average rating was received by LP_01 and LP_07 (i.e. an average of 3.50), and the worst rating by LP_08 (i.e. an average of 1.00). The average rating for individual types of service recipients is shown in Fig. 6.

Upon analyzing the two extreme results, it can be concluded that demand forecasting is better organized in shorter supply chains characterized by fewer intermediaries (a manufacturer who also sells retail). In contrast, the worst scenarios are observed in elongated networks where the manufacturer acts as a wholesaler and lacks direct sales information. The next factor under examination is demand management. Figure 7 displays the average number of pallet spaces allocated for storing a single SKU over the past three months.

Based on the calculations conducted, in most cases, the analyzed value does not exceed 10 pallets per SKU. Only a few networks exceed this threshold, with one (N_15) doing so significantly. The intervals for assessment in this context were also developed using quartiles (Table 11).

Based on the data, the lowest inventory levels were achieved by LP_05 (0.87), while the highest were observed in LP_02 (5.66). The results concerning the evaluated factor in relation to different types of customers are also intriguingly distributed (Fig. 8). The best average result related to the ratings received by networks from 1 to 5 was again achieved by the manufacturer & retailer — this confirms the author’s earlier assumptions that shorter networks with fewer intermediaries are more convenient for coordination by the logistics operator. This type of customer also obtained the lowest value in terms of the average number of pallets maintained per SKU, which was 0.80.

It’s evident that the results obtained from the assessment of this factor are better than those from the previous evaluation. The next factor under review is the organization of transportation. The evaluation of transportation organization was conducted using the assessment of delays caused by the logistics operator, calculated in days based on data from the last three months (Fig. 9).

As revealed by the analysis, several networks are characterized by relatively significant delays caused by the logistics operator, which significantly affects their ability to manage transportation within the network and results in delays in order fulfillment. Table 12 illustrates the allocation of networks to specific rating ranges for the analyzed factor.

The majority of networks exhibited very minor delays (28 networks, accounting for 40.58% of all networks) and average delays (27 networks, representing 39.13% of all networks). Three networks experienced delays exceeding one day. Among the logistic plants, LP_07 displayed the highest average delays (0.55 days), while LP_03 and LP_09 exhibited the lowest delays (0.02 days). Figure 10 presents the averaged results of the analysis of the transportation organization factor in relation to various types of service recipients.

The lowest rating was received by the retailer (2.89), while the highest was given to the wholesaler (4.00). This indicates the difficulty in organizing transportation related to the movement of goods to the retail sector (second to last place was taken by the manufacturer & retailer, confirming these assumptions). This may be due to the relatively wide range of products that require various actions before loading and the challenges of organizing transportation to retail points located in city centers and other hard-to-reach areas.

The next factor examined was logistical information management. This factor was analyzed based on the responses of experts to the posed question (Table 13).

When it comes to assessing the factor related to logistics information management, it is generally provided at a high level in most networks. According to the author, this is due to the fact that one of the primary tasks of operators is to take care of Value-added services (VAS), which is currently ensured through digitization and a focus on optimizing information exchange. The assessed factor performed best for LP_06 (average score of 5.00) and worst for LP_02 (average score of 3.11). The results were also related to types of service recipients (Fig. 11).

As seen in this case as well, manufacturer & retailer performed the best, and once again, manufacturer & wholesaler received the lowest average score, indicating that it is one of the longer networks with multiple intermediaries. In the next stage, the final result was analyzed (Fig. 12). For the final result calculation for each network, a weighted average was used, as proposed by Kramarz and Kmiecik (2024), a methodology the author mentioned in the previous chapter.

The network N_17 achieved the best result in terms of the level of logistics coordination implementation based on the selected factors under investigation (1.3123), while network N_37 performed the worst (0.4625). The arithmetic mean of the network assessment results was 0.8640, and the median, which is the middle value, was 0.8490. If the arithmetic mean is close to the median, it indicates that the data does not exhibit significant skewness. Skewness is a measure that describes whether data is shifted in one direction (positive skewness) or the other (negative skewness) relative to the arithmetic mean. If the mean is close to the median, it means that the data is distributed in a manner close to symmetrical, with no significant shift in one direction. In practice, this suggests that the data distribution may be less skewed and less susceptible to the influence of outliers. However, the mean is not always identical to the median, even for symmetrically distributed data, because both measures capture different aspects of the data distribution. The mean represents the overall central tendency, while the median represents the central position. Furthermore, the standard deviation was 0.2071. The LP with the highest average rating is LP_03 (0.985), but similar results were also obtained by LP_01 (0.950), as well as LP_04 and LP_07 (both received an average score of 0.942). LP_08 received the lowest result (0.669). Figure 13 presents the average results for different types of service recipients.

The manufacturer & retailer received the highest score (1.05), which aligns with the previous findings from earlier sections of the analysis presented in this paper. Conversely, the manufacturer & wholesaler performed the worst (0.67). The other types of service recipients received corresponding scores: wholesaler: 0.89, manufacturer: 0.88, retailer: 0.83. It can be concluded that they were very close in their evaluations. In the next step of the analysis, an assessment of the service recipients was conducted.

4.2. Assessment of service recipients

The evaluation of service recipients was carried out through expert assessments, where experts provided responses to specific questions on a five-point scale. The first factor assessed was the “diversity of SKU” (Table 14).

The majority of responses for this question fell into the three-point scale. The diversity of SKU indicates the challenges faced by the operator in managing warehouse and transportation processes. In other areas of the scale, responses were numerically close to each other. The next factor examined was “warehousing susceptibility” (Table 15), where responses were generally evenly distributed. Only two networks received a rating of 5. This may indicate high demands placed by service recipients on service providers in contemporary markets.

In Table 16, the assessment of “special requirements in the area of storage and transportation” is presented. In this case, the distribution of responses was similar to the previous factor, with the difference that a larger number of networks received the highest rating (i.e. 5).

In Fig. E.1, the cumulative rating of networks covering the total points corresponding to the assigned scale is shown. Analyzing the figure, it can be observed that the assessment of the neural network labeled as N_17 stands out compared to other networks. Although its cumulative rating is significantly lower than that of other networks, it is worth noting that this is not due to a single specific aspect of the assessment, but results from averaging the results from various rating categories. Network N_17 received the maximum possible score of 4.00. It is also worth noting that two other networks, namely N_19 and N_61, received the minimum rating, which was 1.68. Therefore, the largest difference between the highest and lowest rated networks is quite significant, which may be due to significant differences in their performance or quality. The median of the results obtained by different networks was 2.68, indicating that most networks achieved results close to this value. The average result was 2.69, which is a value close to the median, suggesting that the data do not contain significant deviations from the norm. In addition, the standard deviation was 0.58, indicating a small variation in network results around the mean. This means that most networks were close to the average rating, suggesting some stability in their results. In summary, based on the analysis of the figure, it can be concluded that network N_17 stands out as the best, while N_19 and N_61 received the lowest ratings. The differences between networks are relatively large, but most networks are clustered around the average with little variability. As for the assessment of LPs, the best result for averaged scores was achieved by LP_06 (3.33) and the worst was LP_08 (2.17). Here, too, it can be observed that the results obtained were similar, with the following networks receiving corresponding scores: LP_01 (2.58), LP_02 (2.66), LP_03 (3.25), LP_04 (2.91), LP_05 (2.33), LP_07 (3.11), LP_09 (2.67). Figure 14 shows an additional averaged rating for each type of service recipient.

As per the analysis, in this case, the wholesaler received the highest rating, followed closely by the manufacturer & retailer and the retailer. While the top result is slightly different from previous analyses, the worst-performing network remains the one that is the longest, i.e. the manufacturer & wholesaler. In the subsequent sections, the analysis delves into assessing the level of collaboration in each network.

4.3. Assessment of collaboration level

The first assessed element was information flow (Table 17). Experts evaluated the level of cooperation in this area in a fairly even manner, with most of them avoiding the highest and lowest ratings.

Clarity of information was assessed in a similar manner (Table 18), with the exclusion of extreme ratings, both positive and negative.

The assessment of the flexibility factor (Table 19) and participant satisfaction (Table 20) appeared more diversified. In the case of flexibility, the largest proportion of networks received a score of 4 (43.48%), while for participant satisfaction, the most common responses were equivalent to a score of 5 (37.68%) and 4 (34.78%). This may indicate good relationships between service providers and recipients, which are a positive indicator for implementing various changes and improvements in network operations.

Figure E.2 presents cumulative values corresponding to assessments for the specific factors broken down by individual networks. The maximum score, both cumulatively and on average, was awarded to two networks, namely N_17 and N_54, where both achieved perfect scores of 5.00. It is worth noting that Network N_17 once again stands out as one of the best in this comparison, which may indicate its exceptional quality or effectiveness. These networks occupy top positions in the results, emphasizing their outstanding achievements. On the other hand, the minimum values were received by two other networks, namely N_07 and N_61, where the average score was only 1.25. This means that these networks received lower scores. The results in this case show significant differences between the best and worst ratings, suggesting significant variations in the performance of different networks. The average score for all 69 networks is 3.25, indicating an overall above-average level of results. The median is 3.50, suggesting a balance between scores above and below this value. However, with a standard deviation of 0.94, the results are much more varied than in the previous assessment, indicating greater variability in the results of different networks. As for LPs, this comparison is also diverse. LP_04 achieved the highest average score of 3.94, while LP_08 and LP_09 obtained the lowest scores, at 2.50. This suggests significant differences between different LPs. The analysis of results shows significant variation in the results of different networks and LPs, which may be due to differences in quality, efficiency or other relevant factors. Networks N_17 and N_54 stand out as some of the best, while N_07 and N_61 received the lowest scores. The average values for all networks and plants indicate an overall level of results, but the standard deviation highlights significant variability in the results. In the case of the average assessment broken down by service recipient types (Fig. 15), a significant advantage of manufacturer & retailer in terms of cooperation level assessment can be observed. This confirms the author’s earlier assumptions regarding a higher level of cooperation between operators and service recipients in networks with fewer intermediaries.

The last part of the analysis consists of comparing the results obtained in the assessment of the 3PL logistics operator, the assessment of the service recipient and the assessment of the level of cooperation. This comparison will be based on an attempt to find correlations between individual factors.

4.4. Correlations

In the first step of the analysis, the focus was on evaluating the correlation between the assessment of the 3PL logistics operator and the assessment of service recipients, as well as the assessment of the 3PL and the assessment of the level of cooperation (Table 21). The assessment was based on three correlation indicators, and calculations were performed using the cor.test() function available in RStudio. Additionally, this function also provided information related to p-values, allowing each result to be checked for statistical significance.

In the p-value calculations, it’s worth noting the notation “E”. This is exponential notation, used to represent very small or very large numbers in a more concise and readable way. In this case, “E” represents the power of 10, which is used to shift the decimal point to the appropriate position (Maor, 2009). For example, the number 2.84E−13 means the number 2.84 multiplied by $10^{(-13)}$ (which is approximately 0.000000000000284). In other words, it’s a very small number close to zero, expressed in scientific notation to work with very small numbers more conveniently. Since each p-value in the results is represented with such notation, it can be concluded that they are very close to zero, and therefore statistically significant. The first analysis, which focused on the correlation between the assessment of 3PL service providers in terms of logistics coordination and the assessment of service recipients in networks, provided valuable insights into the relationships between these variables. Pearson correlation coefficient values of 0.7425 and Spearman values of 0.7350 indicate a moderate positive correlation. This means that as the evaluation of 3PL logistics providers in terms of logistics coordination increases, service recipients in networks tend to evaluate the services of providers more positively. In other words, organizations that invest in improving logistics coordination within their supply chains are more likely to satisfy their customers. The Kendall coefficient value of 0.5920, although slightly weaker, confirms this positive relationship. This is important from a practical perspective as it suggests that focusing on improving logistics coordination can contribute to increased competitiveness by enhancing positive customer perceptions. The second analysis, regarding the correlation between the assessment of 3PL logistics providers in terms of logistics coordination and the assessment of the level of cooperation in business networks, provided even more compelling evidence. Pearson correlation coefficient value of 0.9335 and Spearman value of 0.9394 indicate a very strong positive correlation. The Kendall coefficient value of 0.8373 further confirms this clear relationship. This means that the assessment of 3PL logistics providers in terms of logistics coordination has a significant impact on the level of cooperation in business networks. In practical terms, this suggests that organizations that receive higher ratings in logistics coordination from 3PL providers tend to exhibit higher levels of cooperation and collaboration in their business networks. This is crucial information for companies because it implies that improving logistics coordination can yield benefits not only in terms of customer satisfaction but also in terms of interorganizational cooperation efficiency. The study also examined the correlations between the final assessment of 3PL logistics providers related to logistics coordination and various elements assessed in terms of service recipients and the level of cooperation. The results were visually presented using scatter bubble charts created in the Statistica software (Figs. F.1–F.7). It’s important to note that in this case, the author focused only on Pearson correlation, and the p-value is rounded to four decimal places (calculations show that in this case, each correlation is also statistically significant). There is a slight correlation between the decrease in the diversity of SKU and the assessment of 3PL in terms of logistics coordination. This suggests that a decrease in the diversity of SKU does not significantly impact logistics coordination. A stronger correlation is observed between warehousing susceptibility and the assessment of 3PL. This suggests that logistics coordination is easier to implement when products are more susceptible to warehousing. A similar level of correlation is found between special customer requirements and logistics coordination. It is easier to implement logistics coordination in networks that operate under standard conditions. A higher level of correlation is observed between information flow and the assessment of 3PL. It is easier to implement logistics coordination in networks that facilitate the free flow of information. A slightly lower but still high level of correlation exists between information clarity and logistics coordination. Clarity in conveying information is important for ensuring quick responses and effective actions in logistics coordination. Flexibility in operations proved to be the most strongly correlated element with the assessment of 3PL in terms of logistics coordination. The network’s ability to react quickly and adapt significantly enhances the opportunities provided by logistics coordination. Participant satisfaction also exhibits correlations, although slightly smaller than in the previous factor. It can be concluded that the level of satisfaction on both sides of the cooperation has a significant impact on logistics coordination. Further discussions related to the obtained results have been expanded in the discussion section.

5. Discussion

5.1. Which network features are supporting the logistics coordination?

Shorter and more direct supply chains are a crucial element of effective operations and logistics management in today’s businesses. Their defining characteristic is the reduction in the number of intermediaries and participants in the process of delivering products or services from the producer to the consumer. In shortened supply chains, there are significantly fewer elements to oversee. This means that managers and employees responsible for managing them can focus on more critical aspects such as process optimization, quality monitoring or cost control. A reduced number of variables makes it easier to make faster and more precise decisions. Shorter and more direct supply chains promote better collaboration among all participants in the process. All parties involved are closer to each other, facilitating communication, information sharing and problem-solving. The absence of unnecessary intermediaries eliminates potential communication barriers. In the dynamic business environment, the ability to respond quickly to changes is crucial. Shorter and more direct supply chains allow for a more flexible and adaptive approach. Companies can more easily adapt to changing market conditions, consumer trends and customer needs. The reduction in the number of steps in the supply chain often translates into greater efficiency and resource savings. Simpler and more direct processes can lead to reduced losses, fewer delays and lower costs associated with warehousing, transportation and inventory management. Such advantages have been repeatedly emphasized in research (e.g. Miller and Miller (2002), Insanic and Gadde (2014) and Rakyta et al. (2016)). This was also confirmed in the research conducted in this paper. The networks in which the service provider served a manufacturer & retailer type of service recipient achieved the best results in the evaluation. It can be concluded that the structure and configuration of the network in which the service recipient operates are crucial in logistics coordination. The more intermediaries involved, the more challenging the functions related to the implementation and management of logistics coordination are. Interestingly, the research revealed that this directness of the network does not have such a significant impact on the evaluation of the service recipient itself, so even a poorer relationship between the service provider and the service recipient is mitigated by the smaller number of intermediaries in the network.

The characteristics most correlated with logistics coordination regarding the Pearson correlation coefficient are: flexibility (0.89), information flow (0.87), clarity of information (0.85), participants’ satisfaction (0.75), special requirements in the area of storage and transportation (0.74), warehousing susceptibility (0.73) and diversity of SKU (0.43). The results themselves are interesting because in the context of logistics coordination, information exchange is not confirmed as the most important factor. Information sharing is considered a critical factor in the final outcome of logistics networks by researchers such as Liu and Kumar (2003) and Liu et al. (2015). Another factor that surprisingly showed low correlation was special customer requirements. Customer requirements are considered one of the most critical factors in the operation of networks (Large, 2007; Large et al., 2011; Chu et al., 2019). However, according to the author, 3PL companies have already learned to meet even the most advanced customer requirements in logistics as part of their standard work based on their knowledge and skills. A factor that also has significant importance, as emphasized in Khan and Bosgraaf (2009), Welsman (2010) and Frankin and Johannesson (2013), is diversity of SKU. However, in the conducted research, it was the factor with the lowest correlation with logistics coordination. This may indicate that 3PL companies can easily handle a small or highly diversified assortment within the scope of logistics coordination, and it does not have a significant impact on their operations.

5.2. Why logistics coordination could be so important?

Logistics coordination is important for several key reasons that arise from the described mechanisms, such as demand forecasting, inventory management, transportation planning and resource planning. Through demand forecasting and inventory management, logistics coordination allows companies to maintain appropriate inventory levels, which, in turn, translates into product availability for customers. This ensures that customers can rely on products being available when needed. Logistics coordination helps reduce costs associated with inventory management and transportation. Optimizing inventory helps avoid overstocking or shortages, which can lead to lower operational costs. Transportation planning allows for route optimization and efficient use of available transport resources, which also results in lower transportation costs. One of the crucial issues is the proper select of 3PL provider (Yan et al., 2003), so the chosen 3PL could impacted the results of logistics coordination. By optimizing logistical processes, a company can gain a competitive advantage in the market. Faster delivery, better product availability and lower operating costs can attract customers and help maintain their loyalty. Logistics coordination can also be closely related to other management areas such as supply chain management and network management. 3PLs are inclined toward inter-organizational learning (Min et al., 2018). As a result, a company can achieve full integration and synchronization of various processes along with other companies in the network, leading to better control over operations and efficiency. 3PLs play a significant role in supply chain integration (Niemann et al., 2018), so the author believes that they could also play a significant role in network coordination. Perhaps logistics coordination would also be able to solve some of the problems related to network coordination. For example, the network coordination paradoxes associated with market, social and hierarchical mechanisms, as presented by Czakon (2009, 2018), would require further research. The findings of this study gain further significance when viewed in the context of recent global events impacting logistics providers. Notably, the situations at the Panama and Suez Canals have highlighted the vulnerabilities in international shipping routes, underscoring the importance of effective logistics coordination. These events, along with the evolving geopolitical landscape, have drastically altered the logistics environment, making the need for efficient and flexible supply chains more critical than ever. The study’s emphasis on shorter, more direct supply chains aligns well with these challenges, offering a framework for mitigating risks associated with global disruptions. The research outcomes offer several actionable insights for logistics managers. First, the importance of developing supply chains that are not only efficient but also resilient to external shocks cannot be overstated. Managers should prioritize simplifying their supply chains, reducing the number of intermediaries and fostering closer relationships with key stakeholders. In author’s opinion, this approach not only enhances efficiency but also provides a buffer against global disruptions like those seen in the Panama and Suez Canal incidents. Additionally, the study’s findings on the importance of flexibility, information flow and clarity in logistics coordination should guide managers in refining their operational strategies. Emphasizing these aspects can lead to more responsive and adaptive supply chain networks. In light of recent global events, logistics managers should also consider diversifying their transportation routes and methods, to avoid over-reliance on any single channel or geography.

5.3. Managerial implications

The study’s emphasis on shorter, more direct supply chains presents a compelling argument for logistics managers to reevaluate and potentially restructure their existing networks. The reduction of intermediaries not only streamlines operations but also reduces complexity, leading to improved oversight and decision-making capabilities. Managers should consider these aspects in their strategic planning, focusing on creating supply chains that are both agile and resilient to unexpected disruptions. The identified network features such as flexibility, clarity in information flow and participants’ satisfaction are crucial for logistics coordination. Managers should prioritize these aspects in their operational strategies. Implementing systems and processes that enhance the flow and clarity of information can lead to more effective coordination, while also boosting satisfaction among all participants in the supply chain. Despite the low correlation found between special customer requirements and logistics coordination, this does not diminish their importance in network operations. Managers should continue to develop competencies and adapt strategies to meet diverse customer needs, ensuring that their services remain relevant and competitive. Addressing diversity of SKU showed the lowest correlation with logistics coordination, managers should not overlook the management of a varied assortment of products. Efficient handling of diverse SKUs can be a competitive advantage, allowing companies to cater to a broader range of customer needs without compromising the efficiency of logistics coordination. Investment in 3PL and beyond: The significant role of 3PL in supply chain integration suggests that further investment in these services could be beneficial. Managers should explore the potential of advanced logistics service providers, such as 4PL or 5PL, to enhance coordination across the entire supply chain. The evolution from logistics coordination to choreography represents an exciting frontier for logistics management. Managers should consider how to transform their current coordination strategies into more integrated, synchronized choreographies, potentially leading to enhanced efficiency and performance in the supply chain.

5.4. Main limitation and future research

Research in logistics coordination and network coordination can open up interesting perspectives for the future. There are many areas that would be worth exploring to better understand these issues and their impact on supply chains and logistics organizations. Research into paradoxes in network coordination can be expanded to their application in logistics coordination. It could be investigated whether the same or similar paradoxes occur in logistics coordination and what their implications are for the efficiency of supply chains. Research can also focus on analyzing how logistics coordination affects entire supply chains. This could include assessing the effectiveness, flexibility and adaptability of the supply chain to changing market conditions. Researchers can examine whether 3PL service providers are adequately equipped to implement logistics coordination concepts that involve the entire supply chain. This could lead to conclusions about the potential need for more advanced forms of logistics service providers (e.g. 4PL or 5PL) or ways to improve them. Some authors point out that coordination alone is not enough, and the natural next step is to move from coordination to choreography (Grange et al., 2020). Future research could, therefore, be based on an analysis of the possibilities of evolving the concept of logistics coordination into a more advanced form. Additionally, this paper does not consider all factors related to network and logistics coordination mechanisms presented in Kramarz and Kmiecik (2024). Future research could fill this gap as well. Building on this study, future research could explore the impact of specific global events, like the Panama and Suez Canal crises, COVID-19 pandemic or geopolitical situation, on logistics coordination. Investigating how logistics networks have adapted in real time to these disruptions could provide valuable insights into the efficacy of different coordination strategies. Additionally, research into how geopolitical shifts are influencing global supply chains would offer further understanding of the dynamic interplay between global events and logistics coordination. The smart idea of connecting the resources of the different 3PL, presented by Karia et al. (2015) could be also useful in the context of logistics coordination and could be examined in the further research.

6. Conclusions

The study presented herein offers an important contribution towards advancing theoretical knowledge on the roles and dynamics of 3PL operators within logistics networks, including insights into network coordination mechanisms. The authors have embarked on a comprehensive examination of the operational environment for logistics operators, highlighting its impact on their decision-making processes and activities within the framework of logistical coordination. Central to this paper is an exploration aimed at understanding the operational milieu of 3PL logistics operators, with a focus on how environmental factors influence their coordination capabilities. Such an approach enhances our understanding of the determinants that influence the success of coordination initiatives. The researchers delve into the significance of logistics networks in facilitating logistical coordination, investigating the attributes and structures of these networks that contribute to effective coordination among stakeholders. Drawing from a study spanning 69 networks in two EU countries, the research provides a solid basis for making informed conclusions, covering a wide range of contexts and market scenarios. The findings of this research stand to benefit both theoretical experts interested in the evolution of logistical and network coordination theories and logistics practitioners. The insights gained contribute to a deeper understanding of the elements that impact coordination success across various scenarios, highlighting the traits of logistics networks that support successful collaboration among supply chain actors. This knowledge is instrumental in guiding the formulation of enhanced logistical management strategies and practices within organizations, as well as improving operations for 3PL logistics operators.

ORCID

Mariusz Kmiecik  https://orcid.org/0000-0003-2015-1132