Abstract
This meta-analysis examines the supply chain alignment (SCAl) and performance relationship and the influence of some moderating variables. Twenty-nine articles (30 correlations) were selected based on a systematic literature review, covering a total set of 6,658 companies. A significant, positive, and reasonably strong correlation was found between SCAl and performance, indicating a medium to large correlation. However, the observed heterogeneity showed substantial between-study variations. Moderator analyses revealed higher correlations in developing vs developed economic regions, overall performance vs specific performance measures, CB-SEM vs PLS-SEM methods, and intra-organizational vs inter-organizational performance measures, and a very similar correlation between specific vs overall SCAl constructs. Despite SCAl having a significant positive correlation with performance in all cases, its magnitude basically depends on methodological moderators and other sources of variation not considered in this research. These results highlight the importance of enhancing SCAl and offer researchers a detailed examination and better understanding of the SCAl-performance link that underpins future research.
Introduction
Today’s dynamic and complex global marketplace accentuates the importance of effective supply chain (SC) management (Garcia-Buendia et al., 2023). As businesses endeavor to remain competitive and responsive to customer demands, a key factor that often determines success and leads to better business results is supply chain alignment (SCAl) (Fawcett & Magnan, 2002; Reklitis et al., 2021; Simatupang & Sridharan, 2002). It is vitally important for organizations striving to respond to dynamic market demands and gain a competitive advantage (Christopher & Towill, 2000) to ensure that all their SC members—from suppliers to manufacturers, distributors, retailers, and customers—synchronize and coordinate their activities, objectives, and strategies (Mentzer et al., 2001).
SCAl is a core capability that contributes to enhancing operational efficiency, minimizing disruptions, optimizing costs, and ultimately, improving customer satisfaction (L. M. Douglas et al., 1998). It is a key to success and sustainability (H. L. Lee, 2004, 2021) that enables firms to attain a long-term competitive advantage by boosting their performance and, consequently, their business. Despite this, research on this topic is scarce, fragmented, largely theoretical (Marin-Garcia et al., 2018; Wong et al., 2012), and faces many challenges (Arana-Solares Iván et al., 2012; Skipworth et al., 2015).
SCAl effectively and efficiently coordinates the resources and skills within a firm and between its SC members (Ogulin, 2014) to achieve operational and commercial benefits (Attia, 2016) through superior performance. Conversely, a lack of alignment can undermine many SC practices (H. L. Lee, 2004). It is a dynamic capability that can achieve a competitive advantage by addressing variable customer demand, economies, and markets (Whitten et al., 2012). However, it demands complex resources (Machuca et al., 2021).
Essentially, SCAl is an enabler that synchronizes material, information, and financial flows throughout the SC (Christopher, 2016), requiring not only the integration of a company’s internal functions but also collaboration and cooperation with external partners. It also improves visibility and information sharing across the SC, which, in turn, improves decision-making capabilities and reduces uncertainty (Mentzer et al., 2008). To guarantee SCAl, all SC functions should be focused on high-level organizational priorities (Mitchell & Kovach, 2016).
Although most of the limited research has found a positive SCAl-performance relationship (e.g., W. Ahmed, 2021; Keating, 2022), not all studies confirm this (e.g., Huo et al., 2021; Skipworth et al., 2015). Literature reviews and meta-analyses that focus on and evaluate the impact of SCAl on performance are also scarce. Wong et al. (2012) reviewed the corresponding literature to identify the key factors that enable SCAl. Feizabadi, Maloni, & Gligor, et al. (2019) focused on four SC journals to develop the antecedents and consequences of SC agility, alignment, and adaptability; the so-called Triple-A. Other literature reviews (Hochrein et al., 2017; Ogulin, 2014; Vasconcellos et al., 2017; Wong et al., 2012) focused on alignment conceptually or did not analyze its impact on performance. Of only three articles that have performed meta-analyses of SCAl, two (Cao et al., 2012; Gerow et al., 2014) did not evaluate its direct effect on performance, while the third (Hou et al., 2023) was conducted in the Triple-A SC capability context.
Based on this finding, a systematic literature review-based meta-analysis was considered necessary to examine the SCAl-performance relationship and evaluate the possible influence of moderating variables. Our meta-analysis consolidates findings and provides a more accurate assessment of the SCAl-performance relationship while considering some performance measures (financial and non-financial) analyzed in previous research. The magnitude and direction of the SCAl-performance relationship outcomes are determined, and the correlation estimates are robust. Some other potentially influential moderating factors or contextual variables have also been identified, enhancing our understanding of the boundary conditions and contextual nuances that shape this relationship. We consider the impact of both traditional control variables, such as industry and economic region, and some less frequently studied moderators, such as SCAl dimensions and performance measure operationalization.
This research contributes to the previous literature by, first, identifying, classifying, and coding empirical articles on the SCAl-performance relationship; second, examining and summarizing the main characteristics of the selected papers; third, grouping and analyzing the performance measures and identifying some moderator variables, and fourth, quantifying the SCAl-performance correlation in the full sample and analyzing whether the moderator subgroup correlations differ significantly.
The following sections complete this article. After this introduction, the theoretical context and underlying assumptions are established. A full analysis, a description of meta-analysis methodology, and the discussion and conclusions follow. The article concludes with findings, knowledge gaps, and recommendations for best practices.
Theoretical Background
The SCAl-Performance Link
SCAl states that a focus company must align all its SC members’ interests with its own (H. L. Lee, 2004) and flexibly adjust the SC network configuration to align its members’ objectives (Dubey et al., 2015). All chain members’ strategies, both internal and external, must be configured to establish common goals, strategies, structures, and processes (Gattorna & Jones, 1998) as SCs are less competitive when their business and information technology strategies are not aligned (Marin-Garcia et al., 2018). So, SCAl positively affects SC flexibility, delivery, quality, and cost (Alfalla-Luque et al., 2018) and plays a major role in SCs achieving a competitive advantage (Attia, 2016). SCAl has also been considered a driver of other dynamic capabilities. For example, Feyissa and Sharma (2016) consider that SC agility and SC adaptability cannot achieve their full potential unless SC members align their interests and are motivated to synchronize their efforts.
SCAl is defined as “the ability for information and knowledge sharing, establishing roles, tasks and responsibilities, and equitably sharing risks, costs, and benefits, with the aim of synchronizing and coordinating processes and activities” along the SC (Alfalla-Luque et al., 2018, p. 8). It helps address the increasing number of challenges of globalization, such as shorter product lifecycles, rapidly evolving technology, and changing management philosophies (Ogulin, 2014), increasing effectiveness through a common strategy and an efficient SC (H. L. Lee, 2004). It is a dynamic capability constructed and constantly readjusted to address variations in consumer demand, markets, and economic structures (Whitten et al., 2012). If SC partners are not aligned, some members could be inclined to optimize their own operations rather than maximize global SC performance. Such self-serving decision-making could affect global performance negatively, resulting in inefficient outcomes (Alfalla-Luque et al., 2024).
There are multiple facets to SCAl, including strategy, customers, suppliers, information/IT, process, and incentive alignment (Marin-Garcia et al., 2018; Mikalef et al., 2014). Therefore, it can be approached globally or from different perspectives (e.g., information/physical flows, suppliers/customers, strategic/operational). Table 1 gives the findings of previous research that has explored the SCAl-performance relationship as a dynamic capability, showing that several studies find a positive and significant SCAl-performance relationship (e.g., W. Ahmed, 2021; Attia, 2015; Charoensuk et al., 2014; Keating, 2022; Pradabwong et al., 2017; Tipu et al., 2019) while others (e.g., Dubey & Gunasekaran, 2016; Huo et al., 2021; Skipworth et al., 2015; Tallon & Pinsonneault, 2011) do not confirm this. The Feizabadi, Maloni, & Gligor, et al. (2019) literature review corroborates these mixed findings, while Cao et al. (2012) find that alignment might not always lead to better outcomes.
Findings on the Relationship Between SCAl and Performance Measures.
Therefore, the literature on the SCAl-performance relationship is divided. However, as SCAl is a dynamic capability, theory suggests that a positive relationship should be expected. Thus, we formulate the following hypothesis:
Hypothesis 1 (H1). There is a positive correlation between SCAl and performance.
The Role of Moderators
The SCAl-performance relationship’s complex and multifaceted nature could be due to the presence of moderating variables. Moderators can be grouped into three categories: substantive (operationalization of the SCAl and performance constructs), extrinsic (industry type and economy type [country developmental level]), and methodological (analysis method used in research and performance measure operationalization).
Construct operationalization is a common moderating variable in meta-analysis (Alfalla-Luque et al., 2023; G. Wang et al., 2018) as variations in operationalization could influence the direction and/or magnitude of the relationship between independent and dependent variables (G. Wang et al., 2018). So, this can be considered a moderating effect that explains between-study heterogeneity (Hancock et al., 2013).
SCAl construct operationalization is included as a substantive moderator. The literature has considered SCAl as both a unidimensional (e.g., Attia, 2015) and a multi-dimensional construct (e.g., Machuca et al., 2021), and also, one or several of its dimensions have occasionally been analyzed separately (e.g., Pradabwong et al., 2017). Meta-analysis categories must contain a sufficient number of works for SCAl operationalization to be used as a moderator. Analysis of previous research shows that sufficient empirical research exists on the dimension Information/IT alignment. However, some other insufficiently researched dimensions for analysis by subgroup have also been investigated, for example, process, incentive, and strategy alignment along the SC (e.g., Pradabwong et al., 2017; Tang & Rai, 2014).
SCAl is frequently conceptualized as a multidimensional capability encompassing information, operations, strategy, and relations. When these dimensions (e.g., information, incentive, strategic alignment) are mutually reinforcing, their combined influence on performance can exceed the sum of their individual effects. Supporting this view, Flynn et al. (2010) show that a general non-disaggregated SC construct explains more variance in performance than any dimension considered separately, highlighting the value of integrated models. In addition, Teece (2018) argues that SCAl functions as a cohesive system and cannot be disaggregated without the dimensions losing their strategic meaning and impact.
Nonetheless, some researchers advocate the use of specific alignment dimensions, arguing that they offer stronger and more actionable correlations with performance as they are clearer causal mechanisms. For example, Wong et al. (2012) find that under environmental uncertainty, specific dimensions of integration, such as information alignment, are more directly related to operational outcomes. Similarly, Huo et al. (2021) shows that alignment dimensions interact differently with firms’ strategic orientations, implying that disaggregation increases the sensitivity of contextual moderators. Following this line, aggregated constructs may dilute or obscure these nuances.
In summary, although both modeling approaches have merit, theoretical, methodological, and empirical considerations suggest that a general SCAl construct is more suitable for capturing alignment’s integrated and systemic impact on firm performance. This leads us to hypothesize:
Hypothesis 2 (H2). The correlation between SCAl and performance is stronger when SCAl is measured with a general SCAl construct rather than specific SCAl dimensions.
Second, performance type has also been considered a substantive moderator in previous meta-analyses of SC variables (Alfalla-Luque et al., 2023; Chi et al., 2009; Hancock et al., 2013; Hou et al., 2023; Kirca et al., 2005; Leuschner et al., 2013; Wowak et al., 2013). Various performance measures have been identified that can be grouped as global or specific. Global performance considers financial and non-financial performance measures (Chahal et al., 2020; Feizabadi, Gligor, & Alibakhshi, 2019; Judge & Douglas, 2009). Specific performance indicators are related to, for example, aspects of operational performance such as quality, time, flexibility, environment, delivery, logistics, and perceived customer satisfaction (Charoensuk et al., 2014; Machuca et al., 2021) and aspects of financial performance such as net profits, financial results, market, economic, sales, revenue, and so on (Li et al., 2015; Skipworth et al., 2015). Table 1 summarizes prior research findings.
The choice between specific indicators or a broader, general construct is important in management research. Venkatraman and Ramanujam (1986) warned that relying on a single type of indicator, typically financial, can obscure important relationships, noting that “a unidimensional composite of a multidimensional concept such as business performance tends to mask the underlying relationships among the different subdimensions” (p. 807). This concern persists decades later. In a review of nearly 2,000 articles published between 2015 and 2019, Bolton et al. (2024) found that approximately two-thirds of studies still relied on a single performance measure and nearly three-quarters did not incorporate multiple perspectives. These findings suggest that management research continues to underutilize general overall performance constructs despite these long-standing theoretical arguments.
However, from both the diagnostic and managerial standpoints, specific performance measures may offer greater clarity and practical relevance. They allow practitioners to link business practices more directly to tangible outcomes. Bozarth et al. (2009) emphasize that operational metrics are often more directly impacted by SC practices than broader, aggregated measures. Although Combs et al. (2005) support the conceptual richness of multidimensional constructs, they also caution that combining distinct performance components may lead to compensatory effects that obscure or weaken observed correlations as each potentially responds differently to business practices. Similarly, Schleicher et al. (2019) argue that performance systems grounded in clearly defined, specific criteria tend to produce more consistent and interpretable outcomes. Zhang et al. (2005) further emphasize the value of targeted performance constructs such as customer satisfaction in capturing the operational impact of SC practices. Collectively, these findings suggest that in many contexts, specific performance measures may offer stronger explanatory power and greater diagnostic utility than aggregated assessments.
Arguments exist both for and against the way that this hypothesis is formulated. However, in line with the previous hypothesis, the arguments suggest that global performance constructs may offer conceptual completeness. We therefore propose:
Hypothesis 3 (H3). The correlation between SCAl and performance is stronger when performance is measured with a general overall performance indicator rather than specific performance indicators.
Extrinsic moderators selected for this research are industry type and economy type (country developmental level).
Contingency theory (Lawrence & Lorsch, 1967) states that contextual factors influence the extent to which business practices are adopted and successfully implemented. The divergence perspective (Ralston et al., 2008) posits that cultural values inherent to each society remain resilient even during industrialization and globalization. Consequently, national culture continues to influence how business practices are interpreted and enacted, ultimately affecting their outcomes. Thus, variations in contextual variables may reveal differences in the adoption and performance impacts of managerial strategies, SC capabilities, and operational practices (Arana-Solares Iván et al., 2019). So, the legal and regulatory environment, cultural factors, economic conditions, infrastructure, logistics capabilities, and supplier and customer networks that vary across regions can significantly influence alignment efforts. Understanding and considering these factors in research and practice could be important for achieving effective SCAl and optimizing performance outcomes. Two primary economic regions have been identified by the extant SC literature: developed and developing (L. Chen et al., 2023; Hou et al., 2023; Machuca et al., 2021; W. Wang et al., 2018). Some scholars argue that SC research has traditionally focused on developed countries (Gebauer et al., 2012), leaving a gap in our knowledge on developing countries (Gebauer et al., 2007; Sharma et al., 2021).
Selecting economic region as a moderator is consistent with previous meta-analyses on SC, although this variable’s influence is controversial. Some studies have found no differences between economic areas. For example, the Chahal et al. (2020) analysis of the SC integration-performance relationship found an insignificant difference between the results for developed vs developing countries. L. Chen et al. (2023) included developed and less-developed regions and found no effect of SC learning on performance. The Alfalla-Luque et al. (2023) assessment of SC agility’s impact on performance found that country type (developed vs developing) did not affect heterogeneity. However, other studies have found significant differences between economic regions. For example, Govindan et al. (2020) found that green and sustainable chain practices led to better firm performance in developed economies than in developing countries. Conversely, the L. Chen et al. (2021) analysis of the SC leadership-performance relationship found that the impact was greater in developing countries than in developed countries and global regions. Similarly, the Abreu-Ledón et al. (2018) analysis of the lean production-performance relationship found a greater effect size in emerging economies than in advanced countries.
In light of this controversy, this research argues that from the perspective of the law of diminishing marginal returns, the incremental benefits of further investment in SCAl tend to decrease as SCAl practices mature and become more institutionalized. In developed economies, where SCAl practices and supporting infrastructure are already well established, additional improvements may yield relatively modest performance gains. In contrast, firms operating in developing economies often start from lower alignment maturity, organizational capabilities, and infrastructure baselines. Consequently, improvements in SCAl can lead to more substantial performance gains as inefficiencies are reduced and coordination mechanisms are strengthened. This pattern is particularly evident in emerging economies, where firms often experience disproportionately large performance improvements when adopting SCAl practices. As Avittathur and Jayaram (2016) note, adopting SC management practices in these contexts can yield significant benefits because they address foundational gaps in process integration and infrastructure. Rai et al. (2006) found that digitally enabled SC integration had stronger performance effects in markets where such practices were relatively rare, often corresponding to developing economies. Similarly, Hult et al. (2007) observed that the benefits of SC integration were magnified under high environmental turbulence, a characteristic more typical of emerging economies. In such contexts, SCAl becomes a strategic capability that enhances agility and resilience, allowing firms to manage uncertainty better and respond effectively to market volatility.
Although developed economies offer more stable regulatory frameworks and broader resource bases, these advantages may not translate into stronger SCAl–performance correlations if the potential for improvement is already exhausted. Instead, developing economies may derive greater marginal utility from alignment initiatives for their capacity to absorb and benefit from advanced practices not yet widely adopted. Therefore, building on the theoretical rationale and empirical evidence, we posit the following:
Hypothesis 4 (H4). The correlation between SCAl and performance is stronger in developing economies than in developed economies.
Industry type is another important moderating variable (Junni et al., 2013) used in SC meta-analysis due to industry-specific characteristics, SC configurations, technological advancements, regulatory and environmental factors, and competitive dynamics. Its effect has been studied on the performance of SC knowledge (Wowak et al., 2013), SC integration (Chahal et al., 2020), SC leadership (L. Chen et al., 2021), SC green management (Geng et al., 2017; Govindan et al., 2020), just-in-time (Sartal et al., 2021), and servitization (W. Wang et al., 2018), for example.
Industries have unique characteristics that could influence SC practices and performance outcomes. Understanding different industries allows researchers and practitioners to account for their nuances and specificities when examining the SCAl-performance relationship outcome. Industry type has traditionally been viewed as a control variable in prior research. The two main groupings typically examined are manufacturing and services, which differ in terms of their core operations: Manufacturing industries entail physical production and transforming raw materials into tangible goods, while service industries primarily involve intangible, knowledge-based offerings (Chase et al., 2006). This distinction has implications for SC practices and performance outcomes, as the effect on performance greatly depends on customers’ subjective perceptions. In addition, measuring this effect may not be straightforward as differences could exist between manufacturing and service SCs, with the latter primarily defined as intangible, inseparable, heterogeneous, and perishable (Chahal et al., 2020).
So, the SCAl-performance relationship could be moderated by the industrial sector, particularly when comparing manufacturing and services. The traditional SC management literature’s greater focus on manufacturing does not imply that SCAl has a stronger correlation with performance in this context.
In contrast, service firms often operate in more dynamic, customer-facing contexts where SCAl may have more direct and visible effects on performance outcomes. The simultaneous nature of production and consumption amplifies the impact of coordination and information sharing. As Sengupta et al. (2006) note, certain types of alignment may produce stronger effects in service environments due to the immediacy of customer interaction and responsiveness. Similarly, Ellram et al. (2004) argue that although services differ structurally from manufacturing, their SCs can benefit significantly from alignment practices, potentially even more so, due to the intangibility and variability of service delivery. Furthermore, Prajogo and Olhager (2012) find that information alignment has particularly pronounced effects in service-intensive environments, where integrating and sharing information is crucial for ensuring consistency and customer satisfaction. These findings suggest that alignment mechanisms, especially those related to information and coordination, may exert stronger effects in service firms, where value is co-created with the customer in real time.
Furthermore, although manufacturing environments benefit from process standardization and infrastructure that facilitate integration (Flynn et al., 2010; Frohlich & Westbrook, 2001; Zhao et al., 2011), they may also be more structurally rigid, and SCAl’s impact on performance may be attenuated by existing process optimization. In manufacturing, SCAl effects are often channeled through established systems, potentially limiting their marginal influence on performance outcomes. Therefore, while manufacturing firms have historically been the focus of SCAl research, there are powerful theoretical and empirical reasons to expect that SCAl practices may have a stronger effect on performance in service contexts. So, we posit the following hypothesis:
Hypothesis 5 (H5). The correlation between SCAl and performance is stronger in the services sector than in the manufacturing sector.
There are many possible methodological moderators, all of which are likely to have significant implications for the results of a meta-analysis and generate implications for theory and applications (Aguinis et al., 2011; Kulinskaya et al., 2008). Including study design- and methodology-related characteristics in a meta-analysis (Lipsey & Wilson, 2001) helps ensure the robustness and reliability of the findings. The methodological moderators chosen for the present meta-analysis were analysis method and performance measure operationalization.
Most studies comparing parameters estimated by CB-SEM focus on assessing the accuracy of estimated loadings or paths, rather than inter-construct correlations (e.g., Binz Astrachan et al., 2014; Becker et al., 2019; Hair et al., 2017; Reinartz et al., 2009; Sarstedt et al., 2017). Dash and Paul (2021) conclude that inter-construct correlations are lower in PLS-SEM than CB-SEM, due to its optimization function, which maximizes the explained variance of dependent variables and tends to overestimate indicator loadings. This overestimation may result in attenuated inter-construct correlations as a greater proportion of within-construct variance is captured than between constructs (Dijkstra & Henseler, 2015). Based on this reasoning, we propose the following hypothesis:
Hypothesis 6 (H6). The correlation between SCAl and performance is stronger when CB-SEM is used rather than PLS-SEM.
A theoretical construct cannot be empirically explored unless it has first been properly specified or made tangible, known as operationalization (Mueller, 2004). This is the most widely employed moderating variable in meta-analysis, as different types of operationalization could affect the size and/or direction of the link between the dependent and independent variables (D. Y. Wang et al., 2018).
Operationalization could therefore be considered a moderating influence that reveals the diversity of studies (Hancock et al., 2013). Different approaches to defining performance could introduce stochastic errors and contribute to heterogeneity in findings due to the impact of variable definitions on actual correlations (Van Wijk et al., 2008). In particular, some performance metrics include internal performance measurements (intra-organizational performance), while others compare a firm’s performance with its primary competitors (inter-organizational performance).
Intra-organizational performance assesses and measures the effectiveness and efficiency of an organization’s individual functions, processes, or departments, i.e., a capability derived from a company’s internal attributes and resources that reflects the organization’s proficiency in executing its core processes and achieving its own performance objectives (Hayes & Wheelwright, 1984). It focuses on evaluating performance outcomes and capabilities within a single organization, thus offering insights into an organization’s internal operations, productivity, and all-round success in achieving its objectives. The organization’s internal data and parameters are considered to evaluate its ability to adapt its products to customer requirements, to quickly launch new products, and its success in reducing delivery and lead times. Understanding and improving intra-organizational performance is important for organizations to enhance their competitiveness, operational efficiency, and customer satisfaction by optimizing internal processes, aligning resources, and improving the global performance of multiple functional areas.
In contrast, inter-organizational performance refers to the assessment and measurement of the global effectiveness and efficiency of a multi-member SC network. It focuses on the collaborative efforts, coordination, and integration of SC partners to achieve shared goals and objectives. Inter-organizational performance measures such as market share and competitive advantage provide a comparative framework that may accentuate the impact of SCAl practices and also give insights into the performance outcomes of the entire SC network and the collective impact of the relationships and interactions among its members. Derived from its attributes and resources, inter-organizational performance is a capability that distinguishes an organization and enables it to outperform its competitors (Hayes & Wheelwright, 1984). It can come from a variety of sources: superior quality, more advanced technology, rapid response to market changes and needs, differentiation of products and services, and so on, and it is more valuable and sustainable when it is difficult to imitate (C. J. Chen, 2019).
Performance measurement through intra- or inter-organizational indicators can influence the observed strength and interpretation of the SCAl-performance relationship. Although the operations management literature lacks clear references regarding the differential effects of proximate versus distal performance measures, research in other management areas provides valuable insights. Human resource management, in particular, consistently demonstrates that intermediate or proximal outcome measures tend to be more strongly correlated with the use of specific management practices than final or distal outcomes (K. M. Douglas & Sutton, 2010; Jiang et al., 2012; Katou, 2017; Kehoe & Wright, 2013; Marin-Garcia, 2013). For example, meta-analytic evidence shows that human resource practices present stronger relationships with employee attitudes and behaviors (proximate outcomes) than with ultimate firm performance indicators (Hancock et al., 2013; Jiang et al., 2012).
Drawing from this theoretical foundation, we argue that intra-organizational measures such as internal operational indicators represent proximate outcomes that are more directly influenced by SCAl practices within the firm. Inter-organizational performance measures such as comparative, relational, or market-based indicators are distal outcomes that, while capturing the strategic and collaborative value of SCAl, may be influenced by numerous external factors that weaken the observed correlation. Given this reasoning, the following hypothesis is posited:
Hypothesis 7 (H7). The correlation between SCAl and performance is stronger for intra-organizational performance than for inter-organizational performance.
Methodology
Meta-analysis is the methodology chosen to address the SCAl-performance relationship in this research. This method allows us to aggregate findings to establish whether a relationship exists and, if so, to determine its size and dispersion (Hak et al., 2018; Hunter & Schmidt, 2004; Y. H. Lee, 2019). Findings of multiple empirical studies can be assessed and integrated, thereby enhancing the robustness of the estimated effect size compared to an individual study (Borenstein et al., 2009).
The systematic literature review to select the studies was performed on recommendations and practices in previous papers (Alfalla-Luque et al., 2013; Andreu-Andrés et al., 2018; Fadahunsi et al., 2019; Pertusa-Ortega et al., 2020; Sartal et al., 2021; Tranfield et al., 2003). In common with other meta-analyses (Alfalla-Luque et al., 2023; Pertusa-Ortega et al., 2020), the automatic search strategy covered the period up to October 24, 2024, in two relevant databases: Scopus and Clarivate-Web of Science (WoS). The search strings were: WoS: (suppl* AND chain* AND (align* OR “Triple-A” OR “Triple A”)) (Title) AND Articles (Document Types); Scopus: TITLE (suppl* AND chain* AND (align* OR “Triple-A” OR “Triple A”)) AND (LIMIT-TO (DOCTYPE, “ar”)). The term Triple-A SC was included as it comprises SCAl. WoS yielded 153 items and Scopus 186. Figure 1 shows a PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) diagram (Liberati et al., 2009) of the process followed to select empirical articles, and the obtained results.
Records identified through a search of the WOS/Scopus databases (PRISMA).
After comparing the results and removing duplicated items, 218 items remained. A two-phase selection and filtering process followed. First, titles and abstracts were screened to identify whether articles coincided with the research objective, with 59 excluded. Then, the content of the remaining 159 articles was examined, and exclusion criteria were applied (see Figure 1). Following the same procedure as other papers (Alfalla-Luque et al., 2023; Geyskens et al., 2009; Iftikhar et al., 2021), analyses using identical samples were excluded: Attia (2016), same sample as Attia (2015); W. Ahmed and Rashidi (2021), same sample as W. Ahmed (2021); Dubey et al. (2015), same sample as Dubey and Gunasekaran (2016); Feizabadi, Maloni, & Gligor, et al. (2019); Feizabadi et al. (2021), same sample as Gligor et al. (2020), and Alfalla-Luque et al. (2018) same developed country subsample as Machuca et al. (2021). Three other studies were excluded as they did not provide compatible item-level measures for the construct of interest. Specifically, Badi (2024) focused on a blockchain platform and did not report item-based measurements aligned with our inclusion criteria. Nazeer et al. (2024) was excluded as the instrument items could not be identified or verified, preventing assessment of measurement equivalence. Keating (2022) was excluded as the study operationalized the construct using a single indicator that was incompatible with item-based measures used in the other studies. Any missing information was initially requested from article authors by e-mail, with a reminder sent if it was not received after three weeks. A final total of 29 articles remained for coding and export to the correlation table.
Table 2 lists the selected articles and some of their characteristics, including the number of dimensions considered to build SCAl. Some of the 29 articles included and evaluated more than one SCAl-performance correlation, as each alignment and/or performance type was treated independently. Each correlation constituted one input record. As Machuca et al. (2021) included two sub-samples (developed and developing countries), Table 2 lists 29 papers but has 30 rows. A sample size of 30 units is consistent with the samples used in previous meta-analyses, for example, Wowak et al. (2013), 35 articles; Gerwin and Barrowman (2002), 26; Mackelprang and Nair (2010), 25, and Nair (2006), 23.
Empirical Papers Selected for Meta-Analysis.
C/R, correlational or regression analysis; 3SLS, Three-Stage Least Squares; QCA, qualitative comparative analysis.
Coding in Excel templates was based on recommendations in earlier works (Alfalla-Luque et al., 2023; Losilla et al., 2018; Villiger et al., 2021). The Pearson correlation was taken as a metric of effect size (Alfalla-Luque et al., 2023; Barroso-Méndez et al., 2015; Sierra-Morán et al., 2024). Information was extracted by labeling the code fragments corresponding to our research questions and filling out a card for each of the references. Missing information was labeled “not available” (N/A) (Knight et al., 2019; Sanchez-Ruiz et al., 2018). Economic region was classified using information from the statistical annex of “World Economic Situation and Prospects 2022” (United Nations, 2022).
In a pilot test, the coding team discussed the procedure and coded two randomly selected articles from the meta-analysis, thus improving the technique and code list. Two coders then independently extracted and coded the information obtained from the articles. In cases of doubt, for example, if the correlations were absent from the publication or numerous outcomes were provided, the coding protocol was revised, with the two coders coding these papers separately and comparing their findings. When their data matched, it was considered a consensus score. Discrepancies were discussed by the coders in an attempt to come to an agreement. A lack of agreement would have been recorded as a coder dispute, and the third coder would have adjudged. However, as the two coders concurred in every case, the agreement index was 100%, and formulas such as Kappa, ICC, Krippendorff’s alpha, and the index of reliability (LeBreton & Senter, 2008; Perreault & Leigh, 1989; Villiger et al., 2021) were not required.
Some articles addressed multiple relationships, for example, between SCAl and various performance measures (Attia, 2015; Geyi et al., 2020; Mikalef et al., 2014; Pradabwong et al., 2017; Pu et al., 2020; Ryoo & Koo, 2013; Sheel & Nath, 2019). We considered a reductionist approach to be a better solution than an integrative approach (López-López et al., 2018) as our data would foreseeably present within-study heterogeneity. However, as there was an insufficient number of studies for a multi-level focus, when correlation values were available for several SCAl dimensions for one type of performance (or several performance dimensions for one type of SCAl), the Geyskens et al. (2009) recommendations were followed. So, rather than calculate a mean value that would assume that the studies’ effect sizes were independent, the Fisher Z-transformation was used (as described in Borenstein et al. (2009), Chapters 6, 7, and 24) to calculate composite effect sizes in order to take into account interdependencies between practices (Abreu-Ledón et al., 2018). The value set correlation was obtained by calculating the average of the transformed values and decompressing the Fisher Z-transformation.
Following the Borenstein et al. (2009) recommendations, we assessed the main effect size, heterogeneity, and moderators for the data analysis. As the obtained articles were considered a research sample based on different populations in which different processes or measurement instruments had been used, the random-effects model was estimated (Borenstein et al., 2019; L. Chen et al., 2021; Hedges & Vevea, 1998; Hunter & Schmidt, 2002). Since no agreement exists as to whether it is better to attenuate (Lipsey & Wilson, 2001) or unattenuate (Hunter & Schmidt, 2004; Mackelprang et al., 2014), and given that both options have advantages and disadvantages from the theoretical and the practical points of view, we present the attenuated results (Geyskens et al., 2009). Unattenuated correlation data are provided in the Online Annex Supplemental Material.
In addition to the traditional heterogeneity statistics commonly reported in meta-analyses (Q, p-Q, τ², and I²) (Higgins & Thompson, 2002), we employed more informative approaches to assess heterogeneity. Following recent recommendations (Borenstein, 2022; Borenstein et al., 2009, 2019; Brannick et al., 2021; IntHout et al., 2016; Kepes et al., 2024; Lin, 2019; Mariano Faggion et al., 2021; Nagashima et al., 2019), we focused on the prediction interval, primarily, and the credibility interval width as these provide more meaningful information about the range of true effects across studies.
We calculated two additional heterogeneity indices (Mackelprang et al., 2014; Nair, 2006): Ratio 1 is the average corrected correlation divided by the estimated population standard deviation and can be interpreted as a z-value and an alternative to confidence intervals for determining whether the effect size differs significantly from zero. Ratio 2 is the ratio of the weighted average sampling error variances to the variance of corrected correlations; values below 0.75 indicate the presence of substantial heterogeneity and are an alternative to prediction intervals for heterogeneity assessment.
We followed established procedures for subgroup analysis (Borenstein et al., 2019; Sierra-Morán et al., 2024), using ANOVAs to compare between-group heterogeneity (Qb) and within-group heterogeneity (Qw). This approach tests the null hypothesis that all subgroups have equivalent effect sizes. A p-value < .05 indicates that at least one subgroup differs significantly from the others in effect size magnitude, regardless of within-group variability
We also conducted a sensitivity analysis to assess the robustness of our findings and identify potential outliers. A Baujat plot was used to detect influential studies and was validated through a leave-one-out procedure.
Publication bias was analyzed with the Trim and Fill technique (Geyskens et al., 2009), enabling us to simultaneously test for publication bias risk and estimate the modified effect size (Borenstein et al., 2019).
Our analysis uses a hybrid approach combining key aspects of two methods, leveraging the advantages of both (Kisamore, 2008). From Hedges-Olkin, we apply Fisher’s Z-transformation, inverse variance weighting, a random-effects model with the HE estimator, confidence and prediction intervals, and Z-tests for significance. From Hunter-Schmidt, we incorporate final reporting in the correlation metric by reversing the Z-transformation. The metafor package in R (Viechtbauer, 2010) was used to perform the statistical analyses (https://github.com/jamg-upv/ART-693MetaAlignPerf).
Results
The total sample of articles (see Table 2) corresponded to 6,658 companies, with 46.7% from developed countries, 50.0% from developing countries, and the remaining 3.3% from studies covering both regions. The questions or items were rated by respondents on predominantly 7- or 5-point Likert-type scales. The most used analytical methods were PLS (36.7%) and CB-SEM (36.7%).
SCAl was measured as a general SCAl construct in 66.7% of cases and as specific SCAl dimensions otherwise (33.3%). The construct conceptualizations in the selected papers (Table A in the Online Annex Supplemental Material) were in line with the reference definition followed in the present research (Alfalla-Luque et al., 2018). Some key points were identified: (1) information and knowledge sharing, emphasizing its importance for enhancing collaborative efforts; (2) process and activity coordination, underscoring the need to align activities and processes to achieve common goals effectively; (3) distribution of risks, costs, and gains; (4) strategic partnerships with suppliers and customers; (5) performance improvement through shared goals, effective communication, and collaboration.
The performance construct, definitions, dimensions, and analyzed items in the sample papers are summarized in Table B in the Online Annex Supplemental Material. In line with Alfalla-Luque et al. (2023), the measures found were related to two types of performance: specific performance indicators (43.3%), and overall or global performance (56.7%). Intra-organizational performance measures were used in 40.0% of studies, and inter-organizational in 50.0%.
Table 3 presents the key characteristics of the studies included in the meta-analysis, including sample sizes, reliability coefficients for both SCAl and performance measures, the originally reported correlations, the type of correlation (attenuated or unattenuated), and the corrected correlations used in our analysis. All correlations were standardized to their attenuated form to maintain consistency across studies. When studies reported unattenuated correlations, we applied an inverse Hunter and Schmidt (2004) correction for measurement error to obtain attenuated estimates. Attenuated correlations were not modified. This approach provides a more conservative analysis, using correlations that reflect the measurement error typically present in empirical research.
Sample Size, Reliability Coefficients, and Correlations of Selected Studies.
Table 3 shows two papers with strong correlations (Charoensuk et al., 2014; Tipu & Fantazy, 2020). Both studies were conducted in the context of developing countries and used reflective (common factor) measurement models, which, taken together, could contribute to more powerful associations. However, we found no clear patterns among the three studies with weaker correlations (Dubey & Gunasekaran, 2016; Mohaghegh et al., 2024; Tallon & Pinsonneault, 2011), with two conducted in developed countries, and the other in developing economies. One focused on services, another on manufacturing, and the third, both. Similarly, their performance measures varied, including operational, financial, and global performance indicators.
Main effects
The forest plot generated from the SCAl global analysis (see Figure 2) gives a comprehensive visual summary of study effect sizes and confidence intervals (CI) and a pooled estimate for attenuated correlation coefficients. All the studies presented positive relationships (effect size) while only one (Mohaghegh et al., 2024) presented a CI that included zero. No clear publication trend emerged across the nearly 20 years analyzed. The pooled attenuated correlation was 0.385 with a CI of [0.318, 0.449]. As the CI interval did not include zero, the combined effect size was considered significant (ratio1 > 1.96). According to recommended cut-off values, pooled correlation values were positive, between “large” (following Bosco et al., 2015) and “medium-large” (Cohen, 1988). However, observed heterogeneity indicates that future studies may well find correlation values outside this confidence interval and that the estimated value may reflect genuine differences across industries, organizational contexts, cultural settings, and methodological approaches rather than simple sampling errors.
“Forest Plot” total sample attenuated correlations.
The prediction interval (PI) [0.006, 0.668] and the credibility interval (CR) [0.013, 0.664] both indicate that, in future research, the attenuated correlation between SCAl and performance could be as low as 0.006 in some populations and as high as 0.668 in others. The tau² estimate (τ² = 0.040) is the real between-study variance on Fisher’s Z-transformation scale. This metric is traditionally used with the Q statistic (Q = 329.20, p < .001) or with the I² index (89.73%) to analyze heterogeneity, indicating that the observed between-study variability in effect size is much higher than can be attributed to sampling error. The ratio 2 value is 0.112, corroborating the high heterogeneity identified by PI and CI.
Although wide, PI and CR do not include negative values at their lower bounds, suggesting considerable variation in the true effect size across different contexts and populations. While the pooled correlation estimate and its CI indicate a positive relationship overall, the wide PI suggests that future studies in certain contexts or populations may indeed find correlations close to zero between SCAl and performance. This does not invalidate the general positive trend observed across studies but, rather, indicates that the relationship may be context-dependent.
Without additional moderator analyses to identify the specific conditions under which low relationships might emerge, SCAl cannot be concluded to universally enhance performance across all contexts. Therefore, while H1 receives general support from the overall positive correlation, substantial heterogeneity suggests the need for more context-specific theorizing about when and where this relationship applies.
Moderation Analysis
Moderator analysis and subgroup results for attenuated correlations are presented in Table 4. We consider that PI provide more meaningful information about the range of the real effects in the subgroups than traditional statistics such as I² or Q. PI are consistently wide across most subgroups, indicating substantial within-category heterogeneity. Seven subgroups showed PI that excluded negative values: SCAl general construct (PI: 0.007–0.660), global performance (PI: 0.009–0.685), developed regions (PI: 0.086–0.591), manufacturing industry (PI: 0.049–0.662), studies using CB-SEM analysis (PI: 0.010–0.783) or PLS-SEM (PI: 0.071–0.507), and inter-organizational performance measures (PI: 0.148–0.499). These subgroups suggest more consistent positive SCAl-performance relationships, although the wide intervals indicate that substantial variability remains even within these categories (albeit to a lesser extent for inter-organizational performance). This suggests that the proposed moderators were insufficient to create truly homogeneous subgroups, indicating that the SCAl-performance relationship is more complex than can be captured by these categorical distinctions. On the other hand, the published study sample only contains two specific works on the services sector, a number that is too small for this subgroup’s estimation to be considered reliable.
Summary of Attenuated Correlations.
Notes: CI, confidence interval; LL, lower limit; UL, upper limit; 95% CI for correlation. Q: Cochran’s Q; pQ: p-value of Q; l2: inconsistency ratio; τ2: estimated amount of between-study variance. PI, prediction interval; LL, lower limit prediction; UL, upper limit prediction; 95% PI for correlation. CR, credibility interval; LL, lower limit; UL, upper limit).
Interpret subgroup estimates with caution due to the extremely small number of studies (n = 2) included in this meta-analysis.
Table 5 presents the Q statistics used to test between-subgroup differences for each moderator. Analysis reveals a consistent pattern across all moderators: while significant between-group differences exist (all Qb values p < .001), substantial within-group heterogeneity persists (all Qw values p < .001). The Qb values indicate that economic region, industry type, SCAl dimension, performance type, intra- vs inter-organizational performance, and analysis method account for meaningful portions of the observed variability in effect sizes. However, the Qw values demonstrate that said moderators fail to form truly homogeneous subgroups as substantial unexplained within-category variance remains. Persistent within-group heterogeneity across all moderators indicates that new categories overlapping two or more of the current moderators or additional unmeasured factors may influence this relationship (potentially including contextual variables not captured in the primary studies, methodological differences, and more nuanced theoretical distinctions).
Q Values for Attenuated Correlations.
It is important to note that despite the persistent within-group heterogeneity, the between-group differences observed in Table 5 are statistically significant and unlikely to be fortuitous. However, despite being significant, the differences in some correlations have virtually no real impact; for example, for specific SCAl measures (r = .396) vs general SCAl constructs (r = .380), or for specific performance measures (r = .366) vs overall performance measures (r = .400). Similarly, across economic regions, developed countries show a lower correlation (r = .365) than developing countries (r = .416). However, other differences are more marked: for example, the methodological moderators, where intra-organizational (r = .429) vs inter-organizational performance measures (r = .335) reveal more variation, as does the choice of analysis method, with CB-SEM studies yielding higher correlations (r = .486) than PLS-SEM studies (r = .305).
Most subgroup correlations fall in the medium-large range (Bosco et al., 2015), developing, CB-SEM, and intra-organizational subgroups having a large correlation.
Analyses were conducted to test the sensitivity of the results. First, two studies (Charoensuk et al., 2014; Tipu & Fantazy, 2020) were identified as potential outliers. However, analysis with a leave-one-out procedure indicated that it had virtually no influence on the overall effect size.
After identifying outliers, we examined whether higher values of the square root of the product of SCAl and performance reliabilities were associated with larger effect sizes. Meta-regression analysis revealed a negligible negative trend with substantial scatter across studies, indicating considerable heterogeneity not explained by this moderator. The reliability moderator accounted for virtually no variance in effect sizes (R² < 0.00), and the moderator test confirmed no significant moderation effect (QM(df = 1) = 0.813, p = .367). These results demonstrate that measurement reliability does not systematically influence the magnitude of the observed relationship, supporting the robustness of our primary findings.
Funnel plot analysis revealed some asymmetry (see Figure 3), with a noticeable lack of studies in the lower-right quadrant of the plot, suggesting a potential under-representation of studies with larger effect sizes and smaller standard errors. To assess the potential impact of this asymmetry, we applied the trim-and-fill method, imputing five hypothetical missing studies (shown as open circles) on the right side of the funnel plot. However, a subsequent adjusted effect size estimate showed no substantial difference from the original pooled estimate. This minimal change suggests that, while potentially present, publication bias does not affect our results.
Trim-and-fill analysis of publication bias (attenuated values).
In addition, we analyzed whether journals with a higher impact factor (CiteScore, 2023) tend to publish papers with higher Z-transformed correlation values than those with a lower impact factor. Meta-regression showed that journal impact factor does not explain the variance in effect sizes (R² < 0.00). The moderator test (QM(df = 1) = 0.2090, p = .648) indicated that the journal impact factor relationship was not significantly moderated, suggesting that there is no systematic association with the magnitude of reported correlations.
Discussion
These meta-analysis results provide evidence to support H1: on balance, the SCAl-performance relationship is positive and statistically significant. This suggests that, on average, greater SCAl is associated with better performance results. This aligns with the theoretical proposal that SCAl is an organizational management lever (Alfalla-Luque et al., 2018; Attia, 2016). Correlation values are medium to high. Furthermore, the fact that practically all the studies present positive effects, with only one having a confidence interval including zero, suggests that the lack of significance in some works could be due more to statistical power, context, or measurement issues than the real non-existence of the relationship. Nonetheless, observed heterogeneity helps explain why the prior literature offers mixed results (Cao et al., 2012; Feizabadi, Maloni, & Gligor, et al., 2019). The estimated correlation should not be interpreted as representing the general company population as in some specific environments, alignment can have higher or lower and even nonsignificant correlations.
Note that the correlation between SCAl and performance does not indicate causality (Rohrer, 2018). Alternative explanations are still plausible, even when average correlation is positive and significant (Antonakis et al., 2010; Rohrer, 2018; Sarstedt & Danks, 2021): inverse correlation (better performing companies can invest more in alignment mechanisms), omitted variables that influence both (e.g., environmental uncertainty, network complexity, digital capability, governance quality), and measurement and common method bias when SCAl and performance are captured with self-reported surveys. However, if a causal relationship does exist, it must manifest as an observable correlation. Thus, the meta-analysis findings meet this necessary condition, although not sufficiently so to indicate causality, and provide a solid empirical base that justifies future research to examine the specific contingency conditions under which SCAl generates a higher or lower impact on organizational performance through designs that enable more robust causal inferences.
SCAl construct operationalization was considered as the first substantive moderator. Results indicate that using general SCAl constructs demonstrates a very similar correlation to specific SCAl dimension constructs (H2). Although the difference is significant, from a practical point of view, it does not represent a relevant change. This finding does not align with previous studies such as Flynn et al. (2010) and Teece (2018), which advocate the use of general constructs that encapsulate the aggregated essence of dynamic capabilities, and appear to indicate that excessive differences should not be expected from the estimated correlation perspective.
Similarly, results indicate that the SCAl-performance relationship is stronger when performance is assessed using general indicators rather than specific performance metrics (H3). This finding aligns with recent methodological critiques (e.g., Bolton et al., 2024), which emphasize the underutilization of multidimensional constructs in management research and call for broader, theory-driven approaches to performance measurement. However, despite being significant, the difference in between-group correlations does not represent a very large change in magnitude. This contrasts with the empirical literature, which favors specific indicators for their diagnostic clarity and practical relevance (e.g., Bozarth et al., 2009; Schleicher et al., 2019). While these studies suggest that specific metrics are more sensitive to the direct effects of alignment practices, our findings imply that global performance constructs may capture the relationship with SCAl’s systemic and integrative nature better or in a very similar way, at least.
In line with contingency theory (Lawrence & Lorsch, 1967) and the divergence perspective (Ralston et al., 1997), the context in which SCAl is applied also appears to moderate its effect, demonstrating the influence of contextual factors on business practices, implementation, and outcomes. Results show that the SCAl-performance correlation is stronger in developing than developed economies (H4), supporting the theory of diminishing marginal returns, whereby firms in advanced economies may already have captured most of the gains from SCAl. Conversely, as firms in developing economies often start from lower baselines, they may experience greater performance gains. However, the correlation values are not excessively different. This finding contrasts with the expectation that stable institutions and advanced infrastructure in developed countries enhance the effectiveness of alignment practices (Swink et al., 2007; Wiengarten et al., 2010). Our results indicate a positive and moderated correlation that is fairly similar in magnitude in both contexts.
H5 was initially intended to analyze differences between industrial firms and service companies. However, the systematic search only yielded two service sector studies. Such a small number renders the subgroup effect estimations unstable (Valentine et al., 2010) as they are very sensitive to each of the studies; within-subgroup heterogeneity cannot be reliably measured, and above all, the between-subgroup differences comparison lacks consistency. This is relatively common in meta-analysis: attempts at stratification by context attributes sometimes “leave too few observations” for an interpretable quantitative synthesis (Hrabec et al., 2022). We therefore report the results as exploratory and defer any comparative inferences until a greater body of evidence exists in services. This limitation reflects a publication bias in the SCAl/performance literature toward studies in manufacturing contexts. Such a lack of empirical evidence in the services sector is a gap in the literature that should be addressed in future research.
Concerning methodological factors, we examined whether the choice of analytical technique influenced the observed SCAl–performance relationship (H6). Contrary to the criticism that PLS-SEM-based studies supposedly present stronger effects due to the method’s flexibility and suitability for smaller samples (Reinartz et al., 2009), our results show that studies using CB-SEM tend to report stronger SCAl–performance correlations. Finally, whether performance is framed through intra- or inter-organizational indicators also influences the strength of the observed relationship (H7). In both cases (H6 and H7), the between-group differences are not only significant but they can also be considered relevant as there is a change of the range in the correlation magnitude.
In sum, this study confirms SCAl’s correlation with performance and indicates that relevant differences stem mainly from methodological choices, whereas differences related to measurement approaches and geographic contexts, although statistically significant, are not substantively meaningful.
In terms of implications for researchers, this research contributes empirical evidence by identifying moderators that explain differences between correlations. Research describing how SCAl relates to different performance factors, including financial and non-financial results, is very scarce. Furthermore, current research frequently examines the phenomenon too broadly, ignoring the unique characteristics of different contexts and circumstances (Ogulin, 2014).
The remaining heterogeneity after the moderator analysis suggests that contextual and additional methodological factors that influence the SCAl-performance relationship could exist that were not examined in this study due to the unavailability of primary studies that systematically report said variables. This limitation is not restricted to the present study; recent SC management meta-analyses have identified similar constraints, for example, L. Chen et al. (2023) on SC learning, Beigi Firoozi et al. (2024) on SC agility, and Hou et al. (2023) on Triple-A SCs, which recognize the impossibility of analyzing multiple potentially relevant moderators due to the limited numbers of available studies.
Failure to identify meaningful moderators suggests that the current empirical literature remains fragmented, with insufficient convergence around specific contextual factors that influence this relationship. In such cases, heterogeneity may reflect inadequate reporting of contextual information in primary studies that prevents effective moderator coding. Consequently, more studies on the SCAl-performance link are needed to increase the volume of data and reduce heterogeneity. In addition, studies should clearly specify the type of company under study; for example, no distinction was made between manufacturing and services in some firms studied in our meta-analysis.
From a theoretical perspective, SCAl as an organizational capability should be affected by factors such as company size and age that influence the ability to develop and deploy resources along the SC in a coordinated way. Others, such as product complexity, investment level in IT, and the degree of environmental uncertainty, could moderate SCAl effectiveness as greater complexity and uncertainty demand higher levels of synchronization and adaptation among chain members. In addition, from the perspective of the resource-based view, variables such as prior experience in inter-organizational collaboration, the presence of formal governance mechanisms, and the specific type of alignment implemented (strategic vs operational, or information-focused vs physical processes) could explain some of the variability not captured in the analysis.
The literature on SCAl suggests that not all forms of alignment are equally effective in all contexts and that the congruency between the type of alignment and the characteristics of the competitive environment could be decisive. However, variations in the ways that studies report these variables make any comparison impossible, ruling out their quantitative analysis in the present meta-analysis. So, future research to examine these potential moderators should be designed to allow the testing of hypotheses specifically derived from the conditions under which SCAl is related to organizational performance.
This research also has practical implications. As managerial resources are finite, decision makers must prioritize where to focus first. Our findings support a positive and statistically significant link between SCAl and performance; however, this evidence is correlational, so it should be interpreted as guidance on where improvements are most likely to bring the greatest benefits rather than as proof of direct causality. In practice, managers can strengthen SCAl by targeting three complementary levers: incentive alignment (clarifying and aligning partners’ interests through shared goals and gainsharing), information alignment (improving transparency and timeliness of data exchange to reduce uncertainty and support coordinated decisions), and process alignment (synchronizing planning and execution through shared routines, standards, and interfaces). Given the heterogeneity observed in the literature, the relative payoff of each lever is expected to vary across contexts; therefore, firms may benefit from diagnosing their main source of misalignment (opportunism and conflicting incentives, lack of visibility, coordination frictions) and prioritizing the corresponding interventions. Overall, SCAl can help mitigate coordination risks and improve end-to-end efficiency and responsiveness across the supply chain network.
This analysis provides researchers with a curated set of peer-reviewed articles on the SCAl–performance relationship and a reproducible process and results, including a meta-analytic assessment of several moderating variables. We have also distilled an integrative conceptual figure (Figure 4) to make this evidence base easier to interpret and reuse. The figure identifies the key dimensions of SCAl (synchronization, coordination, activities, supplier-customer relationships, objectives and strategies, material and resources, information, process), their relationships with other performance dimensions (general, operational, financial, market performance), and the set of contextual and methodological moderators that define the magnitude of this association. The conceptual framework includes both extrinsic factors identified by our analysis (industry type, economic region) and potential moderators suggested in the theoretical literature that could not be examined due to the limitations of primary study availability (complexity, technology maturity, organizational capabilities, stakeholder network characteristics, regulatory environment). It provides researchers with a roadmap for future research that indicates both empirically tested relationships and gaps where greater theory development and empirical evidence are required, especially on the use of aggregated scales to measure SCAl and general performance metrics to assess the success of alignment initiatives.
Conceptual framework of the sources of heterogeneity in the observed correlation between supply chain alignment and performance.
Conclusions
This study contributes to current research with a meta-analysis that evaluates empirical research on the SCAl-performance link. Building on previous research, it provides comprehensive empirical evidence of how methodological and contextual contingencies shape this relationship. Findings reveal a significant, positive, and reasonably strong SCAl-performance correlation across the full sample and all moderator subgroups and confirm that this correlation applies across different performance types, economic regions, and methodological approaches. However, total sample and within-subgroup heterogeneity suggest that, while our understanding of this relationship remains incomplete, the identified moderators represent meaningful starting points for developing more fieldwork that can better account for contextual factors that influence the SCAl-performance relationship.
Hypotheses H1, H3, H4, H6, and H7 were supported, demonstrating that the strength of the SCAl–performance relationship is influenced by the way that performance is measured, contextual variables, and methodological issues. However, only the last appear to have any practical relevance. H2 was not supported by our data as the higher correlation is in the opposite group to the hypothesized group and, in any case, although significant, the difference has no practical relevance. H5 could not be tested as there were only two studies in the services group, emphasizing that research in this sector must be extended.
These insights call for theoretical and empirical design in future research and a nuanced implementation of alignment strategies in practice. Taken together, these findings reinforce the importance of conceptual and methodological rigor when evaluating the SCAl–performance link. They also suggest that researchers and practitioners should consider how SCAl and performance are operationalized and in what context. While the overall relationship is clearly positive, its strength may be contingent upon these critical design choices.
Finally, reporting in published studies must be improved. Articles should contain appropriate and complete information to offer a satisfactory understanding and characterization of the population studied, the methodology used, the assumptions made, the results, and in general, all the information needed for the study to be replicated. Some of the reviewed articles lacked information about the sample (size, origin, economic region, country, type of industry); methodology (informants, scales used, analysis model); statistical information (means, correlations, standard deviations, whether values are attenuated or not); and variables (construct definition, source of the scales). Unfortunately, this lack of information is not uncommon, as was demonstrated by Bayonne-Sopo et al. (2020) in their examination of PLS use in OM, where the degree of detail about such basic parameters as, inter alia, the correlation matrix, was suboptimal in many papers.
Limitations and future research
As in all meta-analyses, this paper has an inherent limitation concerning the generalizability of the results, which ultimately depends on the representativeness of the primary studies included. Although only peer-reviewed articles indexed in the leading academic databases were included in this SLR to ensure data quality, the scope of generalization was constrained by the limited number of available studies. This was especially evident in certain categories, such as the service sector, where the number of empirical studies was notably small.
Academics can find the meta-analysis data, conclusions, and suggestions valuable for future studies. However, further precise conclusions about how SCAl impacts performance cannot be drawn from the current analysis due to the heterogeneity found in both the total sample and the subgroups. More research into the SCAl-performance link is therefore needed to increase data volume and reduce heterogeneity.
Inconsistencies in the conceptualization and operationalization of SCAl are a significant challenge in this field. Most studies treat SCAl as a unidimensional construct, while only a minority analyzes specific dimensions such as information, process, and incentive alignment. This conceptual heterogeneity likely contributes to variability in effect sizes. Future research should seek greater definitional clarity and explicitly specify which alignment dimensions are being studied. Finally, the lack of information in the sample papers precluded the inclusion of some due to missing key data and their failure to evaluate moderating variables such as firm age and firm size, which is a further limitation.
In light of these difficulties, below, we formulate some specific recommendations for the scientific community to facilitate future quantitative syntheses and improve the systematic accumulation of knowledge in the area. First, articles should report not only the items used to measure SCAl and performance, but also the reference scales, explicitly citing the original authors and accurately indicating whether they have been adapted with suppression, addition, or any modification of the wording of the original items. This would enable studies to be compared and any possible sources of methodological heterogeneity to be detected. Second, when the available scales are not suitable for capturing specific aspects of SCAl in some contexts, new useful and reusable scales must be devised and rigorously validated. However, should validated scales be readily available, the scientific community should prioritize their systematic reuse instead of developing new versions, unless some evident theoretical or contextual justification exists that supports such a change. Third, studies that use hierarchical constructs should report the average and standard deviations of the lower-order constructs in addition to the aggregated higher-order constructs, and also the correlations between the lower-order constructs. This information is essential for understanding specific SCAl-performance relationship dimensions and could reveal differential association patterns that remain hidden in aggregated analyses.
Finally, articles should present disaggregated descriptive statistics by subsample when the control or moderating variables are categorical (e.g., industry type, country development level), and lower-order construct correlations with the moderating variables when these are quantitative. This would enable more precise moderation analysis in future meta-analyses and the examination of contingency conditions that cannot be evaluated currently, due to the lack of information reported in primary studies.
Supplemental Material
sj-docx-1-brq-10.1177_23409444261429526 – Supplemental material for Uncovering the Connection Between Supply Chain Alignment and Performance: A Meta-Analysis
Supplemental material, sj-docx-1-brq-10.1177_23409444261429526 for Uncovering the Connection Between Supply Chain Alignment and Performance: A Meta-Analysis by Enrique Bayonne-Sopo, Juan A. Marin-Garcia and Rafaela Alfalla-Luque in Business Research Quarterly
Footnotes
Author’s Note
Enrique Bayonne-Sopo is now affiliated to Universidad Europea de Valencia, Valencia, Spain.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study has been conducted within the frameworks of the following funded competitive projects: PID2019-105001GB-I00 by MCIN/AEI/10.13039/501100011033 (Ministerio de Ciencia e Innovación-Spain), and PY20_01209 (PAIDI 2020-Consejería de Transformación Económica, Industria, Conocimiento y Universidades -Junta de Andalucía).
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
