Transaction Cost Theory, originally advanced by [11], addresses the fundamental question of how firms organize economic activities: whether to internalize operations within the firm or rely on the market. The theory assumes that transaction partners pursue efficiency and profitability while managing risk. Key characteristics of transactions include their frequency and volume, which directly influence costs. Two main uncertainties affect transactions: behavioral uncertainty, stemming from opportunistic tendencies by actors, and environmental uncertainty, caused by external shocks and unpredictable conditions. [25] explain that information asymmetry (where some actors possess information not shared across the transaction) raises costs and risks.

[53] adds that such asymmetry undermines efficiency, as not all parties operate with the same knowledge base. To mitigate these risks, transaction cost theory proposes governance mechanisms that lower uncertainty and facilitate transparent, validated, and secure information flows across the supply chain. In the context of supply chains, [12] note that strategists must address transaction risks to design economic safeguards. [47] further show that the theory is valuable in shaping governance decisions that support resilience and continuity. This makes transaction cost theory particularly useful in evaluating how SCCP strategies reduce opportunism, improve information sharing, and enhance overall performance.

The Kraljic Portfolio Matrix, developed by [26], is a strategic tool that helps organizations classify purchases based on two dimensions: supply risk and profit impact. Items are categorized into four quadrants—non-critical, leverage, bottleneck, and strategic—depending on their importance and the complexity of their supply markets. This approach allows managers to tailor procurement strategies to the level of risk and the potential effect on performance. [55] emphasize that the matrix advances continuity strategies by enabling organizations to identify critical suppliers and manage them proactively. [8], however, caution that the model may oversimplify reality, as it provides a static snapshot that does not capture all evolving risks.

[20] argues that globalization and technological advances increase the need to revisit and adapt Kraljic’s framework for continuity planning. According to [3], using the Kraljic model strengthens continuity management by prioritizing supplier relationships and aligning risk strategies with organizational goals. [4] adds that it enables managers to assess the suitability of adopting continuity practices across different organizational contexts. Overall, the Kraljic Matrix provides a structured way to incorporate SCCP into supplier management by focusing resources on high-risk, high-impact areas.

High Reliability Theory (HRT) emerged as a response to Perrow’s (1984) Normal Accident Theory, which argued that accidents are inevitable in tightly coupled, complex systems. [42] and [44] argue that while systems may be complex and interdependent, organizations can still prevent major accidents by adopting structures and practices that enhance reliability. Such practices include decentralized decision-making, redundant safety mechanisms, and continuous learning. [52] suggest that HRT provides a multidisciplinary framework that risk managers can use to build resilience. [27] further emphasizes that, unlike NAT’s pessimistic outlook, HRT assumes disruptions can be minimized through organizational design.

[46] argue that principles of high reliability (such as sensitivity to operations, commitment to resilience, and deference to expertise) can be transferred to supply chain contexts to guide the development of robust continuity strategies. [2] and [51] demonstrate that HRT can strengthen supply chain resilience by encouraging proactive detection and response to warning signals. [6] add that decentralization improves responsiveness, as decisions made closer to suppliers enhance agility during disruptions. [14] also highlights the usefulness of the Cynefin framework for categorizing crises and designing targeted responses (Figure 1). Thus, HRT underscores the importance of organizational structures and practices that not only anticipate disruptions but also ensure that retail supply chains remain reliable, adaptive, and sustainable under stress.

Figure 1.Conceptual framework.

Supply Chain Continuity Planning (SCCP) refers to a holistic process that identifies potential threats to an organization, assesses their likely impact on business operations, and develops frameworks that enhance organizational resilience and effective response capabilities to safeguard stakeholders, reputation, brand, and value-creating activities ([22]). SCCP enables organizations to prepare for, respond to, and recover from disruptions while ensuring continuity of critical business functions ([7]; [49]; [3]).

In this study, SCCP was conceptualized through three dimensions: improved organizational resilience, developed contingency arrangements, and reduced disruption downtime. Improved organizational resilience refers to the ability of retail chain stores to anticipate, absorb, adapt to, and recover from disruptions ([2]; [1]). Developed contingency arrangements refer to the existence of documented continuity plans, emergency procedures, backup systems, and recovery mechanisms designed to maintain business operations during disruptions ([7]; [4]). Reduced disruption downtime refers to the organization’s ability to restore critical operations within acceptable recovery periods following a disruptive event ([49]; [3]). These dimensions were measured using questionnaire items relating to continuity planning, preparedness for disruptions, contingency arrangements, organizational adaptability, and recovery effectiveness.

Performance was defined as the accomplishment of organizational objectives measured against predetermined standards of accuracy, completeness, cost, and speed ([13]). From a supply chain perspective, performance reflects the extent to which an organization achieves operational efficiency, customer satisfaction, profitability, and competitive advantage ([17]; [30]). The construct was operationalized using four dimensions: profitability, on-time delivery, cost efficiency, and sales turnover. Profitability refers to the organization’s ability to generate financial returns from its operations and was measured through items relating to gross profit margins and revenue growth ([41]; [23]).

On-time delivery refers to the ability of the organization to consistently fulfill customer orders within agreed timelines and was measured through delivery reliability and lead-time performance indicators ([17]; [50]). Cost efficiency relates to minimizing operational and supply chain costs while maintaining desired service levels and was measured through resource utilization and cost reduction indicators ([10]; [9]). Sales turnover reflects the volume and value of sales generated by the organization and was measured through growth in sales volume, market share, and sales revenue ([54]; [41]). Collectively, these dimensions provided a comprehensive measure of retail chain store performance and enabled assessment of the influence of SCCP on organizational outcomes.

Supply chain continuity planning (SCCP) is increasingly recognized as an essential practice for ensuring the resilience of organizations in the face of disruptions such as natural disasters, pandemics, cyberattacks, and geopolitical shocks. [5] define SCCP as the process of creating strategies to guarantee that supply chains remain functional during and after crises. Its primary goal is to reduce disruption impacts on operations, financial performance, and reputation. The key steps include identifying critical suppliers, assessing risks, formulating contingency plans, and continuously testing and improving these plans ([7]; [39]; [49]). [24] stress that proactive continuity planning provides firms with a competitive advantage by minimizing the negative effects of disruptions.

While some risks originate within firms, many are external. Increasing reliance on suppliers and partners means that disruptions in inbound logistics or supplier failures can ripple across the entire value chain, causing severe operational challenges ([38]). [19] developed a decision model emphasizing the secure location of critical inventory to improve disaster preparedness, while [21] underscored the need for structured handbooks to guide continuity management. Similarly, [18] demonstrated how analytical models can quantify SCCP’s effect on supply chain costs.

[3] point out that operational risks—such as technical failures, natural disasters, and deliberate attacks—require diverse mitigation measures depending on their impact. Awareness, prevention, and remediation are seen as the foundation of risk management ([56]; [29]). [48] highlight the usefulness of tools such as Failure Mode and Effect Analysis (FMEA) and Ishikawa cause-effect diagrams in proactively identifying and addressing potential weak points.

[16] explains crises as multi-stage events, starting with warning signals (prodromal), followed by acute and chronic phases, and culminating in recovery. [28] adds that effective SCCP depends on the ability to detect weak signals before crises escalate. [33] further demonstrate that SCCP creates competitive advantage by reducing response times and enhancing supply chain flexibility. Similarly, [1] conceptualize resilience as a multidimensional construct strengthened by continuity strategies.

Despite global recognition, empirical research on SCCP has been uneven. Most studies rely on case analyses, with limited use of survey methodologies to validate findings across diverse firms. [32] were among the first to operationalize SCCP constructs through surveys, while [40] emphasized learning from major global crises to enhance continuity frameworks. [36] and [35] found that firms adopting continuity planning improved accountability, governance, and profitability.

Within Africa, supply chain continuity research has highlighted both challenges and opportunities. [51] found that Ugandan firms improved resilience by investing in redundancy, collaboration, and flexibility. [43] emphasized procurement’s role in strengthening resilience through strategic sourcing practices. However, many African firms struggle with resource constraints and infrastructural limitations, which reduce their ability to operationalize continuity frameworks effectively.

In Kenya, several disruptions have exposed the fragility of retail supply chains. The collapse of Nakumatt, Uchumi, and Tuskys has been partly attributed to inadequate continuity strategies, weak governance, and liquidity crises. [34] argued that public institutions such as the Kenya Bureau of Standards also lacked continuity mechanisms, compromising service delivery. [45] demonstrated that private security firms in Nairobi improved efficiency and profitability by embedding business continuity management practices. More recently, retail chains like Naivas and Quickmart have shown resilience by integrating risk management and continuity planning into operations. These cases illustrate that SCCP is not only relevant but critical for the survival of retail businesses in Kenya’s turbulent environment.

Although significant literature exists globally, African and Kenyan contexts remain underexplored. Many prior studies have focused on manufacturing, logistics, or public sector organizations, leaving a gap in the retail sector. Moreover, while conceptual and case-based research dominates, few studies use large-scale survey data or advanced statistical modeling to test SCCP’s impact. This study addresses these gaps by applying a quantitative research approach, incorporating both primary and secondary data, and analyzing the effect of SCCP adoption on the performance of retail chain stores in Kenya.

This study adopted a longitudinal survey design, guided by a positivist research philosophy, to evaluate the effect of Supply Chain Continuity Planning (SCCP) on the performance of retail chain stores in Kenya. The longitudinal design was chosen because it enables data collection from multiple respondents over an extended period, enabling the assessment of trends, changes, and causal relationships among study variables between SCCP and performance. The study examined 17 retail chain stores in Kenya. From a target population of 23,031 employees, a sampling frame of 1730 managers, supervisors, and supply chain personnel was established. Using the Krejcie and Morgan (1970) sampling technique, a sample of 315 respondents was selected through a multistage sampling approach. In the first stage, retail chain stores were treated as clusters and respondents were stratified into three categories: managers, supervisors, and supply chain personnel. In the second stage, employees were selected from the respective clusters using probability proportional to size (PPS) to ensure adequate representation across the participating retail chain stores. A total of 284 usable questionnaires were obtained, yielding a response rate of 90.16% and providing sufficient data for the multilevel analysis.

Data were collected using a structured Likert-scale questionnaire that was pilot-tested with 33 respondents from selected Nairobi retail chain stores to enhance clarity, sequencing, and content validity. Secondary data were collected from the financial statements, operational reports, and internal records of 16 retail chain stores over a five-year period. Performance was measured using objective indicators such as gross profit margin, delivery time ratios, and order lead times, while SCCP was measured using aggregated survey scores on organizational resilience, contingency arrangements, preparedness, and recovery effectiveness. The aggregated SCCP scores were then regressed against the five-year performance data to assess their influence on organizational performance.

Descriptive statistics (means, standard deviations, and frequencies) summarized responses, while inferential statistics tested hypotheses. Given the nested structure of the data (employees nested within organizations), the study employed Multilevel Modeling (MLM) to estimate effects at both the individual (Level 1) and organizational (Level 2) levels. Descriptive statistics (means, standard deviations, and frequencies) summarized responses, while inferential statistics tested hypotheses. Given the nested structure of the data (employees nested within organizations), the study employed Multilevel Modeling (MLM) to estimate effects at both the individual (Level 1) and organizational (Level 2) levels.

Mixed-effects models were used to test the study hypotheses by examining both the fixed effects of Supply Chain Continuity Management strategies on performance and the random effects across retail chain stores. Model appropriateness was assessed using Intraclass Correlation Coefficients (ICC), Likelihood Ratio (LR) tests, and model fit indices (AIC and BIC), with the best-fitting model adopted for hypothesis testing based on the significance of the estimated coefficients ([15]).

This study examined the influence of Supply Chain Continuity Management strategies on the performance of retail chain stores in Kenya using a two-level hierarchical data structure. Level 1 comprised employee respondents who provided perceptions on the study variables, while Level 2 comprised the retail chain stores within which the employees were nested. The multilevel structure enabled the assessment of both employee-level effects and variations across the 17 retail chain stores included in the study.

Mixed-effects models were estimated using Maximum Likelihood Estimation (MLE). Model selection and comparison were conducted using Likelihood Ratio (LR) tests, Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC), which are appropriate for comparing nested multilevel models estimated using ML. Following model selection, bootstrapped standard errors were computed to improve the robustness of parameter estimates and statistical inference in the presence of minor deviations from normality. Consequently, all reported LR tests, AIC values, BIC values, and hypothesis tests are based on models estimated using Maximum Likelihood Estimation (MLE).

The random-intercept model at the employee level with 1 level 1 predictor would be specified at level 1 and level 2 in hierarchical form separately as;

The second equation can be substituted into the 1^st equation to yield a formulation given by;

In the equations:

Y i j is the performance as perceived by employee i nested for entity j;

X i j is the measurement of predictor variable X as perceived by employee i nested for entity j;

β 1 is the coefficient of X at level 1;

γ 0 j is the intercept which has a separate specification equation due to the 2 levels assumed to cause variation in performance. In the intercept equation;

β 0 j is the level 1 intercept which is the average performance for the entire population; and

μ 0 j is the entity specific effect (cluster specific) random intercept.

The independent predictor of the study is the Supply Chain Continuity planning. Strategies are however covariates that are expected to impact performance of the retail chain store thus variation in performance as influenced by predictors are not just due level 1 effects but due to variations of the independent variables as level 2 random effect covariates. The study thus adopted multilevel random coefficient modelling (MRCM) as the technique model’s random coefficient estimates as proposed by [31]. The correct specification would therefore include the covariates (independent variables) in the intercept equation as;

The second equation can be substituted into the 1^st equation to yield a formulation given by;

where;

Y i j Performance as perceived by employee i nested for entity j;

X i j Measurement of predictor variable X as perceived by employee i nested for entity j;

β 1 Coefficients (slope) of the predictors at level 1;

μ 1 j Cluster-specific random slope (coefficients) of the predictor;

γ 0 j intercept which a separate specification equation due to the 2 levels has assumed to cause variation in performance. In the intercept equation;

β 0 j Level 1 intercept which is the average performance for the entire population; and

μ 0 j The entity specific effect (cluster specific) random intercept.