In this research methodology, we conducted a research paper review during the preparatory stage using the following primary terms: “drivers for supply chain sustainability,” “critical success factors of sustainable supply chain adoption,” “critical success factors in sustainable supply chain,” and “Studied and analyzed the factors that influence a sustainable supply chain”. This process has started with the finding of critical success factors and ended with the ranking of critical success factors, as shown in Figure 1. Reviewing the academic and corporate critical success factors led to the identification of 15 critical success factors, which are listed in Figure 2. By taking into account the opinions of both academics and experts in company management, critical success factors have been modified for application in the Bangladeshi food industry.

After collecting expert opinions from the different food industries in Bangladesh, the critical success factors prioritized the basis of individual weight and used Smart Interpretative structural modeling software. The interpretative structural modeling is a systemic tool; this tool can be used to discover contextual links and it is also a structural method that derives its foundation from the intricate web of connections. In a digraph model, the individual connections and overall structure are shown. It helps to impose direction and order on complex relationships, which is beneficial.

Step I:

Path of the partnership between two critical success factors (e.g., I and j) was figured out using the following four symbols:

• V-CSF: I facilitates j.

• A-CSF: j facilitates i.

• X-CSF: i and j facilitate each other.

• O-CSF: i and j are not connected.

We created an SSIM for SSCM practice implementation across the Bangladeshi food sector using these four symbols.

Step II:

Creating the Reachability Matrix: Create an initial reachability matrix by transforming the SSIM. There are two entries in the binary matrix: 0 and 1. It also abides by the following guidelines:

Rule1: For every V, 1 for (i, j) and 0 for (j, i)

Rule2: For every A, 0 for (i, j) and 1 for (j, i)

Rule3: For every X, 1 for (i, j) and 1 for (j, i)

Rule4: For every O, 0 for (i, j) and 0 for (j, i)

As a result of adhering to these guidelines, we were able to develop both a preliminary reachability matrix (Table 2) and a decision matrix (Table 3), which take into account the transitive interactions between the various considerations. Table 3 displays the results of the analyses conducted on the driving power and dependence power of each factor [31] [32] . The driving power, which is determined by adding up all of the consecutive components, is the force of one variable that influences another. Dependence is the total of all the values in Table 3 column and represents the extent to which the variable and the other variable are impacted.

Step III:

Cross-impact matrix multiplication (MICMAC) Analysis: The Cross-impact matrix multiplication method effectively correlates to the cross-impact matrix multiplication that is used for classification. The ability of measurements to be multiplied was the inspiration for the development of MICMAC. In this article, using it, we were able to evaluate the CSFs in terms of their execution and have a better understanding of the ISM-based model. It entails classifying the CSs that have been discovered according to their level of dependency and driving strength, both of which were assessed using the final reachability matrix [33] . MICMAC includes a graphic depiction of the factors based on their driving and reliance power in four clusters: autonomous, dependent, linkage, and independent. MICMAC also includes an analysis of the relationships between the elements.

Both the reliance power and the driving power of autonomous factors are rather low. Dependent variables have significant dependency power but poor driving power. Connective elements have a powerful reliance power as well as a powerful driving power. The driving power of independent forces is high, but the dependency power they exert is low. It is an illustration of the classification of the CSFs into these four zones for the purpose of assessing the application of SSCM practices.

The Cross-impact matrix multiplication (MICMAC) model in Figure 3 proves that there is no autonomous CSF, demonstrating the applicability of each of the fifteen CSFs under consideration to this investigation. Found CSFs 3 and 11 (tax reliefs of certified companies, financial benefits, and safe and quality food) to be dependent and CSFs 4, 5, 9, and 14 (business ethics, well-defined metrics for sustainability tracking, contribution from profit and resource, and resource savings) to be independent. CSFs 1, 2, 6, 7, 8, 10, 12, 13, and 15 (Sustainable Procurement Policy) Training and capacity building, enhancement of the company image, collaboration with multitier suppliers, good return on investment, business-to-business pressure, availability of information and transparency, elimination of duplication, and quality variation are all linked variables [33] [34] .

Step IV:

Level Partitioning: An extracted each component’s reachability and antecedent sets from the finished reachability matrix. The component itself, as well as any other variables that it may help enable, are included in the reachability set. The element in question, as well as any other potential variables that have an impact on it, make up the antecedent set. Calculating the intersection of these two sets for each CSFs is another one of the tasks. When the reachability and intersection set of a CSFs are identical to one another, that CSFs is said to be at level 1 [34] . After the first round, we threw away the level-1 CSFs and proceeded with the same technique until all of the levels of the CSFs were found. In this particular scenario, Safe and Quality Food is located at level 1 and constitutes the very pinnacle of the ISM hierarchy. The ISM hierarchy concludes with business ethics, which occupies the lowest level possible, as shown in Table 4.

Step V:

Diving Reachability Matrix: Dividing up the reachability matrix into its several layers the process of processing the reachability matrix results in the formation of several tiers of the structural model. It consists of a multitude of subsidiary stages. A reachability set, also known as Set R, is created for each variable or factor based on the rows of the matrix. This set is constructed in such a way that it has all the column numbers that indicate the factor stated in the affected counterpart of the system. Alternatively, each variable or factor also has an antecedent set, or Set A, which is generated using the matrix’s columns. This set is constructed in such a way that it contains the entire row numbers that indicate the factor stated in the influencing counterpart of the system. These influential and affected parameters are read for the cell values that are equal to 1. Once again, the combination of set R and set A may be found by using the equation (Set R) (Set A) = Set RA = Set C as shown in Table 5. A comparison is made between Set C and Set A for each variable, and the variables for which the compared sets are found to be identical are removed from consideration. The variables are taken out of the equation as a result of this elimination. The results of this elimination are then applied to each of the added sets of variables. Performing this action will cause all of the remaining variables’ reachability sets and antecedent sets to be updated. The procedure is carried out once again during the subsequent iteration to get rid of all the variables. All of the elements that were eliminated during a given iteration are stacked up to generate a collection of iterations after each one. Each iteration represents a level of hierarchical structure in the overall model, and the variables in that iteration are partitioned in a manner that corresponds to that level.

Step VI:

By rearranging each variable in the reachability matrix according to the level it occupies, one may create a canonical representation of the removal as shown in Table 6 and categorization of variables into distinct levels described before. The outcome is a matrix that is the same as either the lower triangular matrix or its transpose is represented in Figure 4.

This diagram (Figure 4) is transformed into the Interpretative structural modeling (ISM)-based model. The Interpretative structural modeling (ISM) hierarchy model also ensures that no issue at a given level will assist any challenge at a higher level.

Step VII:

As previously discussed, converting a matrix into row canonical form, sometimes referred to as row reduced echelon form, “reduced row-echelon” form, or Gauss-Jordan form or reduced conical matrix as shown in Table 7. This is done by applying the Gaussian elimination approach to transform the matrix into echelon form.

The issue of sustainability is now of international importance. Food manufacturers in Bangladesh need to use SSCM techniques if they want to compete successfully in both the local and global markets. Since certain foreign suppliers and clients of Bangladesh’s food manufacturers and process industries have started implementing sustainability measures, this will undoubtedly lead to the wider spread of SSCM practices across Bangladesh’s food sector. However, businesses will be unable to pursue a sustainable supply chain in the right way unless they first get a grasp of the critical success factors. This research set up fifteen CSFs based on the judgments of specialists. The fifteen critical success factors were then used as inputs to an interpretative structural modeling (ISM)-based model, and subsequently, the Cross-impact matrix multiplication (MICMAC) analysis was used to make predictions. With sustainable assistance, the most crucial CSFs were isolated and prioritized. By using Interpretative structural modeling (ISM), we were able to identify the interrelationships between these critical success factors. Finally, managers may benefit from MICMAC at the operational and strategic levels by taking the right measures. The following are some of our most important discoveries:

Technique for Order of Preference by Similarity to Idea Solution (TOPSIS) is a multi-criterion decision-making (MCDM) ranking approach that selects, from a limited pool of candidates, the one that is closest to the ideal answer and farthest from the negative ideal solution [35] [36] . The best criteria and sub-criterion values (A+) make up the positive ideal solution, while the worst attribute values (A) characterize the worst-case situation. To rapidly determine the optimal solution, TOPSIS computes a proximity coefficient for each option. In its original form, the technique relied on the decision-makers’ own subjectively assigned weights to various criteria. We also need to rank the success factors after assigning weights to them. Use the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) technique as a result. Here, the industrial specialists determined the weights of the various criteria and elements using established methods. And after the closeness coefficient has been established, we ascertained the rank that the critical success factors evaluate.

Here, business ethics is ranked first, well-defined metrics for evaluating sustainability are ranked second, but two (02) factors—resource savings and contribution from profit and resources—have occurred to rank third. Which specialists have chosen resource savings from?

Business ethics, Well-defined Metrics for Sustainability tracking and Resource savings are the top-ranked CSFs. In Bangladesh, there are fewer ethical issues and work, as it is an emerging economy. As a result, there is limited engagement with eco-friendly technology because there are less well-defined metrics.

Here, business ethics holds the 1st position, while the well-defined metrics for sustainability tracking hold the 2nd position, but for the 3rd position, two (02) factors—revenue savings, contribution from profit, and resource—have taken place. From which specialists have chosen resource savings. Business ethics, well-defined metrics for sustainability tracking, and The highest priority critical success factors are resource savings. Because Bangladesh is a developing economy, there are fewer ethical problems at work. Consequently, there is little use of environmentally beneficial technologies because there are less well-defined metrics. Based on research, managers will be better able to narrow in on the most crucial elements of a sustainable supply chain. They will be able to narrow in on the specific areas where they need to concentrate most to successfully implement SSCM. Thus, the findings will contribute to the paradigm shift toward sustainability now taking place in Bangladesh’s food business. Strategic considerations should be given to those variables with strong driving power, while those with high reliance power should focus on performance and outcomes. If the independent elements are constantly enhanced, businesses will see improved performance. By specializing in the critical success factors, businesses may get an edge in the market by producing goods with fewer negative impacts on the environment. To stay competitive, other businesses will be under pressure from consumers and other interested parties to embrace these same tactics.

Critical success factors have been identified and ranked using a variety of techniques. The technique for order of preference by similarity to ideal solution (TOPSIS) is a quantitative multi-criteria decision-making procedure that applies a full and complete set of information on criteria.

Managers may easily put TOPSIS to use since it does not need much specialized training or understanding. The TOPSIS method investigates each potential outcome and selects the one that is least distant from the ideal answer while still being preferable [37] .

Analyzing the ISM-MICMAC, the best 10 critical success factors are being collected and established into the TOPSIS method. In the TOPSIS method, the criteria are (environment, social impact, company profit) and the critical success factors are shown in Table 8. Corporate specialists provide these values. The values for this factor are to be taken as the standard unit’s box.

Step I:

Each value is normalized by being changed to where m is the dataset’s row count and n is its column count. The researchers are changing along the rows, while j changes along the column, as shown in Table 9.

Equation for Normalization:

X ¯ i j = X i j ∑ i = 1 n X i j 2 (1)

Step II:

Calculating the Euclidean distance for each row from the ideal worst and ideal best values, as well as the ideal best and worst values First, figure out the ideal best and ideal worst values: Now that this has happened, we must determine if there has been a (+) or (−) impact. If a column has a (+) effect, the ideal best value for that column is its maximum value, and its (−) impact is its ideal worst value.

For components in all rows from the ideal best to the ideal worst, we now need to determine their Euclidean distance.

Calculate the Euclidean distance from the ideal best.

S i + = [ ∑ j = 1 m ( V i j − V j + ) 2 ] 0.5 (2)

Calculate the Euclidean distance from the ideal worst.

S i − = [ ∑ j = 1 m ( V i j − V j − ) 2 ] 0.5 (3)

Step III:

Technique for Order of Preference by Similarity to Idea Solution (TOPSIS) score and ranking calculation by using the TOPSIS score for each row based on the distance positive and distance negative that are now available.

Calculate TOPSIS Score.

P i = S i − S i + + S i − (4)

Now order is based on the TOPSIS score, where the greater the score, the higher the rank dataset will be ranked as shown in Table 10.

The current research identified key success elements while implementing sustainable food supply chain management by studying the literature, speaking with industry professionals, polling them, and establishing a bi-level criterion hierarchy. Sustainability is the foundational goal of the hierarchy with environmental, social, and company profit.

Finally, the MCDM experiment was conducted to determine 15 important success factors in the ruling elite of a sustainable food supply chain. The experts were requested to provide responses to the inquiries. The ISM-MICMAC and TOPSIS methods were used in conjunction with expert opinions. Since these methods might yield different ranks, the results revealed which factor is more important. Resource savings and safe and quality food were found to be the most important environmental factors.

Moreover, sustainable procurement policy, training and capacity building, tax reliefs for certified companies and financial benefits, enhancement of the company image, and business-to-business pressure were verified as the most crucial social impact factors.

Business ethics, well-defined metrics for sustainability tracking, collaboration with multitier suppliers, good return on investment, contribution from profit and resource, availability of information and transparency, elimination of duplication, and quality variation were the most important company profit factors.

According to the hierarchy, business ethics, well-defined metrics for sustainability tracking, and contribution from profit, resource, and resource savings had the largest weights among those factors.

Found six levels from the Level Partitioning Iterations Table 10, with CSFs 4 at level 1, CSFs 9 at level 2, CSFs 5 and 14 at level 3, CSF 15, 13, 12, 10, 8, 7, 6, 2, 1 at level 4, CSF 3 at level 5, and CSF 11 at level 6. After an ISM-based model by first placing each CSF at its appropriate level and then using the Final Reachability Matrix (FRM) table to illustrate the relationships between the CSFs. Additionally, six levels were determined using the Level Partitioning Iterations Table 10, with CSFs 4 at level 1 and CSFs 9 at level 2.

In practice, MICMAC is equivalent to the application of cross-impact matrix multiplication to the classification process. The ability of measurements to be multiplied was the inspiration for the development of MICMAC. In this study, we made use of it to get deeper insight into the ISM-based model we were working with as well as to assess the CSFs from an implementation point of view. It entails categorizing the CSFs that have been discovered according to their level of reliance and driving power, both of which were set up in the final matrix. At level 2, CSFs 5 and 14 at level 3, CSF 15, 13, 12, 10, 8, 7, 6, 2, 1 at level 4, CSF 3 at level 5, and CSF 11 at level 6. After placing all of the CSFs at their level, we depicted a relationship between the CSFs as per the Final Reachability Matrix (FRM) table, resulting in an ISM-based model. MICMAC involves creating a graphical depiction of the components depending on their level of driving and reliance power. These factors are then grouped into one of four categories: autonomous, dependent, linkage, and independent.

Both the reliance power and the driving power of autonomous factors are rather low. Dependent variables have significant dependency power but poor driving power. Connective elements have a powerful reliance power as well as a powerful driving power. Independent variables have significant driving power but minimal dependency power. The MICMAC picture depicts the classification of the CSFs into these four zones so that SSCM techniques may be evaluated and implemented.

The MICMAC Figure reveals that there is no independent CSF, which proves that all fifteen of the CSs that were taken into consideration are applicable in this research. CSFs 3 and 11 (tax reliefs for certified companies, financial advantages, and safe and quality food) were determined to be reliant on one another. CSFs 4, 5, 9, and 14 (business ethics, well-defined metrics for sustainability monitoring, contribution from profit and resource, and resource savings) must be independent of one another.

And the connecting factors are CSFs 1, 2, 6, 7, 8, 10, 12, 13, and 15 (sustainable procurement policy, training and capacity building, improving the company image, working with multitier suppliers, good return on investment, business-to-business pressure, transparency and information availability, removing duplication, and quality variation). Since they pave the path for the efficient and successful implementation of linking and dependent SFs, independent CSFs are viewed as having a high priority for adoption in SSCM procedures. As a result, implementing independent CSFs should be given high importance.

In Sustainable Development Goals (SDGs):

We found the following similarities between our factors and the sustainable development goals (SDGs).

1 = No Poverty

2 = Zero Hunger

8 = Decent Work and Economic Growth

9 = Industry, Innovation and Infrastructure

12 = Responsible Consumption and Production

15 = Life on Land

If the mentioned success factors are implemented, the sustainable development goals as shown in Figure 6 will be achieved because this research work is theoretically and practically based on the sustainable food supply chain. A critical success factor plays a crucial role in the sustainable development of the food supply chain management, especially in No Poverty (SDG 1), Zero Hunger (SDG 2), Decent work and Economic Growth (SDG8), Industry, Innovation and

Infrastructure (SDG 9), Responsible Consumption and Production (SDG 12), and Life on land (15) [38] [39] . And the main theme of sustainable development goals is the elimination of poverty in all its forms everywhere. Achieving food security, improving nutrition, and promoting sustainable agriculture promoting inclusive and sustainable economic growth, employment, and decent work for all ensuring sustainable consumption and production patterns. Protecting, restoring, and promoting sustainable use of terrestrial ecosystems, managing forests sustainably, combating desertification, and halting and reversing land degradation and biodiversity loss secure a prosperous future, harness economic growth and development, and foster social solidarity across the sustainability of foods and nutrition.