Embracing supply chain digitalization and unphysicalization to enhance supply chain performance: a conceptual framework

3.1 Definition of research database and keywords

The research was carried out exclusively in Scopus. A single database was used to improve the rigor, clarity and replicability (Cammarano et al., 2022b; Paré et al., 2015). The main reasons for using Scopus are twofold. Firstly, it is a large multidisciplinary database covering published articles, books, book chapters and proceedings in the humanities and science disciplines (Bramer et al., 2016; Festa et al., 2018). Secondly, it is possible to search by specific research areas. This aspect makes the study possible when considering that the SLR results highlight the fragmented and complex literature concerning DTs in supply chain management. The areas considered in this study are detected from the articles selected through the SLR and published in international journals within the subject area “business, management and accounting, economics, econometrics and finance” in Q1 and Q2 quartiles according to Scimago Journal Ranking and ISI Web of Science.

The 87 technologies identified for the analysis, the 11 macro-technologies in which they are categorized and the relative search keywords used in the search string in the title and keyword fields are reported in Table 1.

Regarding the proposed taxonomy, current literature lacks a general and acknowledged classification of DTs even if there are several articles. However, these classifications refer to specific industries and areas of applications as well as refer to one or relatively few macro-technologies (Cañas et al., 2021; Ghobakhloo, 2020). Thus, the suggested taxonomy of DTs tries to summarize and merge available classifications with the categories of DTs emerging from the study of thousands of scientific articles carried out for this study. In Table 1 for each macro-technology, references to papers on SC management are reported to justify their relevance in the research field. For each technology, the query implemented on SCOPUS is the combination of the following elements: [SUBJAREA] AND [PUBYEAR] AND [KEYWORDS]. In particular, [SUBJAREA] refers to the query conditions to filter the results with the journal within the aforementioned subject areas, whereas [PUBYEAR] allows to consider only papers published from 2019 to 2022. As for [KEYWORDS], only papers disclosing at least one keyword within either the title or keywords fields are selected. For example, the following query is used for blockchain:

[SUBJAREA] AND [PUBYEAR] AND (TITLE (“distributed ledger”) OR KEY (“distributed ledger”) OR TITLE (“blockchain”) OR KEY (“blockchain”) OR TITLE (“cryptocurrency”) OR KEY (“cryptocurrency”) OR TITLE (“smart contract”) OR KEY (“smart contract”) OR TITLE (“non-fungible token”) OR KEY (“non-fungible token”) OR TITLE (“NFT”) OR KEY (“NFT”))

3.2 Criteria for identifying emerging practices

A content analysis was carried out by a group of management engineering experts involved in the study. The first step is to manually identify any emerging practices based on the use of DTs in the business world in each article. The term “emerging practice” indicates a new non-standard way of implementing a DT within a business process in order to achieve a specific positive impact on SC performance. Thus, emerging practices are investigated with this methodology, according to a multidimensional perspective and summarized as the combination of the implemented DT, the business process using it and the expected impact on the performance (Figure 1).

The experts read scientific articles and intercepted those that clearly reported how DTs could be used in companies in order to promote emerging practices and gain a business advantage. The search on Scopus provided 78,108 studies covering a time horizon ranging from January 2019 to May 2022. From this articles list, a sample of 21,488 articles, equally distributed in various technologies, were searched. However, 1,131 documents were not downloadable. Therefore, 20,357 articles were analyzed. The articles suitable for the classification are those in which it is clear how the technologies impact on SCs and in what business process they have been used. For the purposes of this study, 1,585 suitable documents associated with SC management and DTs were selected. The selected articles contain all the information necessary to identify any emerging practices, that is which technologies are applied, in which business processes and with which impacts on market and organizational performance. Since the articles are useful for the analysis only it is possible to find out an emerging practice exhaustively described and well-connected with the business process and the expected impact on SC performance, a large number of the papers were excluded. Literature on SC management is featured by a significant range of studies that are not eligible for this research, such as (1) literature reviews, (2) conceptual papers that focus the analysis on the value of DTs, (3) documents that discuss technical issues, (4) articles that refer to the public sector, (5) articles describing emerging practices outside the field of SC management and (6) documents that do not formally clarify in which specific business process the emerging practice can be implemented, as well as not empirically demonstrating the positive impact on SC performance.

3.3 Classification scheme

The classification scheme for emerging practices is necessary to classify articles in a standardized way with specific labels as well as to ensure a rigorous classification of the research results. When emerging practices were detected in the article, they were classified with specific labels: the technology used, the business process and the relative impact achieved. Only emerging practices associated with specific business processes for SC management were considered, as shown in Table 2. The business processes taken into account concern all the activities that engage SC actors such as suppliers, distributors, wholesalers, manufacturers, couriers, 3PLs and final customers. The classification takes into account the main areas of implementation of DTs reported in literature on SC management. References to prior literature were reported in Table 2.

In addition, this study aims to analyze the impacts that emerging practices, based on DTs, bring to SC management. In Table 3, the impacts have been divided according to whether they are market or organizational. The former refers to the impacts on external stakeholders, the latter to internal impacts. For each label, a definition of the impact nature is reported, as well as references to prior studies that have considered it important for SC management.

The manual classification of an emerging practice requires the attribution of three labels-technology, business process and impact-in order to standardize in which contexts it is used with which impacts. These labels, manually detected by the team of experts, are then used as categorical variables to present the research results. In this way, an emerging practice can be synthetized with a tuple technology-process-impact. Moreover, the emerging practices in which the use of BD was considered were signalled with a specific flag.

An example of the manual content analysis performed by the team of experts is provided in Figure 2. A practice implementing IoT to capture Big Data about suppliers and support their evaluation and selection has been detected from the work of Gottge et al., 2020. The experts read the paper, captured an emerging practice and assigned labels to define the combination technology-process-impact. While the information about DT, business process and impact are mandatory, the flag big data is optional and depends on the presence of a formal reference to BD.

As previously reported, this study selected 1,585 suitable documents associated with SC management and DTs. Within these documents, one or more emerging practices have been found, that is one or more tuples technology-process-impact were recorded. Overall, the number of detected emerging practices, as well as tuples, is 5,060. In the next section, each tuple will be a statistical unit of analysis. It is necessary to highlight that in some cases the emerging practice has been associated with multiple DTs, different business processes and impacts. Therefore, to preserve the three-dimensional perspective, multiple tuples have been assigned to the same emerging practice to cover all the possible combinations technology-process-impact.

4.1 Descriptive statistics

This section will provide descriptive statistics on each categorical variable under investigation. By counting the number of tuples, it is possible to give an overview of which DTs, processes and impacts are the most investigated in literature. Table 4 illustrates the number of emerging practices classified for each technology. The most frequent practices related to SC management are associated with artificial intelligence, blockchain and the internet of Things with more than 500 practices. The number of emerging practices associated with immersive environments, open and crowd-based platforms and 3D printing is less than 300. The reason is that these DTs are less associated with the SC management. 3D printing is often associated with production activities, open and crowd-based platforms are linked to the concepts of innovation and crowd involvement and, finally, immersive environments are less consolidated technologies since they are highly innovative and disruptive. On the contrary, the literature on technologies based on traceability, authentication, certification and forecasting systems is rich of well-detailed examples, case studies and applications for SC management. The number of emerging practices that have combined the use of Big Data with other technologies is 1,012, about 20% of the total. Table 4 shows how the technologies closely connected with Big Data are those of data management such as artificial intelligence (25%), Internet of Things (32%) and computing (38%). Technologies such as robotics (2%), immersive environments (8%) and open and crowd-based platforms (7%) are less associated with the use of Big Data for SC management. For example, robotics is mainly related to material flows and less to information flows management.

Table 5 illustrates the distribution of emerging practices according to the business processes considered for SC management. Business processes such as operations control and operations planning cover more than 40% of the emerging practices. Both delivery and transportation play an important role in the distribution of emerging practices. This means that these technologies are often adopted in logistics. The technologies are less applied in business processes such as 3PL 4PL Couriers-Outsourcing, distributors and wholesalers, raw materials management, supplier payment and post-sale service with less than 100 classified practices.

Tables 6 and 7 respectively show the emerging practices that have market and organizational impacts. On the one hand, the results show how often DTs are mainly used to achieve better customer satisfaction and better product/service quality and differentiation. On the contrary, DTs are used less to improve brand reputation, attract investors and receive incentives from governments. It should be considered that the emerging practices that aim to achieve market impacts are about 22% of the sample.

On the other hand, the results illustrate how emerging practices can allow for the achievement of different organizational impacts. DTs aim to improve, automate, optimize and streamline SC management processes and operations. As a result, automation, optimization and efficiency allow to reduce managing operations costs. In this study, the technologies presented improve the data and information management processes among the participants of the SC. Nevertheless, impacts such as innovation, knowledge and technology management as well as employee engagement have been less considered in the literature considering the use of DTs.

4.2 DTs, business processes and impacts

This section reports the three-dimensional analysis of emerging practices considering the relationships among DTs, business processes and impacts. Figure 3 shows a framework in which the impacts appear on the rows and the business processes in the columns. The most frequent technologies in the sample are shown at the intersection between business processes and impacts. A filter was applied in which only the technologies that have a number of emerging practices greater than 10 for the specific business process and impact are present. To obtain further insights from this study, the market and organizational impacts have been separated and distinguished. For the purposes of this research, impacts such as attracting investors, competitive advantage, employee engagement and government incentives have been eliminated since for these impacts the emerging practices present are lower than 10. The emerging practices of the following business processes for SC management such as raw materials management, 3PL 4PL couriers – outsourcing, post-sales service and reverse logistics were not reported because the number is less than ten emerging practices analyzed. Finally, the grey boxes in Figure 3 show emerging practices in which Big Data is more frequently used in conjunction with the other technologies mentioned.

Starting from the supplier evaluation and selection, buyer–supplier relationships, order management-purchasing, supplier payment and inventory, warehouse, transportation, distributors and wholesaler, there is not a minimum number of ten emerging practices, based on DTs, useful that can achieve specific market impacts. On the contrary, DTs for these specific business processes are mainly used to achieve organizational impacts. In particular, blockchain technology can be used in various business processes such as supplier evaluation and selection, buyer–supplier relationships, order management – purchasing and supplier payment to improve SC relationships management. Blockchain technology enables various features such as security, authentication and certification of shared data. The actors belonging to the SC can trust and collaborate with the other participants in the network since trust is guaranteed by the security and decentralization technology features. Therefore, the technology supports specific business processes that engage the interaction among two or more members of the SC who could exchange crucial data or information. For the supplier payment, the use of cryptocurrencies can lead to facilitated financial exchanges for the cross-border shipments.

For business processes such as buyer–supplier relationships and order management-purchasing, blockchain technology is promising for costs reduction. Specifically, the use of smart contracts and the elimination of intermediaries allows for a time reduction for the order management activities and consequently in a resource costs reduction. Smart contracts, if properly implemented, could automate different order management activities between several players in the SC and reduce the activities cost. Furthermore, the absence of external intermediaries could guarantee a further elimination of any additional expenses. The presence of huge transactions within the blockchain platform could promote the Big Data analysis and understand, for the order management-purchasing process, which suppliers are most promising and support a more detailed analysis for the raw materials procurement.

It is interesting to note that for the buyer–supplier relationships several technologies can be used to achieve better SC relationships management. In addition to the blockchain, other technologies such as artificial intelligence, digital applications, Internet of Things and proximity technologies have been considered associated with this impact. These technologies allow to collect data and information on SC relationships such as product orders and contacts with different players using Internet of Things, RFID, proximity sensors, record them on a shared platform (i.e. blockchain), carry out specific analysis for forecasting on data exchanges with the players in the SC using machine learning algorithms and, finally, communicating with the members of the SC through digital application (i.e. web application, social media and network, mobile application) in order to strengthen the relationships with certain actors. The widespread adoption of RFID technology provides supply chain managers “requires a higher level of automation of the decision-making processes” (Morenza-Cinos et al., 2019, p. 1020). In this context, Big Data are often analyzed with advanced predictive techniques to make better and more accurate forecasts.

Regarding inventory management, the technologies mainly used in this business process aim to achieve specific organizational impacts such as: costs reduction, efficiency and productivity, information management and SC relationships management. Most of the emerging practices are based on the use of artificial intelligence. Regression and metaheuristic algorithms are often used to improve demand forecasting and better manage the amount of inventory quantities. The use of metaheuristic algorithms leads to an optimization of the stock level by reducing the cost of stock in the warehouse. These algorithms are often used in combination with Big Data to make advanced forecasts for inventory management. In addition to costs reduction, the use of algorithmic techniques allows to improve the efficiency and productivity of the inventory. Based on the data collected, artificial intelligence can modify orders, guarantee the delivery of the requested goods to warehouses on time. Proximity technologies such as RFID, Bluetooth, NFC, QR code and datamatrix support aligning the physical inventory with the virtual inventory by collecting information in real time. In this way, it is possible to make decisions in real time, avoiding errors from manual and repetitive practices. These technologies not only guarantee complete traceability and transparency of the items in stock but also facilitate the monitoring of available quantities. To achieve a better level of SC relationship management, both artificial intelligence and Internet of Things technologies are applied in the inventory. The Internet of Things ensures that raw materials, semi-finished and finished products are tracked within the internet. The use of machine learning techniques leads to managing the relationships among partners better as well as ensuring an optimal level of service based on the inventory stock.

Business processes such as operations planning and operations control consider different technologies that aim to achieve market impacts. Artificial intelligence allows to improve customer satisfaction, product/service quality/value/differentiation and increase revenues for operations planning. Specifically, regression, deep learning and artificial neural network algorithms make it possible to predict market demand by analyzing market data, planning production and distribution in an optimized way. For example, association rules algorithms could identify the most common combination of purchases by customers and detect product features that can attract the consumer tasting. The operations planning carried out with these algorithmic techniques ensures an increase in product turnover by reducing the unfilled order due to the absence of the product on the shelf. Artificial intelligence techniques are often used to achieve other organizational impacts within operations planning such as: costs reduction, efficiency and productivity, energy efficiency, information management, risk reduction and time reduction. These impacts are closely linked to each other. For example a correct and optimized activities planning with metaheuristic algorithms improves the efficiency and productivity of operations, saves time and resource costs and, finally, achieves energy gains. Algorithmic classification techniques often make it possible to classify issues and highlight opportunities for improvement in risk reduction by making the SC more resilient. Regarding the information management impact, artificial intelligence and geospatial technologies are also used for the operations planning. For example, GIS and web mapping allow for traceability, monitoring and control operations to collect information on sites in order to plan specific decisions. The larger the information volume, the more important it is to consider techniques to process Big Data. From the analysis of this amount of data, hidden information useful for the optimization of operations planning could often be intercepted. To achieve a greater efficiency and productivity for the operations planning, various technologies can be used such as computing, DTs, geospatial tech, immersive environments and Internet of Things. Geospatial technologies guarantee data collection through technologies such as GPS, GIS and GNSS. The Internet of Things makes it possible for items, equipment and products to be connected and available on the internet by facilitating operations planning activities. The use of cloud computing promotes the collection of a large amount of data for planning. The web or mobile applications allow to manage the activities planning in a simple and fast way. Finally, immersive environments tools such as digital twin and virtual reality assist to streamlining operations planning activities by simulating their behavior on a virtual space before implementing it in the real world. These analyses can be performed on Big Data which can be considered crucial for resolving issues on the operations planning of SC.

One of the business processes in which there is a considerable number of technologies that guarantee the achievement of different impacts is the operations control. Regarding the market impacts, to achieve a greater customer satisfaction, several emerging practices use natural language processing algorithms in combination with Big Data to monitor the level of customer satisfaction by analyzing the contents of online reviews. Moreover, to improve the product/service quality/value/differentiation different technologies could be adopted to achieve this impact. Technologies used for product tracking such as RFID, GPS, GIS, IoT and Bluetooth can improve the information exchange on the entire SC. The information is visible to all the players in the network and therefore the product could reach a higher market value. The unexpected increase of this information makes it necessary to use technologies such as cloud storage and cloud computing in order to store these data and analyze them with predictive analytics techniques based on BD. Technologies such as artificial intelligence, computing, geospatial technologies and Internet of Things aim to reduce operations control activities costs. The automation developing with these technologies aims to reduce the workload of human resources. Technologies such as cloud computing reduce the cost of IT hardware infrastructure. Artificial intelligence techniques allow for cost savings on the predictive control of machine downtime, while also ensuring a suitable level of inventory. According to Scuotto et al. (2017), the presence of human resources with ICT skills is relevant for the application of ICT to supply chain management. The technologies considered aim at efficiency and productivity for operations control activities. However, the sample under analysis does not consider 3D printing technology within this business process to achieve this specific impact. The main purpose of these technologies in operations control activities is that the processes are carried out correctly. Many of these technologies are used for monitoring such as mobile applications, open-source software, sensors, RFID, Internet of Things, GIS and cloud computing. Other technologies assist in monitoring and guaranteeing the achievement of efficiency and productivity such as through the use of autonomous mobile robots, cobots, drones, virtual and augmented reality. In SCs, it is increasingly important to monitor the aspects of energy efficiency with a sustainability perspective. The most used technologies are artificial intelligence, computing and the internet of Things. The scope of these technologies is to track energy consumption and verify at a later stage which processes, operations and activities produce greater energy inefficiency. Tracking and data management technologies are crucial for improving the information management for operations control activities. Big Data is increasingly used in these detailed analyses. For the control and monitoring activities, it is necessary to know in real time the products shipped, the stock in the warehouse in order to solve several issues. Technologies such as blockchain and computing can provide greater visibility and transparency on various activities as well as improve control and monitoring. The impact of innovation, knowledge and technology management for operations control activities is achieved through artificial intelligence which can identify solutions to hidden issues. In a supply chain, “the effective use of knowledge management technology is critical to effectively plan and predict future operations” (Desouza et al., 2003, p. 130). To reduce risks, technologies such as artificial intelligence, geospatial technologies, Internet of Things and proximity technologies can trace different activities and avoid opportunistic behaviors by the actors in the SC network. Furthermore, machine learning algorithms applied to Big Data can improve the SC resilience by predicting potential effects produced by sudden catastrophic events such as the Covid-19 pandemic (Paul and Chowdhury, 2021). To monitor SC relationships, the technologies that can be used effectively are blockchain, Internet of Things and proximity technologies. These technologies can integrate and coordinate the actors in the SC and monitor the participants to have a correct behavior. Finally, artificial intelligence, Internet of Things and robotics technologies can reduce times. The use of robots reduces the time spent on checking and packing activities. In addition, the automation produced by the combined use of these technologies optimizes human resource times.

Robotics in warehouses improves various aspects such as cost reduction, efficiency and productivity, flexibility and time reduction. These impacts are correlated as cobots, drones and industrial robots reduce the time and human resources costs to carry out warehouse operations on material flows. Furthermore, greater efficiency and productivity are guaranteed, as the robots are programmable to carry out a significant number of warehouse activities without working breaks. These technologies guarantee greater flexibility in the warehouse operations management even in the presence of unexpected events. In this context, drones could help to improve the activities and operations related to the warehouse. These devices would not replace warehouse workers but would work with them to perform warehouse tasks more quickly and effectively.

Artificial intelligence techniques are often associated with achieving different impacts in transportation. Metaheuristic algorithms, deep learning and artificial neural networks can optimize the transport load and identify the most suitable routes in terms of cost and time reduction as well as efficiency and productivity. However, to achieve improvements in terms of information management and SC relationships management, other technologies are needed that allow for data collection and storage in a secure, shared and decentralized manner such as the internet of Things, geospatial technologies and blockchain. These technologies can monitor the shipments and products status in the trucks in real time. Thanks to artificial intelligence techniques, such as deep learning and artificial neural networks, it is possible to optimize transportation processes, prevent errors and detect potential issues and risk factors during the transportation. The use of machine learning techniques on Big Data allows for a better route planning.

Thanks to its intrinsic features of security, disintermediation, decentralization and encryption, blockchain technology allows for collaborative relationships of the actors such as distributors and wholesalers. The authentication and certification of transactions via blockchain allow to have trust in the technology, thus avoiding intermediaries.

Technologies such as artificial intelligence and robotics are widely used in delivery to achieve various impacts such as: customer satisfaction, cost reduction, efficiency and productivity, energy efficiency, risk reduction and time reduction. In particular, the use of unmanned vehicles can support the product shipment without the additional costs of human resources. Moreover, drones could be programmed with machine learning techniques to optimize routing and quickly reach their final destination. The use of Big Data is often needed to conduct and infer these predictions. These technologies can improve customer satisfaction in terms of perfect orders. The use of unmanned vehicles can allow for energy savings since they can avoid city traffic and reduce the number of risks related to accidents at work during the last mile delivery.

For sales and sales channel activities, artificial intelligence and digital applications are widely applied to improve customer satisfaction. New sales channels such as social media, mobile applications and web applications are essential in the e-commerce and m-commerce world. Managing data from these platforms allows to apply natural language processing techniques to online reviews as well as make improvements to the final product to achieve high customer satisfaction. These new sales channels have encouraged the growth of an omnichannel experience that allows the customer to buy products with a smartphone and to pick them up in a physical store.

Finally, digital applications are crucial in customer service activities to achieve better results in terms of brand reputation, customer satisfaction, revenues and efficiency and productivity. These technologies allow to get in touch with the customers and help them with the purchased products, increasing the brand reputation, customer satisfaction and revenues. The use of a mobile application support direct communication with the end customer, such as push notifications in the form of alerts or product recommendations based on each customer's search history. Augmented and virtual reality, metaverse, social robots and proximity technologies such as sensors on the shelves, can help the customer for the final product purchasing, better comprehension of their needs and ensure a suitable purchase. The metaverse fills the gap between e-commerce and physical commerce, transforming the user's shopping experience: from static product catalogues to real-time simulations of items such as clothing. As information on the web increases, it is increasingly necessary to manage Big Data in order to carry out an accurate analysis. Finally, chatbots and contact center automation can increase efficiency and productivity by reducing the workload performed by human beings.

5.1 Scientific implications

The proposed research methodology systemizes the current practice as well as the potential of digitalization and unphysicalization according to a conceptual framework of three dimensions: technology, process and impact. The SLR aims to algorithmically and objectively investigate the recent literature on SC digitalization and unphysicalization, standardizing the information incorporated into thousands of scientific documents. The study can be seen as offering progress in terms of reviewing the main opportunities for SC management, while also identifying the main areas in which research on DTs is currently being caried out. While some SLRs already exist, the authors undertake a unique analytical perspective, resulting in an original set of variables for further investigation in the area. This study represents an important evolution in organizing the current literature on SC digitalization and unphysicalization into a framework in order to delineate the state of the art on SC digitalization and unphysicalization and suggests promising opportunities for future inquiry. The output is a database of emerging practices that can be implemented within specific SC processes, promising the enhancement of SC performance. The findings of this work help incorporate knowledge about SC digitalization and unphysicalization into a three-dimensional framework that can be easily used for defining strategies and organizing SC according to the expected impact on SC performance. A theoretical implication is related to the knowledge management referred to the DIKW model (data, information, knowledge and wisdom) and more in detail on how the knowledge emerged from these emerging practices can be collected (information), selected and analyzed (data) to obtain (through technologies such as artificial intelligence) knowledge and reach the wisdom. Another possible theoretical implication can be related to the strategic management or the influence of the knowledge/wisdom (obtained from the emerging practices) on the strategic decision-making process.

Since these emerging practices are based on the use of DTs, the database suggests how SC digitalization and unphysicalization could affect SC management and performance. By focusing on the most successful opportunities of the implementation of DTs and the most recurring practices reported by scholars, this research provides a conceptual framework for the implementation of DTs to support SC digitalization and unphysicalization, as synthesized in Figure 3. The research demonstrates that depending on both the business process and impact on performance only a few technologies result in a consolidated stage of implementation, revealing probable developments of the practice over the next years and allowing to affirm that they will probably support the development of future best practices in specific business processes.

Moreover, the methodology proposed for the SLR supports the classification of scientific papers and can be replicated to help academics to further understand the technologies, exploring the trends in SC digitalization and unphysicalization, while also identifying the areas where the technologies are most desirable and can be implemented.

5.2 Research gaps and future directions

Although the literature review revealed that many aspects have been investigated about DTs and the role of SC digitalization and unphysicalization, some elements should be further analyzed, in order to both provide academics with a comprehensive view about the theme and support practitioners in taking decisions in the field. In consideration of the information collected from current literature, it is possible to provide a structured proposal for promising future research direction based on the review findings and enriched by insights from the quantitative analysis of the emerging practices. According to the analysis and discussion, this paper has identified the following five research gaps, that could be the basis for prescriptions and a broad agenda to guide future research and theory development.

5.3 Limits of the research

The limitations of the research regard the time-consuming activity of the manual content analysis of scientific papers and the subsequent process of classification of emerging practices performed by the experts. The information about emerging practices could be dispersed within the entire document, so that it is impossible to rely on automated systems to classify emerging practices. Furthermore, to classify emerging practices, an effort has to be made to convert the information into the proposed taxonomies. This limits the implementation of the methodology, which requires the analysis of a huge amount of documents to be able to obtain patterns and insights regarding the most promising areas of implementation of DTs. Additionally, gray literature was not considered, with several applications and case studies not having been considered. Similarly, many pioneering applications of DTs within SCs could have been implemented by companies but not reported by scientific literature.

Despite the aforementioned limitations, literature is widely recognized as the most reliable reference for forecasting the future development and implementation of cutting-edge technologies. Therefore, since literature anticipates business applications, the proposed conceptual framework could support both scientists and practitioners in exploring the usefulness of emerging practices in SCs.

5.4 Managerial implications