Transferring lean knowledge within multinational networks

Abstract

Transferring lean knowledge is crucial for multinational companies; however, it is very difficult. Nevertheless, our knowledge in the area is limited. This paper investigates (a) how lean knowledge is transferred within multinational networks and (b) the type of organizational context that supports it. We use four case studies with 18 interviews in our exploratory study. Our results indicate that companies use a sophisticated knowledge transfer system with a fine web of practices that serve to transfer both explicit and tacit knowledge. The revealed organizational contexts show different patterns of division of labour (and responsibilities) between a central lean office and the particular subsidiary’s lean office. Our findings highlight the importance of transfer coalition, which consists of the cooperation of lean experts and top management during the knowledge transfer process. Top managers play a key role with their weak inter-unit ties, and they provide guidance for lean efforts and allocate sufficient resources to them. Further findings show that the evolution of lean transitions is not linear, and each of our case companies has witnessed turnaround during their journey.

Keywords:

• Knowledge transfer
• lean production
• multinational company

1. Introduction

Since the end of the 90s, respected researchers have discussed lean as a new paradigm (Schmenner and Swink 1998), and many others have highlighted its contribution to performance improvement (Hines, Holweg, and Rich 2004; Huo, Gu, and Prajogo 2016; Demeter and Losonci 2013; Netland 2013; Negrão, Godinho Filho, and Marodin 2017; Shah and Ward 2003). However, lean related studies usually focus on the plant level. The interests of the operations management community towards the ‘modus operandi’ of corporate multi-plant improvement programmes (MPIPs), pervaded by lean principles and tools (Netland 2013), have recently gained momentum (Netland and Aspelund 2014; Boscari, Danese, and Romano 2016).

Such MPIPs lead to improving individual plant performance. In addition, coordinating and managing the transfer of lean related knowledge in MPIPs can considerably improve efficiency even at the network level. Consequently, transferring lean knowledge in the internal network has become an important strategic issue.

Although the exchange of knowledge and learning within a network is one of the strategic capabilities of multinational companies (Shi and Gregory 1998), little is known about how multinational organizations acquire, develop and transfer lean related knowledge within MPIPs (Boscari, Danese, and Romano 2016; Inkpen 2005; Maritan and Brush 2003; Secchi and Camuffo 2016). Scholars argue that difficulties are encountered with regard to both the effective coordination and selection of appropriate knowledge transfer (KT) practices. Hence, units fail to improve performance or require many more resources and time than planned (years, decades) to deploy a lean system (Inkpen 2005; Maritan and Brush 2003).

In general, KT is a well-developed area (Argote 2012), and we have a good perspective on the transfer of internal best practices (O’Dell and Grayson 2004; O’Dell et al. 2003). However, the integration of this body of research into a more specific context, such as lean production (LP), is superficial but crucial, for at least three reasons. (1) LP has become the new way of creating value all over countries and industries, therefore this specific context is very frequent, and worth particular attention. (2) formalized and central lean systems provide a good field to apply existing KT theories, especially because these systems are already in the phase of supporting primarily knowledge diffusion, as opposed to knowledge creation. (3) LP is not a simple best practice, but it has generally been interpreted as a strategic organizational practice (Kostova 1999) that requires changes in values and mindsets. Not only knowledge but values should also be transferred, which require cultural change (Losonci et al. 2017). Therefore, transferring lean related knowledge is more complex than a ‘pure’ technology transfer, usually the focus of KT researches.

Altogether, the recent strategic importance of multi-plant lean programmes in multinational corporations, the difficulties of their effective deployment at the unit and network levels and the value orientation of lean systems highlight the necessity for an in-depth empirical investigation in the specific context of lean.

Our descriptive case-based study explores the KT practices that are related to LP (in this paper, we refer to it as lean KT) in the internal networks of four multinational companies (MNCs). We collect and analyse lean KT practices. Additionally, we explore the supporting organizational context behind the KT system. A better understanding of the phenomenon can help managers to develop a more consistent policy that intensifies lean-related KT and hereby, improve performance with less effort.

After the introduction, the literature is summarized, and two research questions are elaborated. Then, we describe the data collection, methodology and case companies. Next, the similarities and differences of cases are identified and discussed. Finally, we present our conclusions.

2. Literature review

In this chapter, we present the relevant literature on KT between organizations, and we devote a separate subchapter to a discussion of the available literature on the field of transferring lean knowledge.

2.1. Knowledge transfer

‘Everything that describes a firm (…) operating rules, manufacturing technologies, and customer data banks (…) becomes an aspect of its knowledge’, and for an MNC the key competitive issue is how to efficiently create and transfer this knowledge (Kogut and Zander 1992, 384.). By knowledge transfer, we mean a process ‘when knowledge acquired in one organization affects another’ (Argote 2012, 148), in other words, ‘when knowledge is developed from the experience of another unit’ (Argote and Miron-Spektor 2011, 1128). There has been a great deal of research that has discussed the components of KT, such as the (nature of the) knowledge to be transferred, the characteristics of the sending and receiving unit, the transfer practices, mechanisms, methods and channels (hereafter uniformly referred to as transfer practices), and the contextual factors that include the relationship, the inter-organizational dynamics of the units (Easterby-Smith, Lyles, and Tsang 2008; Waehrens, Cheng, and Madsen 2012; Pawar and Rogers 2014) or the location of units (Szász, Scherrer, and Deflorin 2016).

The nature of knowledge, especially its ambiguity, tacitness and complexity (Easterby-Smith, Lyles, and Tsang 2008), has a great impact on KT. Not knowing the real value of the transferred knowledge on the receiving side (called causal ambiguity) can reduce interest (Argote and Miron-Spektor 2011). Transferring tacit knowledge, which is embedded in people (Polányi 1966), is always more difficult than transferring explicit and codified knowledge (Ferdows 2006). It is even more difficult to transfer strategic organizational practices that are ‘dominant, critical, or crucial for achieving the strategic mission of the firm’ (Kostova 1999, 308). The key difference of strategic organizational practices from product- or technology-related knowledge is that not only are they knowledge-based, but they are meaning- and value-based, all of the facets of which must also be transferred. Today, we already know that lean production is not only about the knowledge of the lean toolbox, but it is much more complex and requires a new mindset from employees and leaders (Womack and Jones 2003; Liker and Convis 2012). Thus, we consider it to be a strategic organizational practice.

The motivation and knowledge of the sending and the receiving units also influence the success of KT (Argote 2012; Ernst and Kim 2002). While on the sending side, KT capability and motivation for transfer are the critical factors, on the receiver side, absorptive capacity, that is, the capability to exploit external knowledge (Cohen and Levinthal 1990), and the motivation to learn are the most crucial (Easterby-Smith, Lyles, and Tsang 2008).

There are several KT practices that have demonstrated varying degrees of success (Argote 2012; Alexander and Childe 2013). Knowledge can go from one organization to another through moving personnel (which provides opportunities for observation and communication, or through the transfer of experienced personnel to recipient organizations), technology, templates, routines, or through social networks and alliances. Moving personnel can efficiently facilitate the transfer of knowledge (Rosenkopf and Almeida 2003) and can considerably improve the strategic role of a subsidiary (Vereecke, Dierdonck, and De Meyer 2006). The transfer of knowledge that relies on information technology can serve exploitation (Kane and Alavi 2007). For example, open knowledge repositories in the form of organized databases that are publicly available facilitate exploitation by sharing and creating standard knowledge. Jensen and Szulanski (2007) have shown that templates (defined as ‘an organizational practice that is currently in existence, observable, composed of a single or connected set of processes and consciously used in the replication process’, [1727]) are also adequate to transfer knowledge between organizations.

There are more complex approaches that involve many of the basic transfer practices that are mentioned above. O’Dell and Grayson (1998, 2004) identified four practices that help in KT: (1) benchmarking teams that are responsible for identifying best practices, (2) best-practice teams for KT that consist of managers and/or experts getting together regularly in face-to-face meetings and through electronic channels, (3) knowledge and practice networks that are organized from the bottom up by technical and functional personnel, (4) internal assessment and audits that range from technical assessment to award programmes. They also differentiate between technology (main focus on database and directory, central staff for facilitation) and team approaches (main focus on teams and people network at unit level).

It is widely acknowledged, that there are many contextual factors that influence KT. Technological and geographical distance makes the search for and adaptation of knowledge less efficient and more expensive (Rosenkopf and Almeida 2003). Information technology can support KT; however, it cannot serve as a substitute for it. Information should be clearly classified. Furthermore, one individual should enter the obtained data, and others should be motivated to read it (O’Dell and Grayson 1998). Both require adequate culture. The institutional distance of the cooperating units (i.e. national culture) can make communication difficult. The organizational culture of a recipient unit and the attitudes of its members – especially (a) the experts who are the key holders of knowledge, (b) the managers who are responsible for the changes, and (c) their cooperation, which is referred to as a transfer coalition (Kostova 1999), affect the depth and sustainability of KT. Power relationships, trust and risk, organizational structures and social ties among the parties also influence inter-organizational dynamics (Easterby-Smith, Lyles, and Tsang 2008). The competitive relationship of such parties may lead to opportunistic behaviour (Dyer and Nobeoka 2000). Finally, while in operations management both the content of KT practice and the sector act as contextual factors, in general KT-focused studies, neither a specific sector is highlighted (very different organizations are under scrutiny) nor is a specific best practice selected (O’Dell and Grayson 1998, 2004; O’Dell et al. 2003).

2.2. Transferring lean knowledge

Studies that describe multinational routines to transfer lean knowledge can be divided into two different streams: the first considers the external sources (buyers or suppliers) of lean knowledge (Boyle, Scherrer‐Rathje, and Stuart 2011; Dyer and Nobeoka 2000; MacDuffie and Helper 1997; Sako 2004), and the other is interested in KT within internal networks (Henriksen and Rolstadås 2010; Inkpen 2005; Maritan and Brush 2003; Netland 2014; Secchi and Camuffo 2016). In this paper, we focus on the latter.

The available studies about lean-related KT with a multinational network focus are very limited. Each of them relies on manufacturing firm case studies. However, they significantly vary considering the number of case companies (from 1 to 7), and their depth in relation to our two focus points, namely practices and organizational context. See the following paragraphs and the structured overview of relevant studies in Table 1.

Table 1. Key finding(s) in the literature on lean knowledge transfer within MNCs.

Author(s)		Inkpen (2005)	Maritan and Brush (2003)	Netland (2014)	Secchi and Camuffo (2016)	Henriksen and Rolstadås (2010)	Boscari, Danese, and Romano (2016)
Methodology		Case study	Embedded case study	Multiple-case	Multiple case study	Case study	Case study
Home country (countries)		US (headquarter)	US	Four Scandinavian multinationals	Italy	Norway	Italy
Unit of analysis		GM and NUMMI	Four US subsidiaries	Elkem, Jotun, Volvo, Hydro	Seven multinational plants operating in Italy	Norwegian supplier’s network	Italian company, US and Chinese subsidiaries
Data		Interviews, visits	Interviews with employees from different organizational levels	Workshop, visits, master theses, discussions, internal documents	Interviews, visits, master thesis	Open source, strategic documents, interviews with key personnel	Retrospective data collection, 17 months real time tracking, interviews with key managers and employees
Elements of organizational context		Central team: “a team of people who understood lean production” (p. 449)	Central consulting team, plant-level team, template factory	Central office (facilitating sharing), regional lean coordinator, lean coordinator at the plant, local/plant-level lean teams	Degree of HQ’s and plant’s involvement in lean roll-out, resource allocation, production system design varies significantly	Experts at the mother plants and at the satellites (centre of excellence)	Change team is in charge for transferring lean to subsidiaries, corporate lean office, lean coordinator, involvement of subsidiaries’ managers
Prac-tice(s)	Manual	–	–	–	- Manuals - Adoption sequence of lean practices (principles-based or detailed description)	–	- Intranet - Manuals - Phone calls, e-mails
	Training	–	–	- Academy with courses and qualifications, training programme	- Training	–	- Training
	Visit	- Visits	–	–	–	–	- Short and long visits, plant visits
	Audit	–	–	- (Annual) audit	- Audit	–	- Audits
	Workshop, study teams, projects	- Workshops - Study teams	–	–	–	–	- On site projects - Study and change teams
	Assignments, rotation	- Long and short assignments to the partner - Executive in residence	- Rotation of experts and lean manager among units	- Rotating people among plants	–	–	–
Key finding(s)		- Learning system instead of exposing individuals to new knowledge - Proactively managing kt - Spectacular results at green field plants - Social interactions are important - Greater efforts was initiated when a former programme member became the ceo of gm	- Both central consulting team and a particular unit can shape lean knowledge - Rotation of a lean manager from a successful unit to another unit facilitates progress	- Transfer creates a common improvement language - Shared platforms for all plants or separate systems for divisions - Plants do not need to develop a system - Global/divisional/regional resources to maintain improvement system	- Lean roll-out process depends on the codification of lean knowledge, on the autonomy of the unit and on the contextual ambidexterity	- Use of ict solutions - No coherent policy of lean km - Two way flows of knowledge between mother plant and satellites - Externalization if the most challenging step - Incremental innovation: related knowledge is spatially sticky, so spatial proximity is important	- Effective supporting practices change over time - International teamwork is a main practice - Dedicated lean team at hq to coordinate efforts

2.3.1. Practices of lean knowledge transfer

Our focus is on the building blocks of a coherent lean KT policy that is built on a set of carefully managed KT practices, and that offers a pattern on the interplay among practices, internal stakeholders and types of knowledge.

While every author dealing with lean KT considers knowledge transfer practices, the number and the list of proposed practices differs. Based on the common findings of earlier studies (see Table 1 Practices), we argue that training programmes (training material, qualification, study groups) and manuals, visits, audits and assessment systems, short and/or long assignments and rotation might be the key practices of lean KT. Unfortunately, based on this ‘short list’ the standardized processes (Inkpen 2005) boosting KT cannot be elaborated.

A coherent policy must also consider the role of internal stakeholders. According to one part of the literature, we can emphasize the role of central offices and top management. For example, Netland (2014) described corporate level improvement offices’ personnel as the key players in the development and transfer of lean knowledge and programme-related information to and from plants. Inkpen (2005) highlights the CEO’s role. He investigated the KT between General Motors (GM) plants and NUMMI (joint venture of GM and Toyota) for two decades and found that the real change in KT occurred when a person who had participated in previous KT programmes became the CEO of GM. In addition to such top-down relationships in KT (e.g. from headquarters to subsidiary, the role of CEO) there are important bottom–up (e.g. from plant to HQ) and bottom–bottom transfers. Maritan and Brush (2003) examined four subsidiaries of a North American diversified company to follow the KT of flow (lean) manufacturing. Its subsidiaries established their own lean teams and worked together several times. It is evident at this point, that internal stakeholders’ role and responsibilities are crucial, however, they are mainly treated independently from specific KT practices. It is worth noting that among the internal stakeholders the unit level top management is usually ignored.

In addition, the available description about the practices on the ‘short list’ is rather superficial regarding the type of knowledge flow. For example, as Henriksen and Rolstadås (2010) underscored, the transfer of both explicit and tacit knowledge has crucial importance in the lean context. Clearly, some of the revealed practices effectively support the transfer of explicit/codified knowledge (e.g. manual, training materials), while others support the exploitation of tacit knowledge (e.g. rotation, workshops). Ferdows (2006) points out the importance of both tacit and codified knowledge, and also the importance of fit between the type of knowledge and the practices used. However, these relations are just illustrated in his paper, and they are not lean specific.

Based on this review, we have obtained some knowledge of how multinational companies transfer lean knowledge within their internal networks. However, the vision of a coherent policy supporting KT cannot be constructed, since we do not observe any pattern of practices and conclude only with a ‘short list’ of them. Researchers do not link practices to the nature of knowledge and it is also obvious that the revealed practices are mainly related to lean experts/managers.

RQ1 Who are involved within MNCs in lean knowledge transfer, and what are the applied practices to transfer lean knowledge?

2.3.2. Organizational context of lean knowledge transfer

In the literature, the organizational structure (e.g. hubs at the corporate and/or regional and/or unit levels) that supports lean KT in multinational corporations and the roles (and activities) of different hubs in KT are usually addressed only indirectly and superficially.

Several authors discuss the organizational structure that is related to knowledge management. In his comprehensive description Netland (2014) argues that in many corporations, there are local lean coordinators or teams at sites. If subsidiaries do not have lean teams, they usually appoint lean experts. If a central lean initiative is launched, other levels of lean offices are also introduced, for example, regional, divisional or corporate level lean offices. While types of lean hubs at different levels can be clearly classified, there are no guidelines on how the structure of the organisation and the structure of lean hubs can be matched.

The establishment of lean hubs’ structure raises the problem of ‘division of labour’, namely how roles and tasks are assigned to different lean hubs, among them tasks and activities related to acquiring, development and transfer of knowledge. In centralized systems, the central lean office has the resources to acquire and develop knowledge, and it works with selected subsidiaries in pilot programmes (Netland 2014). However, even a corporation with a central lean office can face a situation in which a template unit has more mature lean knowledge (Maritan and Brush 2003) than other plants or even the central lean office. This phenomenon questions the superior position of the central lean office, and it highlights the importance of a plant’s expertise in knowledge acquisition, development and even transfer towards the central lean office and other units. Secchi and Camuffo (2016) also conclude that the division of several knowledge-related activities (e.g. describing formalized lean system, assessment, roadmap) varies significantly between HQs and plants in multinational companies. Furthermore, many administrative (e.g. reporting, audits) and organizing activities (e.g. events, conferences, meetings) should be addressed in addition to knowledge-related activities. Unfortunately, neither these activities nor the possible division of these activities among lean hubs, are described.

To summarize, in the structure of lean hubs the important role of global and local teams (or coordinator(s)) are recognized. Two main activities are also revealed. Global offices are involved in coordination and organization (e.g. audit, reports), and they directly contribute to the development and sharing of knowledge. The local office is mainly responsible for the plant-level exploitation of lean knowledge.

RQ2 How does a central/divisional/plant lean office (lean knowledge hub) facilitate the transfer of lean knowledge?

2.3.3. General challenges in lean KT – the influence of maturity and roll-out process

Two dilemmas underscore both RQs. (1) Our review revealed that the focus of many papers is on the lean implementation phase, and little emphasis is placed on the analysis of more mature lean organizations (an exception is Netland 2014). (2) There is an implicit assumption of linear roll-out in the lean community ‘through “cascading” a predefined, codified lean production system from HQ to the plants following a predetermined sequence of steps…’ (Secchi and Camuffo 2016, 63). In contrast to this simplified approach, Secchi and Camuffo (2016) concluded that lean roll-out processes are more effective if they are less codified, more decentralized and contextually ambidextrous (individuals do both exploration and exploitation, e.g. routine work and kaizen). These characteristics (codification, centralization and (non-)linearity) can also influence KT practices and organizational contexts.

3. Methodology and data collection

Because not enough knowledge has thus far been developed in the researched area, we chose case-study research for this study (Yin 2003). The aim to provide an empirical basis for lean KT also motivated the application of this methodology.

We conducted a convenience sampling in Hungary to search for subsidiaries that are more mature in lean implementation (at least in their internal network) (see Table 2). The four case subsidiaries have headquarters (HQ) in different countries, and they represent different manufacturing sectors. This diversity of origin and industry helps to obtain insight into potentially different approaches. Each MNC applies a formal and lean based MPIP. Two of the case subsidiaries are part of large multinationals with above 100 factories worldwide (German and U.S.). The two other MNCs are also among the market leaders in their own sectors; however, they are smaller in terms of the number of employees and the number of factories. When we write about a specific case unit we refer to its nationality. Thus, the German subsidiary (factory, plant) means that it is the German corporation’s Hungarian site.

Table 2. Overview of case companies and interviews.

Corporation (origin)	German	US	Mexican	Danish
Product	Small machine tools	Connectors and sensors	Cylinders	Pumps
No. of employees global/local plant	290,000/2100	70,000/1500	20,000/1200	20,000/2000
No. of plants	ca. 440	ca. 100	30	16
Formal LP system	YES	YES	YES	YES
Lean maturity in internal network	Advanced	Advanced	Advanced	Advanced
Interviewees (no. of interviews)	(7) Lean manager; lean coordinator; quality manager; value stream manager; value stream coordinator, purchasing manager; project manager at department manager	(3) Lean manager; lean coordinator; value stream manager	(3) Lean manager; lean coordinator; production manager	(5) Lean manager; lean coach; lean coordinator; quality manager; plant manager

Altogether, we conducted 18 interviews at the four subsidiaries (Table 2), each of which lasted between 45 minutes and 2 hours. Two researchers conducted the semi-structured interviews. If the interviewee asked for the questions, the researchers sent them in advance by e-mail. The interviews aimed to identify the history and organizational background of lean management at the corporate level and at the site, as well as to map the KT practices that are used by the subsidiaries to receive and send knowledge to and from other subsidiaries/central/regional offices in the internal network.

The researchers participated in plant visits at each subsidiary and collected publicly available information about the MNCs. The data collection was complemented by students under researchers’ supervisory activities who either work at the case units or are in the process of writing a thesis about the topic at the case units.

4. Case study analysis

In this section, we proceed with a within-case analysis. Interviews were transcribed and served as the basis of the analysis. The researchers carefully read the transcribed interviews and collected the details of the practices, along dimensions such as the purpose, the involved people and/or organizations, the frequency of actions, the related tasks and the importance of the practice. After individual mapping, the researchers discussed their findings. In cases of disagreement or a lack of information, researchers got back to the companies by e-mail or phone for further information.

To obtain a full contextual, synthesized and comparable picture, case descriptions (based on interviews, other collected documents and available published sources) were written about each subsidiary. The case descriptions were shared with company representatives who adjusted in case of incorrect information and approved from a confidentiality perspective the text.

4.1. The German subsidiary

The German multinational company has four divisions: (a) mobility solutions, (b) industrial technology, (c) consumer goods and (d) energy and building technology, in 440 companies with 290,000 employees. The subsidiary, Robert Bosch Power Tool Ltd. was established in 2003, operates in the consumer goods division and employs approximately 2100 people. The MNC began its lean journey in 2001; in other words, the subsidiary made efforts towards lean from the very beginning. The MNC started lean by implementing lean tools. The MNC developed its own LP system (a tailored version of the Toyota production system) that is completed by a sophisticated audit system to measure the level of implementation at the subsidiaries. In approximately 2010, the MNC realized that it was not an efficient policy to apply lean methodology for its own sake, and it turned towards a goal-oriented approach. The MNC substantially transformed the audit system that not only assesses the level of use but also the impact of using the tools on its strategic objectives. In its division, the subsidiary has the best performance among the subsidiaries. They believe that they can learn a lot from the other divisions of the MNC.

The subsidiary organizes itself around value streams. A value stream can be considered to be a product group with a relatively diverse product portfolio. Not only manufacturing but representatives of other functional areas (e.g. purchasing, logistics) are assigned to a value stream. They have four value streams. Their current initiative is to break the value streams into smaller value streams to obtain homogenous products within them.

The lean organization of the MNC is extensive. Globally, it has approximately 30 people in the lean group at the corporate level. It also has a lean group at the divisional level. Locally, the subsidiary has 8 people in the lean group, and it has 6 additional lean people who are directly assigned to value streams. Global and divisional lean experts usually have deep prior experience that was obtained at the subsidiaries. Lean groups at the global and divisional levels provide handbooks, organize events and audits, and facilitate lean efforts at subsidiaries by providing information and expertise. They regularly visit subsidiaries and check the results of their advice. They also take part intensively, sometimes for several months, in etalon projects, which later are offered for benchmarking visits by other subsidiaries. They do not give directives for factories; however, they set challenges. Progress in factories depends on the ambitions of the factory directors. In addition to lean handbooks, the company operates an intranet where lean projects are documented and shared. These project descriptions are mainly used for ideas and/or to find relevant contacts. There are also regular verbal knowledge (information) exchange opportunities. The global lean group organizes a global lean conference biannually. The conference presentations are carefully selected by the global lean group, and the best ones are awarded. Another meeting, which is called a manufacturing conference, is also organized biannually at the global level for plant managers. At this conference, there are factory visits in small groups into the factories of the selected country to share knowledge. The factory directors also have quarterly meetings. There are yearly meetings for lean groups (coordinators and managers) at the divisional level, which provides a forum for discussions on various themes. Plant managers and lean experts obtain much information from these regular events. A lean audit is a further formal procedure, and it is a good source of knowledge. An audit involves the assessment of the subsidiary’s lean system by two lean experts from two different subsidiaries.

Many activities occur on an as-needed basis. Based on the relevant information that is required, experts, engineers, or even operators are sent to other subsidiaries to study optimal solutions to implement at home. The employees who engage in these travels are selected based on their progress and their commitment to lean improvement. They also have incoming benchmarking visits, at least monthly, which are handled by the local lean group. The company maintains a 2-year junior management programme for talented potential future managers. The participants in the programme are required to spend several months at other subsidiaries where they engage in a dedicated task. Occasionally the lean group is involved in this programme.

4.2. The U.S. subsidiary

The subsidiary is part of a market leader global company with American roots that is currently headquartered in Switzerland. The global company produces half a million different precision products in approximately 100 plants, with 70,000 employees worldwide. The company’s products are available in 150 countries.

The subsidiary, TE Connectivity Hungary Ltd. operates with 1500 employees in North Hungary in the transportation solution business line that belongs to the automotive division and the EMEA region. The product variety (2800 active products), size differences (from nanotechnology to 10 m long) and volumes (millions per product) are all at a high level in the plant.

The global company has used six sigma since 2004, and it started to implement lean in 2007 with a large pilot in the U.S. Several dozens of people were trained for months from that region who later became responsible for lean implementation.

Lean has a global organizational unit with approximately 14 to 15 people, 90% of whom are American. They are former plant or supply chain managers or experts from the outside. The global lean group has developed its own standard materials, training, hierarchy, and auditing system. Its key purpose is to ensure operating excellence in the company. Group members receive special training in a lean tool. For example, a TPM expert controls TPM projects across regions. Within the global lean office, there are 6 regional lean managers, each of whom is responsible for one region (such as EMEA in our case).

In Hungary, a group of 3 people was formed in 2004 that was dedicated to process improvement, first along six sigma, and from 2005 on towards lean. There were several joint kaizen projects with other plants (1–4 weeks) in the EMEA region. From 2007, the projects were replaced by a programme-based approach. The plant needed between three and four years to achieve a relatively mature level of lean; however, developments still continue with the lead of the local operations excellence team, which integrates six sigma, lean and most recently digital manufacturing in its approach.

In the beginning, lean implementations mainly occurred in manufacturing. However, for the last 3 years, lean philosophy has also appeared in business processes such as purchasing, finance and engineering. Usually, one person receives lean training in a given field, and they slowly transform their field. The local lean group currently uses a toolbox of 15 lean tools. There is an expert from the group who is dedicated to each tool, so one expert is assigned 2–3 tools. The global company has developed its own evaluation system to measure operating excellence maturity. The factory is evaluated by each tool and each performance measure annually using stars (1 to 5, 5 stars is the highest level), and the lowest number of stars represents the overall level of factory maturity. The plant is currently at 4 stars.

There are many different forms for exchanging knowledge in the global network, which includes a book of lean standards, various lean training materials and project descriptions are available on the company’s intranet. Offline training, elective or mandatory, depending on the employee’s position, is also provided. There is a deployment plan that is prepared yearly, which contains yearly objectives and the means and the project plans to achieve them. Based on these plans there are monthly regional teleconferences for local lean managers. There is an annual global operations leadership meeting for factory directors and other top managers with sessions. One of these sessions focuses on lean (others focus on issues such as technology, or digitalization). Manufacturing meetings with specific targets (for example, to see the results of a pilot project on site) also take place for plant directors and operations managers two-to-three times per year. The case plant has a closer relationship with the other EMEA region plants.

4.3. The Mexican subsidiary

The Mexican MNC has over 35 plants worldwide with approximately 21,000 employees. They develop and manufacture aluminium components for powertrain and body structure applications in the automotive industry. The case subsidiary, Nemak Győr Ltd. produces cylinder heads with approximately 1200 employees. There were already many attempts in several subsidiaries in the late 90s to implement lean; however, these efforts remained at the subsidiary level. In the early 2010s, there were intentions by local directors to lift lean efforts in the case subsidiary. A global standardization programme of lean tools started later, in 2014. The MNC plans to implement standard lean tools one after the other in half-year waves. Thus far, two lean tools, 5S and standard work have been on the agenda, and value stream mapping will be next.

Both at the global and regional (in our case European) level there is a chief lean manager. There is no independent regional lean group or lean manager. Subsidiaries’ lean groups are organized into a network at the regional level. The chief lean manager who is responsible for the European area is the lean manager of a subsidiary’s lean group. The case subsidiary currently has seven members in the lean group, and, additionally, 3 students also help their work. In comparison, the subsidiary has best practice in TPM, and they are average in the use of other lean tools. Knowledge sharing primarily occurs at the regional (European) level. The MNC has an online platform for project descriptions to share best practices. This online database started in late 2015. Thus far, the local lean group carefully studies uploaded projects, translates some relevant ones to broadcast on factory lean TV, and communicates directly to field managers. They also use descriptions to get inspiration for future projects.

The lean experts meet three times per year for three days each. During these meetings, there is a factory visit and discussion on improvement opportunities. The chief lean manager who is responsible for the European area visits each factory at least once per year. There are monthly video conferences for lean experts with the chief lean manager who is responsible for the European area after their self-reports are sent to the European headquarters with regard to the actual lean tool implementation. In these reports, they should achieve a required level of implementation, which today forces subsidiaries to hire additional lean experts. There are yearly self-audits with external control (by lean experts from other factories who are assigned by the chief lean manager who is responsible for the European area). After the self-audit, three projects should be started at the factories, and their results are controlled by the external expert at the end of the year. Occasional knowledge sharing occurs when workers or experts go to other factories to observe best practices; however, it is still a rare activity. In theory, people can also participate in kaizen events at other factories; however, due to language problems, this subsidiary has not thus far taken this opportunity. While it is beyond lean KT, a new initiative that is called knowledge sharing groups was established. This group contains experts on each production step from different factories who meet bimonthly and evaluate a factory in detail.

4.4. The Danish subsidiary

The Danish MNC operates in the small machine tool business, producing pumps. It has 16 plants worldwide and more than 50 sales companies, and altogether it employs approximately 20,000 people. The subsidiary, Grundfos Hungary Ltd. operates four plants (of 16 worldwide) and a logistics centre with altogether 2,000 employees. Serious efforts into the direction of lean started approximately in 2010 at the subsidiary. The MNC has not mentioned lean in its strategy map; however, it has mentioned shop floor excellence. The global shop floor excellence centre has developed a formalized programme. Subsidiaries that follow this programme have had to set their own goals. In the first phases of implementation, the subsidiary focused on lean tools. One year ago, relying on the help and advices of the Danish subsidiary, they contracted a consulting company to provide lean coaches to develop managerial commitment and knowledge in one of the plants. We conducted our research at that plant.

There is a small central organization for lean at the global level with approximately three people. Four people work in the lean group that locally supports the four plants in Hungary. The primary knowledge exchange occurs at the regional level (Asian, American, European regions). The case plant has the most intense relationship with a subsidiary in Denmark. The subsidiaries in Hungary and Denmark are ahead in lean maturity compared to the other units within the MNC. At the global level, there are training materials that have been developed for different managerial levels. The training of specific lean tools is dedicated to local lean managers. With regard to the division of labour, there are subject matter (lean tool) experts among the lean managers who are responsible for a key lean tool in their subsidiary and for its training for foreign managers. The MNC also has a so-called sharepoint, where project descriptions and learnt lessons are shared (although rarely used, mainly only as a contact point). The lean managers have an annual meeting at the global level. They discuss lean strategy, development, good practices and challenges. The European lean managers have two annual meetings at a European plant. This 3 day meeting also serves as a platform for reviewing lean progress. Furthermore, it is organized as a problem-solving exercise with an output of a plan that must be executed by the local managers. These local managers have the opportunity to take some weeks to visit other factories to facilitate plan implementation. There is a video meeting quarterly for lean experts about progress and projects. The video meeting is more about sharing information.

KPIs and the focal points of lean strategy are set by the global lean centre and the local lean managers. If it is approved by the top management of the MNC, then lean implementation and action are set by the local management. The progress is self-assessed by relying on a standard lean audit that is developed by the global lean centre. It has not been determined whether they have achieved their goals. There is good cooperation with the Danish subsidiary. The Danish managers teach tools and challenge (coach) the subsidiary’s managers on an as-needed basis. The local lean experts can also visit the Danish factories.

5. Cross-case analysis and discussion

After the within-case analysis, we put the practices into a preliminary table that is similar to Table 3. We refined this preliminary table in an iterative way. In case of any doubt, we returned to the company representatives. In the table, practitioners could also review the practices that were used by other companies, so they could provide additional, possibly missing or forgotten, information. Our company contacts verified all of the information that is contained in the paper.

Table 3. Knowledge transfer practices identified in case companies.

	Practice	Short description	German	US	Mexican	Danish	Frequency
Explicit	Manual, handbook	description of tools, principles	x	x	x	x	Always available
	Project descriptions	finished projects with evaluations	x	x (selected projects by central office to be implemented in plants)	x	x	Always available
	Training	Train specific tool in the internal network	Lean tool expert	Lean tool expert (+on-line training))	–	Lean tool expert	On request
	Video meeting	Lean experts		monthly	–	monthly	Regular
	Global conference	Factory directors, top managers	Lean focus, biannual	One session is on lean, annual	–	–	Regular
	Manufacturing conference	Factory directors, on site	General, biannual	target specific, 2-3/year	–	–	Regular
	Lean manager, expert meetings	Global and regional	x	x	x	x	Regular
	Factory director meeting	Special focus on lean	x	–	–	–	Regular
	Audit	On site (S.A. = self-assessed)	Other subs.	S.A. and regional lean managers	S.A. but controlled	S.A.	Regular
	Benchmarking	operators, lean experts (factory directors)	x	x (especially to the Czech plant)	x (limited)	x (especially to the Danish plant)	On request
	Short and long visits	lean experts, lean managers and other employees	x	x	–	x	On request
Tacit	On site lean project	Lean experts lead	–	x (especially important at the start; expert from other plant on request)	–	x	Regular
	Sample project	Best practice development supported by HQs	x (value stream structure)	x (digital lean)	–	–	On request
	Junior management programme	Lean group can be a part of it	x	–	–	–	On request

Cross-case analysis is organized as follows. Since the study of lean KT in mature lean companies is rather exceptional, we first highlight the common characteristics of the case companies’ lean journey. Section 5.1. details the non-linear lean transformation and Section 5.2. describes how the ‘push’ of MPIP resulted in a solid knowledge base at the plants. Section 5.3. and 5.4. turn to the actual lean KT policy.

5.1. A non-linear lean journey – a convergence to central program and mindset change

Considering the lean history of case subsidiaries, each of them changed directions during their lean journey. After some years, the German subsidiary began to utilize lean to achieve its strategic goals instead of simply considering lean implementation as the objective. By coping with challenges in sustaining lean tools the Danish subsidiary realized its lack of managerial commitment and moved towards forming a lean culture by hiring lean coaches and by motivating stronger cooperation between Danish and Hungarian top managers. The Mexican subsidiary started a local lean initiative on its own, and the plant has restarted its lean programme according to the newly developed central standards. Their central lean programme contains classical lean elements, and, because it is a global initiative, the plants can now learn from each other more easily. Finally, the US subsidiary’s approach has also undergone considerable changes. The US subsidiary has completed Six Sigma with lean in its process improvement efforts and now operates a mixed methodology programme. The U.S. subsidiary has completed a tool phase, and it now emphasizes mindset and cultural changes that support its value stream structure. This overview underscores the major changes in the approaches of lean programmes at the subsidiary and MNC levels. Each of our case subsidiaries has undergone a lean transition with substantial changes in its goals, applied tools and principles. Our experience is similar to that of Secchi and Camuffo (2016), Inkpen (2005) and Scherrer‐Rathje, Boyle, and Deflorin (2009), who found deviations from linear lean evolution. This substantial change in the approach during the years (from a tool-based programme followed by efforts in terms of mindset changes) should also have an influence on the evolution of KT system as detailed in the next section.

5.2. The coercive implementation of lean tools pushed by a central office – a basis for absorptive capacity

Almost all of the interviewees thought that they had been wrong to use the initial tool-focused approach, and that was the reason for the change. However, analysing the history of the subsidiaries from a knowledge management perspective, starting to implement lean tools first seems to be necessary to build a relevant level of absorptive capacity (Cohen and Levinthal 1990) to provide a common language (Netland 2014), which can serve as a platform for further learning (Ernst and Kim 2002). Building this absorptive capacity alone by relying exclusively on local resources is difficult. Even if the subsidiaries are motivated, such as in our Danish and Mexican cases, they still require certain levels of push, orientation and external target. Therefore, the existence of a central unit seems to be necessary, especially at the beginning of the development. In terms of this issue, our results contradict those of Secchi and Camuffo (2016), who found that a decentralized development is more advantageous. We propose that this argument is valid for subsidiaries that already have enough knowledge and capability in lean. In the first stage of development, global push and orientation (e.g. context milestones, a well-framed lean programme, training materials, assessment systems, training the trainers through on-site programmes) are fruitful for building the required level of absorptive capacity and a common language to understand each other. Later, the role of the central unit can change to become a facilitator by organizing global events and supporting subsidiaries with information and experience.

5.3. Transfer practices of explicit and tacit knowledge

There are many different practices of KT (Table 3). Some of them provide a platform for explicit knowledge exchange, while others serve the flow of tacit knowledge. Explicit knowledge exchange shows high similarities. An intranet is used in each case for the sharing of lean tool handbooks, training materials and project descriptions and evaluations. We argue that explicit knowledge sharing is important at the start to build absorptive capacity. So, our organizational level experience is consistent with Letmathe, Schweitzer, and Zielinski (2012) who reached a similar conclusion with regard to individual learning when people had to learn new tasks.

There are quite different approaches with regard to the exploitation routines of project descriptions. The German and Danish subsidiaries consider these descriptions to be a ‘library’ and a potential source of information (e.g. whom to contact, main ideas) to launch a similar project. The lean experts of the Mexican subsidiary select the most promising ones and share them within the plant to motivate others towards their implementation. However, only a limited number of projects are adapted. The U.S. MNC has a unique routine to motivate the internalisation of explicit knowledge. Their central lean office regularly reviews project descriptions that are sent by the plants and selects certain promising ones. The central lean office expects local teams to adopt the selected ideas in a few months. This routine shows an efficient way to exploit explicit knowledge (i.e. project description) throughout the network. Ultimately, this routine is about the internalisation of explicit knowledge in weak tie inter-unit relationships (i.e. loose coupling between the units and people of units) (Hansen 1999). This routine underscores that the internalization of a relatively less complex lean project can be successfully managed. Furthermore, because the descriptions of the best projects are shared in the network by the central lean office, plants usually strive and even compete to produce such projects to raise awareness of their lean efforts and expertise.

For more complex knowledge, there are regular and on-request KT practices at the subsidiaries. The palette of practices is more colourful in this context, especially if we take into consideration the role of the involved employees. In most of the cases, tacit KT occurs on an as-required basis, which shows the importance of the motivation of the recipient in the process (Easterby-Smith, Lyles, and Tsang 2008). In the German subsidiary, directors play a key role in regular exchanges (e.g. global conference and manufacturing conference). Their meetings constitute sophisticated information sharing practices in which both explicit and tacit knowledge elements are shared. These conferences provide the opportunity to exploit weak inter-unit ties, so they can decide where to send their employees to acquire new knowledge or develop skills. It can also increase the commitment of decision makers to lean management (Boyle, Scherrer‐Rathje, and Stuart 2011). Therefore, the top-down approach in this German case is much stronger than it is in the other three cases. In addition to the top-down approach, the German subsidiary has a bottom-up approach, in parallel, as is also discussed by Henriksen and Rolstadås (2010). Although directors obtain advice and ideas from the divisional lean office, the directors make the decisions. In theory, we discuss the development of managerial commitment and involvement as important success factors of lean implementations (Womack and Jones 2003). We believe that the German case represents a good example of this phenomenon, both by involving plant managers into weak inter-unit ties (conferences) and by challenges that are set by the divisional lean office. Their involvement clearly demonstrates top management’s orientation towards lean culture.

In the other cases, lean managers, lean experts and their strong inter-unit ties (i.e. cooperation between the lean units of various subsidiaries, mainly within regional networks) are the real vehicles for the transfer of lean knowledge. At these subsidiaries, factory directors might be influenced; however, they do not necessarily follow the ideas of the lean experts. Different practices for the involvement of plant directors into the lean transformation and into the transfer of lean knowledge are in place at certain sites. The CEO of the U.S. subsidiary is the former leader of the local lean group. At the Danish corporation, the Hungarian top managers are (lean) coached by Danish top managers. We acknowledge that these practices can facilitate the transfer of lean knowledge; however, these are contingent and provisional solutions.

Our work confirms the findings of Netland and Aspelund (2014) that both creation and diffusion of knowledge has strategic importance in MPIPs. At the same time, we have substantially extended the ‘short list’ of six–nine KT practices presented in previous works (see Table 1), and have provided a ‘long list’ of about more than 10 practices (see Table 3) for managers to select from to effectively support KT. Our findings offer a diverse set of KT practices, from document sharing through occasional and regular meetings to auditing.

We highlight an interplay among the practices on the ‘long list’. Although some of the revealed practices can be considered more efficient for sharing knowledge, less efficient practices are still required to provide the platform for better ones. Altogether it is a complex system of practices (even with changing emphasis on specific practices during the lean journey), thus building upon and supporting each other is required to achieve successful KT. This interrelated nature of KT practices has not been emphasized previously in lean KT literature. Although we originally tried to find the most useful practices for KT, we realized that it is the subtle network of various practices that makes the transfer happen. This complex system of practices approach contrasts the experience of Boscari, Danese, and Romano (2016). They referred to international teamwork as a primary practice of lean KT. However, we cannot differentiate between primary and secondary practices because the revealed practices are complementary, and their integration leads to a ‘fine web’.

Our findings also draw attention to how different internal stakeholders can boost lean journey. Established practices to involve top managers such as plant and/or operations directors, whose roles were overlooked previously, into lean KT can strengthen the transfer coalition (Kostova 1999). In certain cases, where only lean experts/managers are involved in the transfer, the real decision makers can easily decide to ignore suggestions from the lean team. Hence, we argue that a sophisticated system balances between strong-tie-oriented, expert related practices and weak-tie-oriented, manager related practices. The latter is especially important to present management’s commitment and their mindset change.

The revealed lean KT system shows some unique characteristics when interpreting it based on the best-practice KT literature (O’Dell and Grayson 1998, 2004). At our case subsidiaries, the groups of lean experts actually act as a best practice teams, which is one of the four basic transfer practices. We have also pointed out that a coherent KT policy utilizes top managers’ weak-tie relations which is hardly covered by the four basic practices. Furthermore, the four basic practices are more about teams and people’s network, our practices are more practical since they depict the daily routines of teams and networks.

Some contextual factors have also influence. Lean experts/managers prefer similarity (in size, technology, product) and plants in close geographical distance (Rosenkopf and Almeida 2003). These similarities are much more important than cultural barriers and communication difficulties in KT, which surprisingly were not mentioned at all. At some of our case subsidiaries, the contextual factors are formally embedded in the lean KT system, for example, the loose geography-based network of lean managers at the Mexican company, and the very strong divisional barriers that are faced by the German and U.S. plants, because they can only collaborate with lean teams from the same division.

5.4. The role of lean offices

Each case subsidiary follows a corporate level lean programme that is designed by the central lean office. However, there are many remarkable differences in plant autonomy and in the organizational context with regard to the division of labour and responsibilities in knowledge-related and administration-related activities. (Table 4)

Table 4. Models of organizational contexts of case companies.

Dimension	Case company	German	US	Mexican	Danish
Corporate level lean initiative		YES	YES	YES	YES
Plant autonomy		YES priorities are based on business targets and audit	YES priorities are based on business targets and audit	limited HQ issued roadmap	YES remarkably, recently priorities are set by the template plant
Central lean office (corporate level)		YES	YES	YES	YES
Divisional lean office		YES	YES	not applicable (one division)	not applicable (one division)
Regional lean office		YES	YES	No dedicated team, network of regional lean managers	No dedicated team
The role of dedicated central/divisional/regionnal lean office	Administration	Conferences, events, training, reports	Conferences, events, training, reports	Mainly regional network is important	Conferences, events, training, reports
	Knowledge	Holds training, develop principles and tools, works on pilot programmes	Holds training, develop principles and tools, works on pilot programmes	Develop standards, mainly regional network is important	NO, knowledge is at the template factory
Model of organization context		Central lean office with full functionality	Central lean office with full functionality	Regional chief lean manager dominated model	Template factory dominated model
Possible explanation of differences		No. of employees and many large plants	No. of employees and many large plants	Less employees with only some large plants	Less employees with only some large plants
		Lean fits very well the manual assembly process	Lean fits well the manual-automated assembly process	Assembly and manufacturing, technology-oriented corporation	Assembly and manufacturing, relative low proportion of labour expenses

Two of the subsidiaries have a centralized organizational context behind lean KT. Their lean organization fits the existing organizational structure (product or region based). The German and U.S. models have central lean offices with full functionality: they organize events, collect reports and they are also knowledge hubs.

Further, we revealed two mixed models of lean organizations. In these models the role of the central hub is different. The Mexican corporation launched a global lean standard under the supervision of the central office. However, its model is dominated by the regional chief lean manager who is the key person in organizing activities and supervising progress in his region. The model of the Danish corporation can be characterized as a template factory dominated model: the acquisition and sharing of knowledge are fully delegated to the plants. The template factory is a plant in Denmark, and the ‘number two template factory’ is the case plant in Hungary.

Although the four case subsidiaries have different organizational contexts, each of them can be considered successful in implementing lean. However, we have witnessed a stronger and a weaker shift to a more centralized lean system. The strong shift of the Mexican corporation is evident. Today, lean is a global standard for them, and it is not a decision that was made by local top managers. The Danish company has also strengthened the role of the central lean office (e.g. renewed the audit system) and it has focused its strategy on lean. However, its template-dominated model has not yet been fully challenged. Despite these clear differences in the models of lean organization, there is a substantial similarity in the deployment of the lean system: the subsidiaries themselves should find local practices and need to develop the absorptive capacity to be able to adapt and implement the obtained knowledge.

As mentioned above, the case subsidiaries are more mature in lean compared to the peer subsidiaries in their own networks. This indicates that the organizational support and expectations from the multinational HQ (provided for each subsidiary) are necessary but not enough to be successful. The variance of lean maturity of the different plants of Volvo was also emphasized by a case that was recently published (Netland and Ferdows 2014).

Our findings underline the previously classified structure of lean hubs (Netland 2014) ranging from central lean office to local lean expert. We also emphasize that in the case of central lean programmes the structure of lean hubs fits the organizational structure. A novel finding of ours is that there is no one best way to organise lean hubs’ structure. More specifically, there can be considerable differences between larger and middle-sized MNCs. The former move on a more deterministic trajectory than middle-sized MNCs regarding their lean hubs’ structure. We have also some limited evidence that central lean office has a more important role in knowledge related activities in larger plants. In both contexts, the role of template unit offers some room to manoeuvre in the given structure, usually for larger plants.

Looking at the level of local lean offices, since each of our case plants is large, so they have enough resources for lean programmes (i.e. full-time employees in a local lean office and thus a high level of absorptive capacity), and they also tend to gain the headquarters’ support (Tsai 2001). Especially our U.S. and German case subsidiaries, which are templates (i.e. responsible for developing new lean knowledge for division, or MNC level with central lean office support) show that a central network position is useful for getting easy access to new knowledge (Tsai, 2001).

6. Conclusions and limitations

Creation and diffusion of knowledge in lean based MPIPs is a strategic issue for MNCs, however, they are still struggling to effectively coordinate related efforts in their internal networks. Although practical examples prove that multinational HQs make efforts to build learning systems for the transfer of lean knowledge among units, there are only a few studies that address this topic. Merging knowledge management findings with the operations management studies could provide guidelines for companies on how to organize themselves. Our research illustrates this transfer of lean knowledge for four cases with a complex picture of the practices and organizational contexts from the subsidiaries’ perspectives. Our work provides several novel insights into lean KT. Findings 1–3 relate to the use of practices (RQ1), while 4–5 give insight into organisational context (RQ2).

First, to offer building blocks for a coherent lean KT policy, we link practices, the nature of knowledge and internal stakeholders. Based on our cases we highlight the following characteristics of a sophisticated lean KT system: (a) it relies on many different practices that support the transfer of both explicit and tacit knowledge; (b) it enables the flow of knowledge among units in top-down, bottom-bottom and bottom-up relationships; (c) it ensures the deep and active involvement of (local) top managers and experts to form an effective transfer coalition; (d) it mixes regular and ad hoc (or on request) events; (e) its daily operations are supported by a lean hub structure that fits into the existing organizational structure of the company. Most probably a KT system that is based on these characteristics is required to support the transfer of other types of best practices that are meaning- and values-based (e.g. Total Quality Management). We argue that although these characteristics are not unique, all of them are necessary for a successful lean KT. In other words, managers aiming to effectively diffuse lean knowledge in internal network should consider the deployment of all practices from the ‘long list’.

Second, we have presented a ‘fine web’ of KT practices and identified the different roles of practices. The involvement of management in the lean transition is a top priority, and we have some evidence that a well-functioning KT system presents the opportunity to deeply involve top management. Lean-focused regular conferences can especially strengthen managers’ participation in transfer coalition and demonstrate their acceptance of the lean culture. Among these practices we have also identified some that were not previously associated with lean KT (e.g. conferences, management programmes). Furthermore, we have also found evidence that in addition to strong-tie and expert-focused interactions, companies should make effort to institutionalize practices that support the internalization of explicit knowledge in weak-tie relationships. Understanding the importance of weak-tie relationship and the involvement of all levels of the hierarchy are essential practical implications of this research.

Third, while the tacit nature of lean knowledge has been highlighted by the previous research (e.g. Henriksen and Rolstadås, 2010), the distinction between tacit and explicit knowledge has helped us to understand their roles in lean transition. We argue that at the start of the lean journey, it is necessary to learn the basic lean tools to have a common platform upon which to build (to have the absorptive capacity). Managers should consider that this start should be effectively supported by practices that also transfer explicit knowledge. In the later phases of the lean transition, practices that promote tacit knowledge have importance.

Fourth, our findings indicate that similar KT systems are deployed in very different organizational contexts. A lean organization that is dominated by central and divisional offices is one viable solution in MNCs. It is also evident that multi-plant global companies that do not have a divisional structure can provide different approaches to linking local teams and lean knowledge hubs, that is, template plant and loose networks of lean managers. We have also pointed out that the best option is the template factory role for local top managers committed to lean initiatives to upgrade their plant’s position in the internal network.

Fifth, our cases clearly show that MNCs’ MPIPs have undergone a considerable transition in the last 10–15 years. We believe that many other companies have matured their lean journey and have changed their tool-oriented approach to a culture-oriented approach over time. In our sample, the German and U.S. corporations are good examples of this phenomenon. The approach that they represent today surely emphasizes the values and meanings that are embedded in LP. Thus, we can speculate that the approach and perhaps also the organization of today’s multi-plant programmes are different from that of the 1990s. For example, there has also been a change in the direction of knowledge flow. While the transfer of lean knowledge from developed countries to developing ones is still important (Mefford and Bruun, 1998), our cases have shown that knowledge – especially through template units – also flows in the other direction.

The main contribution of this paper is to reveal the links among KT practices, stakeholders, the nature of knowledge and KT frequency, and based on these links proposing a coherent policy for lean KT. By doing this we believe that the OM community has a better understanding of the ‘modus operandi’ of lean KT. It is important to note, however, that this coherent policy is more a necessity than a sufficient condition for better performance, as indicated by different stages of lean implementation at subsidiaries belonging to the same multinational company.

Our research is not without limitations. An important deficiency is that we have only the viewpoint of one subsidiary in each company. It would be of interest to study subsidiaries and the HQ of the same company (such as in Netland 2013) to investigate the same phenomenon from several perspectives. However, our objective was to collect as many different practices as possible for lean KT, thus companies that originated from different cultures seemed to be a better alternative.

Furthermore, we only address the practices of KT and not how this knowledge is internalized and exploited by the subsidiary itself. Therefore, we cannot state that one way of knowledge sharing is better than another because we have not measured the efficiency and effectiveness of the KT process. Therefore, we do not know the extent to which various KT practices, as well as the whole KT process, have contributed to the success of the case subsidiaries. Nevertheless, the contribution of KT practices is implicitly there because the interviewees most likely spoke about the practices that have the greatest impact on their operations.

We also did not address other external sources of knowledge in our paper, although conferences on lean management and knowledge exchange with other local companies (that have similar contexts) can also contribute considerably to lean improvements.

Finally, an important limitation is the number of cases. Our intention is to conduct further case studies, and after arriving at a relatively full list of practices to perform a survey to obtain more robust results.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Research, Development and Innovation Office – NKFIH under Grant 112745 titled ‘The impact of plant roles on knowledge sharing in manufacturing and supply networks’.

Notes on contributors

Krisztina Demeter

Krisztina Demeter is a full professor and the Head of the Department of Logistics and Supply Chain Management at the Corvinus University of Budapest. She teaches operations management and related topics. She has/had responsibilities in International Society for Inventory Research, European Operations Management Association, Global Manufacturing Research Group. She is also involved in the International Manufacturing Strategy Research network. Her major research areas are manufacturing strategy, global operations management, servitization, lean management and industry 4.0. She has several international publications.

Dávid Losonci

Dávid Losonci is an associate professor at the Corvinus University of Budapest, Hungary. He works for the Department of Logistics and Supply Chain. His main research interests cover production and service management, more specifically lean production and Industry 4.0. In relation to lean production, he deals with the role HR, the transformation of organisational culture and proper analysis of attitudes and competencies of employees and managers. He has recently joined a research team dealing with Industry 4.0.