Network Theory

Network Theory

‘It is not a world of individual and isolated transactions between companies. Instead, each node or business unit is bound together with many others in a variety of different ways through its relationships‘
Hakannsson and Ford 2002 pp133

No organisation exists in isolation. It is through the total sum of direct and indirect interactions that an organisation embodies meaning. Whether we are considering the individual employee, their department or an international industry, there are a multitude of differing relationships. Each relationship is different and its own demands, pushing and pulling to get what it wants in a manner comparable to a child. To exist within this “kindergarten” network we must understand how each relationship affects us.

This section looks at driving forces behind the introduction of the network concept and discusses the emerging challenges for management. A general view is discussed focussing on a specific perspective proposed by the IMP Group .

1 A changing landscape

It is through the study of dyadic relationships that the familiar concepts of supply chains and supply chain management were established by Oliver and Webber (1982). Initially they concentrated on the internal flow of resources within an organisation but quickly extended beyond the boundaries of the focal firm (Womack, 1990; Harland et al. 1996). By viewing the world as a collection of simple buyer-supplier relationships, supply chain strategy became the driving force behind many of the large multinational companies of the 1970’s and 80’s (Moller et al., 1999; Snow et al. 1992). This strategy was favoured by management as it gave a straightforward picture of their business area. It allowed resources to be allocated to the key primary suppliers and all others to simply be ignored.

Towards the end of the 20th century markets and technologies were changing dramatically. Moller et al. discuss four of the key driving forces in their work Business Relationships and Networks: Managerial Challenge of Network Era (Moller et al., 1999). Their framework, shown in figure 3, centres on a focal organisation expanding the horizontal and vertical relationships both upstream and downstream. The driving forces are depicted by arrows in the outer corners.


Driving forces of change on a firm’s horizontal and vertical relationships - Adapted from Moller et al. 1999

The effects of the driving forces on the market are discussed.

Horizontal Changes
Upstream changes can be attributed to the strong tendency over the last 20 years for organisations to outsource all but their non-core activities (Prahalad and Hamel, 1990; Quinn, 1999). This has acted to both widen and lengthen traditional supply chains creating multiple tiers of numerous suppliers. Globalisation has opened up more potential supply routes but in turn has increased the level of competition.

Looking downstream, globalisation has created many new potential customers but direct access can be difficult resulting in the rise of powerful distribution channels i.e. Tesco and Walmart. The Internet offers a potential route to establish direct contact but creates the need for specialist creation, maintenance and promotional roles that most manufacturing companies will have no experience of.

Vertical Changes
As the effects of globalisation are felt companies are increasingly looking to their competitors for support. A strong trend for strategic alliances helps companies keep abreast of the increasing complexities (Roy et al. 2005). The cost of developing new technologies is ever increasing alongside the risks involved. Failure could mean the end of individual companies so alliances are formed to spread this risk.

‘companies are forced to collaborate in view of shortening product life cycles, rising R&D costs and growing knowledge specialization‘(Ritter and Gemunden, 2003).

Each of the effects discussed have acted to increase both the number and importance of relationships. In this new landscape the large enterprises of the 70s and 80s were now like dinosaurs unable to adapt to the new demands (Snow et al. 1992). Their supply chain strategy was not suited to deal with the complexities of multiple relationships. This new landscape called for an alternative view.

The image of atomistic actors competing for profits against each other in an impersonal marketplace is increasingly inadequate in a world in which firms are embedded in networks of social, professional, and exchange relationships with other organizational actors“ (Gulaiti, 2000 p 203)

We see here the introduction of the term network which has wide implications for management strategy. The conclusion of analysis by Moller et al.. can be summarised in their statement:

We seem to be entering an era where traditional markets are being replaced by networks of interrelated firms and other actors“. (Moller et al., 1999 p 414)

2 Resultant Management Challenges

The emergence of a network era creates problems for traditional management. Two such issues are discussed here.

Chaos Management
How can a single firm hope to gain control of an entire network? There seems to be a growing consensus that this aim is impossible to achieve (Hakannsson and Ford 2002). Ford (2002) states that ‘the inherent complexity of intercompany relationships and networks means that it is unrealistic to imagine that they can be wholly designed by any one party, still less that their evolution can be solely the result of conscious one sided plans’. A network is a chaotic place to exist, but this does not mean that an attempt to predict or control possible outcomes is entirely futile: active steps can be taken. Theory suggests that by taking a network view of the focal organisation it is possible to manage ones position to gain an advantage over competitors (Jarillo and Stevenson 1991).

“In the management perspective the problem becomes one of how a company can cope with change in the network when it has virtually no possibility to predict with any accuracy any future state.“ (Hakannsson and Snehota, 1995, p 270)

Network Boundaries
Before we can realise any such network, we must first appreciate that we will never be looking at the full picture. Any network is simply a selected snapshot of a much wider one. Mattson (1988, p 132) comments ‘We can regard the global industrial system as one giant and extremely complex network since there exists always some path of relationship that connect any two firms’. He goes on to say ‘However, for obvious analytical reasons, a total network must be subdivided according to criteria such as interdependence between positions due to industrial activity chains, geographical proximities etc’. In truth every single organisation is connected in some way to every other. We can potentially view the entire world as a huge network of interacting bodies. Unfortunately such a network is so unbelievably complicated that no meaningful analysis can be performed. The key to useful network analysis is explicitly stating our boundaries; we must define our research space.

However, it is almost impossible to state a general formula for defining a network as it is dependent on the criteria that we wish to observe. Fombrun (1982 p282) states that ‘there is no agreed boundary to an inter-organisational network, the choice of the boundary should reflect the purposes of the researcher and the research hypotheses of the study’. This sentiment was suggested considerably earlier by academics working in the much more mature area of social networks theory.

“Clearly some limit must be put on the number of links to be taken as definitive for any specific network, otherwise it would become co-extensive with the total network. This difficulty is resolved by fixing the boundary of the network in relation to the social situation being analyzed...There can be no general rule.“(Conway et al., 2001 citing Mitchell, 1969)

A network only exists once you define the area you will be working within. These boundaries themselves are human constructs. It is the very act of definition that creates the space in which we shall be working, such that a different definition will give a drastically different picture. A number of different network perspectives exist but for effective analysis an organisation must consciously choose how it wishes to view its competitive landscape.

3 IMP Perspective

As the business market changed, a group of academics who were unhappy with traditional approaches formed. The IMP group was formed in the mid 70s by a group of researchers over five European countries. Their aim was to observe and discuss the way in which markets are created, operate and how organisations interact over long timescales (IMP Group, 2009). This group published many pieces of work (Hakannsson, 1987; Easton, 1994; Ford, 2002; Ritter, 1999; Johanson and Mattson, 1994; Hagg and Johanson, 1983), stimulating much debate and moving academic opinion towards viewing markets as networks (McLoughlin and Horan, 2002). Integral to the belief of the IMP group was the idea that no organisations exists in isolation (Hakannsson and Snehota, 1990). All exist as part of a vast network of interacting business relationships (Moller, 1999).

By building on previous work on social-network theory the IMP group aimed to provide a more suitable framework for viewing this new market (Anderson, 1994). Within the following section we shall discuss the IMP perspective by considering such a framework.

To assist with the conceptualisation of this network perspective, the author would like to draw a parallel between organisations within a market and a collection of interacting intelligent particles. Particles are positioned through space and interact via different forces such as electrical, magnetic or gravitational. They will settle into positions depending on the vectoral total of all their interactions. The notion of intelligence implies these particles have a limited amount of control over their position and attempt to actively manipulate the interactional forces at play by adjusting their internal attributes. This model shall be referred to as Organisations as Intelligent Particles (OIP) and use d to help explain the overall conceptualisations throughout this study.

Four lenses are suggested through which we can view a collection of organisations (Easton, 1992). The lenses ‘structure’, ‘position’, ‘relationship’ and ‘process’ each give a different perspective revealing different threats and opportunities. Each will be discussed in turn:

If all organisations act completely independently then the network will have no structure. Resources will be consumed and reformed in isolation. As independence decreases and interdependence increases the network will start to take shape. The structure is governed by both the strength and type of interactions between each organisation. The greater the interdependence the more rigid and defined the structure will become (Easton, 1992) so all networks can be visualised as falling on a spectrum. At one end we have a completely randomly distributed mass of non-interacting organisation and at the other a rigid structure with very limited internal movement.

To claim that a network has structure, we automatically assume that it has an edge. As discussed later, it is not until we define the required criteria for inclusion that a network can exist (Mattson, 1983). So a word of warning: networks are social constructs and not exact models, therefore analysis is never explicit and always open to interpretation. As long as researchers bear this notion in mind then meaningful research can be conducted but if forgotten, theory can easily be applied out of context to the detriment of an organisation (Easton, 1992).

Drawing on the OIP concept we view structure as the collective shape the particles take. With the absence of interaction we would simply see a collection of particles dispersed evenly across our research space. As interactional forces are introduced particles arrange themselves into either one single structure or a number of interacting structures which can exist independently. This conceptualisation allows us to analyse the structure at two levels. The first is its overall shape within the defined research space ; we consider its physical shape if it is long and multi-levelled, broad or narrow etc. Here we can make judgements on how fit for the purpose the current shape is. The second level of structure is the way in which the particles are connected: in this conceptualisation the particles are connected via interactional forces and the strength and frequencies of these interactions will vary depending on the ratio of specific attributes between two particles. We can describe the network structure or substructures in terms of their physical properties such as brittleness, elasticity, density, fluidity etc.

The positional lens takes an individual organisation as its focal view point. The network is defined around this focal-organisation and the relationships they hold with others will define its position. Therefore a network can be described as an aggregation of interlocking positions (Easton, 1992). Considering this from a different angle we can say that an organisation is in fact defined by the net effect of the relationships it holds within a network (Ford et al. 2002). Where an organisation appears within a network may have strong implications on the way it acts and its corresponding strategy. Its position is defined by factors largely out of the organisation’s control which can lead to them making assumptions on their position which are quite different to reality. Ignorance, arrogance or a poor understanding of the surrounding network may cause this misunderstanding resulting in the wrong strategy being implemented. This lack of positional awareness could be very damaging for a company.

Demands will be made by other firms which will attempt to pull the focal-organisation into new positions. The focal-organisation in turn is able to make demands on other firms to manipulate their position. Active steps can be made to improve a single organisation’s position but due to the sheer number and complexity of all possible interactions, this is extremely difficult to achieve. Ford (1997) states “the inherent complexity of intercompany relationships and networks means that it is unrealistic to imagine that they can be wholly ‘designed’ by any one party, still less that their evolution can be solely the result of conscious one sided plans”. Hakannsson and Ford (2002 p 139) state, ‘Each company will try to develop its position in the network relative to other companies, by influencing the knowledge and understanding within other companies and the direction in which each relationship develops’. It is through these desires and demands that organisations change their position and the network evolves.

Considering the positional lens from the point of view of OIP we start by considering an individual particle. We can describe this focal-particles position relative to the other particle’s within the structure. Its position will have an effect on the strength of interactional forces as distance increases or other particles block the way. These particles may try to change their position but cannot just reposition themselves within a structure by magically appearing there; they must slowly travel by manipulating the existing interactional forces and establishing new ones. The further they wish to travel, the more effort will be required. Bearing in mind the fact that other particles may be competing for the same positions or attempting to influence the focal-particles position, we can start to understand what a mammoth task repositioning really is.

The process lens can be applied at two levels. The first more simplistic level is concerned with the internal workings of an organisation; the way in which resources are utilised to create products, services or knowledge (Hagg and Johanson, 1983). First level processes will be relatively well defined and efficient as they are created by a single organisation. It is through these processes that companies compete. Second level processes are the cross-organisational tasks that a network performs (Easton, 1992). These will inherently be more complex and potentially much less efficient. They have been created from co-ordination across multiple organisations each of which with their own vested interest. By studying both first and second level processes we can start to understand why a specific network has taken the shape it has.

Applying this lens to the OIP concept we can think of processes as the way in which interactions take place. First level processes are concerned with the way in which a particle creates and arranges its internal attributes to affect the result of future interactions. Second level processes are concerned with the reasoning behind substructures arrangements; how the network is planning to operate. Particles only have direct control over first level processes and varying control over second. By making changes at the first level, particles may be able to influence second level processes to their advantage. It is through this first level process change, and process change alone that a particle is able to re-position itself.

The final lens suggested by Easton (1992) was that of relationships. A relationship lens is applied at the dyadic level as it is concerned with the assumptions and perceptions between two firms. The IMP model was the first to consider the network not as a collection of organisations, but a collection of business relationships. Actions were not considered as short-term reactionary exchanges but as a result of long-term planned decisions (Hakannsson, 1987). A distinction should be drawn between relationships and interactions. Interactions are short term, concerned with the here and now, whereas relationships consider long term perceptions, expectations and ideas of trust (Johanson and Mattson, 1984). The nature of a relationship is defined by the aggregation of all previous interactions, both direct and indirect. Research into relationships within a network is a truly vast area with many academics working solely within this field. It attempts to understand the logic behind human decisions - a huge effort in itself - and stems from the well established social-network field crossing over to areas such as psychology, anthropology and at times philosophy (Conway et al.,2001 citing Mitchell, 1969).

"no one relationship can be understood without reference to the wider network."(Hakannsson and Ford 2002 pp134)

The absence of relationships from a network would result in a perfect free market model (Easton, 1992). Companies would compete purely on a transactional basis with no trust, loyalty or bias. Relationships are the human part of a network and as such are hard to explain and even harder to model.

"relationships are likely to be complex and long-term and their current form is the outcome of previous interactions between the business units."(Hakannsson and Ford 2002 pp134)

Within the OIP model it is the relationship aspect that gives our particles intelligence. Particles memorise the resultant effects of previous interactions. If the result was negative then a particle will protect itself from similar future interactions. If a previous interaction was deemed positive, then the particle will attempt to influence the surrounding network to increase the likelihood of similar occurrences. Relationship effects act to increase the unpredictability of networks as it is impossible for an outside particle to fully understand the previous interactions between two others.

All four lenses discussed offer a different way of viewing our network, highlighting unique threats and opportunities, but they are still only a viewpoint. It is through human interpretation that any meaning and practical implication are revealed.

4 Network Competence

"Traditional markets are being rapidly replaced by networks. This poses major managerial challenges for industrial and high technology companies."(Moller et al., 1999, p 413)

The birth of the network era called for a change in management and management theory. A competence based framework was devised by scholars within the IMP group which allowed formal measurement of an organisations capability to manage a network. This final section will discuss this framework.

In the early 90s Prahalad and Hamel (1990) first proposed the concept of core competencies. The aim was to propose a construct for understanding the underlying roots of competitiveness (Ritter and Gemunden 2003). They proposed that within an organisation it is possible to highlight and measure the areas which allow them to perform business well. These areas will be consistent across all organisation and the most successful organisations will perform well in each of the competencies (Quinn 1999). Prahalad and Hamel’s concept initially concentrated on technological competencies but this was extended to include managerial competencies (Day,1994; Dosi and Teece, 1993).

The term competence can be defined in two ways according to the literature:

  1. To describe resources and preconditions, i.e., qualifications, skills, or knowledge, necessary to perform certain tasks without considering the actual execution of the task (Ritter and Gemunden, 2003).
  2. To describe a process of specific activities an organisation undertakes (Day, 1994).

In their conceptual paper “Managing technological networks: the concept of network competence” (Ritter and Gemunden, 1997) Ritter and Gemunden incorporated both definitions of competence and applied a network perspective. They built a framework which outlined the key competence areas for managing within a network.

"A company’s degree of network competence is defined as the degree of network management task execution and the degree of network management qualification possessed by the people handling a company’s relationships"(Ritter, 1999)

The structure for the framework can be seen in figure 4, with the competence areas shown in green.


Ritter and Gemunden’s Network Competence framework (Adapted from Ritter and Gemunden 2003)

The framework itself makes a formal distinction between the two definitions of competence. The left hand branch describes the tasks that need to be performed in order to manage the relationships both on a dyadic and cross network level. The right hand side describes the skills and expertise an organisation needs to effectively carry out these tasks. The following section will define the specific competencies attributed to each area.

Network Management Tasks
The framework distinguishes between the competencies needed to manage relationships at the one on one (dyadic) level and as the network as a whole. This idea is conceptualised in figure 5.


Conceptualisation of the nature of the task execution branch of the network competence model (Ritter and Gemunden 2003)

Relationship Specific Relationship specific tasks are competencies that should occur with each individual relationship. They are tasks or roles which should be in place to ensure healthy relationships at the dyadic level (Chiu, 2009). The specific competencies are:

Initiation Any relationship must start with first contact. This competence describes the way in which the relationships are put in place. Initiation may occur from organising or attending trade shows, setting up formal or informal communities of practice or through recommendations from existing suppliers.
Exchange In order for any relationship to be maintained there must be a continuous exchange between the two parties. This exchange can take many forms, for example, transfer of products, services, money, information, know-how, and personnel (Ritter, 1999).
Coordination In addition to the exchange competency the action needs to be coordinated in such a way that they are in tune with each other (Mohr and Nevin, 1990). This competency establishes formal rules and procedures that govern all exchange activities.

Cross Relational Tasks
The previous competencies build on much of the previous research in to dyadic relationships. Taken in isolation they can strengthen individual ties but if a network view is to be taken these relationships must be co-ordinated by a set of cross relational tasks. Ritter and Gemunden identified four such competencies.

Planning To be able to get maximum benefit from all dyadic relationships planning is needed to ensure the actions now have a positive impact on the future state of an organisation. This can be a very complex competency as it is attempting to predict the future state of mostly chaotic system. The quality of planning can be improved by careful monitoring and analysis of a multitude of factors such as internal resource availability, PESTLE analysis or competitors’ movement (Ritter, 1999).
Organising Each plan must be organised in such a way that individual responsibility is assigned to its realisation and the necessary resources are allocated. Individual actions should be verified to ensure they do not have a negative impact on other relationships (Ritter and Gemunden, 2003).
Staffing Personnel within the focal organisation need to be assigned to deal with specific relationships. The personnel must be aware of the cross relational issues in order to be best placed to deal with them.
Controlling The final task is the ability to control the final output of the network. This competence refers to the extent of which the higher management influence decisions lower down the chain.

Network Management Qualifications
To be able to effectively perform the relational tasks, an organisation must have a certain level of understanding. When considering technological networks there is a minimum level that an organisation must reach to be able to operate within it. Ritter and Gemunden describe these competencies as the qualifications an organisation possess.

Specialist Qualification Specialist qualifications are those specific to the type of network an organisation operates within. To be able to operate effectively within a technical network, a certain level of technical understanding is required. Some networks may call for detailed understanding of certain aspect of law where others may call for a deep understanding of economics. Some more creative networks may require a deep understanding of the perceptions of particular market segments (Ritter and Gemunden, 2003).

Not all areas of understanding will have formal qualifications associated with them. This competence includes the tacit concept of experiential knowledge.

Social Qualification The Social Qualifications competence represents the softer skills needed to maintain effective relationships. It includes several dimensions such as communication ability, extraversion, conflict management skills, empathy, emotional stability, self-reflection, sense of justice, and cooperativeness (Ritter and Gemunden 2003).

The network competence framework in its entirety, offers the ability to formally measure an organisation’s network management capability. By using this model within the context of the network theory discussed within this section, a new perspective emerges. This can greatly assist with creating management strategy, increasing understanding of a complex market and most importantly for the purposes of this research study, potentially assist with the management of innovation.

Resource Name Description Resource Type
Innovation Management a Network Perspective I submitted this study as part of my requirement for completing an MSc in Management at the University of Bath. It looks at the overlap between innovation management and network management theory. The results of the literature review were applied to a programme of innovation projects from within the UK Ministry of Defence. The results highlight a strong link between network management capability and innovation success and examines the implications for wider management and academic theory. Presentation
Innovations and Networks Combined This m2i webpage looks at the overlap between Innovation Management and Network theory. M2i wiki page
Network Theory This the main page from with wiki on Network Management.,

The concept of looking at a department or entire organisation as a node within a much wider network. This perspective can help you appreciate the mammoth task of trying to influence this. || M2i wiki page