The past and the future of business marketing theory

Industrial Marketing Management 42 (2013) 394–404

Contents lists available at SciVerse ScienceDirect

Industrial Marketing Management

The past and the future of business marketing theory Ian F. Wilkinson a, b,⁎, Louise C. Young b, c, 1 a b c

Discipline of Marketing, University of Sydney, Australia Department of Entrepreneurship and Relationship Management, University of Southern Denmark, Denmark School of Business, University of Western Sydney, Australia

a r t i c l e

i n f o

Article history: Received 1 December 2011 Received in revised form 1 July 2012 Accepted 1 August 2012 Available online 6 March 2013 Keywords: Complex adaptive systems Business relations and networks Dynamics and evolution Agent based models Mechanisms

a b s t r a c t A complex systems approach to understanding and modelling business marketing systems is described. The focus is on the dynamics and evolution of such systems and the processes and mechanisms driving this, rather than the more usual comparative static, variables based statistical models. Order emerges in a self-organising, bottom up way from the local or micro actions and interactions of those involved. We describe the development of our thinking regarding this approach and its main features, including the development of agent based simulation models and the identification and modelling of underlying mechanisms and processes. We conclude by discussing the implications of this approach for business marketing theory and research. © 2013 Elsevier Inc. All rights reserved.

1. Introduction Patterns in business history matter, as they provide key insights into the way business systems operate and are an important determinant of their present and future. Our research has for many years focussed on uncovering these patterns. Exploration of the evolution of ideas that guide research and of their evolutionary paths is similarly beneficial (Wilkinson, 2001). This article describes the ideas that have underpinned our theories of business marketing systems in part by considering the way they have evolved. We believe we are now at a major transition point in the social sciences, which sees exploration and explanation of the deeper processes of evolution as central to scientific advancement. Advances in computing power and software enable us to tackle these important issues of theory in ways that were not previously possible. We can now test far more realistic theories and models of market systems that deal with the inherent complexities, dynamics and processes of social systems, including market systems. Business markets are complex adaptive systems in which order emerges in a self-organising, bottom up way from the actions and interactions of people and firms and other types of organisations

⁎ Corresponding author at: Discipline of Marketing, University of Sydney Business School, The University of Sydney, Sydney, NSW, Australia, 2006. Tel.: +61 2 9036 7610; fax: +61 2 9351 6732. E-mail addresses: [email protected] (I.F. Wilkinson), [email protected] (L.C. Young). 1 School of Business, University of Western Sydney, Locked Bag 1797 Penrith South NSW 1797 Australia. 0019-8501/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.indmarman.2013.02.007

involved. Control and power are distributed through networks of interconnected, interdependent business actors. This challenges traditional notions of management and actor-centred theories of performance. A person or firm's behaviour and performance cannot be understood as simply the product of its own resources, skills, competences, orientations and motives. Actors operate in the context of other people and firms with whom they are interconnected in various ways. Behaviour and performance depends as much, if not more, on what others do, believe and want than on their own resources, skills, competences, orientations and motives. Context matters. Context is created by the history of past interactions, interconnections, events and the like. Increasingly we recognize that the study of the history of business systems provides insight into their present state and possible future(s) (Young & Bairstow, 2011, 2012). Similarly, consideration of our evolution to a “complex systems science” framework for understanding and researching business markets provides insight into the nature of context, the value its study can provide and ways of effectively researching it. This article is organised as follows. First, we describe what we mean by a complex adaptive systems' view of marketing, and business marketing in particular, and its relevance for advancing theory and research. Next, we briefly trace our intellectual journey towards this perspective. We describe some of the main ideas we have encountered and the way these have shaped our thinking and research within a complex systems science approach. In the final sections we consider the implications of complex systems thinking for business marketing theory and research as well as for management and policy.

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

2. A complex systems view of business marketing 2.1. Complex systems thinking Complex systems science is a way of thinking and doing science that revolutionises the way we can understand and model the behaviour of complex adaptive systems like firms, markets and business generally (Allen, Maguire, & McKelvey, 2011; Epstein, 2006; Jorg, 2011). Complex adaptive systems (CAS) are characterised by distributed control, self-organisation and emergent behaviour. In complex systems the overall behaviour and structure of a system emerges in a bottom-up, self-organising way from the local or micro actions and interactions taking place over time among networks of interconnected actors in an environment. They are sensitive to starting conditions and are non-linear in the sense that small changes can have disproportional effects. There are also top-down feedback effects in which large scale patterns of behaviour emerging, such as price levels, industry structures and trends, affect local actions and interactions. People, households, firms, markets, supply chains, distribution systems, business relations and networks are all CAS interacting and adapting to each other. They are adaptive because the rules governing behaviour are not fixed but evolve over time in response to the experience and outcomes occurring and due to environmental effects. Market systems have long been recognized as complex and adaptive. As Wroe Alderson and Reavis Cox, two of the founding fathers of modern marketing theory, stated in their foundation work: “a market changes day to day through the very fact that goods are bought and sold. While evaluation is taking place within a marketing structure, the structure itself is being rendered weaker or stronger, and the changes in organization which follow will have an impact on tomorrow's evaluations. Marketing theory will not provide an adequate approach if it ignores this interaction between the system and the processes which take place within it” (Alderson & Cox, 1948 p. 151). However, sixty-five years ago there was not the capability to meaningfully research the interacting processes of complex business systems. Reflecting the research capabilities of the time, linear, comparative static, variables based, statistical models of market systems in which time and process are largely absent came to dominate research in marketing and system–process interactions were studied using heroic assumptions or were ignored. Complex systems models and theories call for an additional type of explanation to extend the variables based models (which seek to explain variance and co-variance) that we are used to. This form of explanation is described by Herbert Simon (1968): “To ‘explain’ an empirical regularity is to discover a set of simple mechanisms that would produce the former in any system governed by the latter.” (p.445). The term “mechanism” is often used informally to suggest a causal explanation, a reason for something happening. Campbell (2005) describes them as the processes that explain the causal relationships among variables. Hedström (2005) offers a more complete definition: “mechanisms consist of entities with their properties and the activities that these entities engage in, either by themselves or in concert with other entities…a constellation of entities and activities that are organized such that they regularly bring about a particular type of outcome.” (p. 25). Mechanisms are not variables but they can help explain why patterns of covariance occur among variables measuring different aspects of a system and the results of experiments. Researchers constructing variable-based models or designing experiments refer to underlying mechanisms and process in order to formulate their theories and hypotheses and design their experiments. But a variables-based statistical model or an experiment, no matter how well measured and designed,

395

does not deal directly with the underlying driving mechanisms and processes. Variables are only the manifestations and reflections of their workings. Managers operate in a world of mechanisms and processes. They do not manage variables; they manage people and processes, resources and money. They do this using various mechanisms of control, influence and communication and they assess their own and others' behaviour by taking measures of the behaviour and outcomes occurring. Measuring is another type of mechanism. In a later section of this paper we will describe the main types of mechanisms driving the behaviour and evolution of business relations and networks. There are signs of growing interest in marketing in complex systems thinking and methods. More articles are appearing using such concepts and describing their relevance and use (e.g. Easton, Brooks, Georgieva, & Wilkinson, 2008; Rand & Rust, 2011; Watts & Dodds, 2007; Wilkinson & Young, 2002). 2.2. Researching complex adaptive systems Complex systems science is not only a different way of thinking about the way socio-economic systems behave, it also involves a different type of methodology. There are two approaches to doing science: “Collect observational, survey or other forms of data and analyze them, possibly by estimating a model; or begin from a theoretical understanding of certain social behavior, build a model of it, and then simulate its dynamics to gain a better understanding of the complexity of a seemingly simple social system” (Liao, 2008, p. ix). The kinds of simulation models built to study complex adaptive systems are known as “agent-based models” (ABM) because they start with the behaviour of individual agents or actors in the system and the way they act and interact and these “grow” evolving systems. These types of models increasingly surround us in social media, games, and other areas of science. Such quasi-ecological models are necessary because the behaviour of a complex system cannot be reduced to the behaviour of its parts in any additive way. The parts are interconnected and interdependent and time and order effects matter. Sometimes this type of research is referred to as “computational social science” or “computational economics” because complex system models are not amenable to traditional closed form mathematical solutions. The underlying equations of motion of such systems, which do exist, are far too complex and nonlinear to solve (Borrill & Tesfatsion, 2010; Leombruni & Richiardi, 2005). The only way to solve such equations is to count them out using a computer and examine what happens under different conditions. The mathematical equations defining the rules of behaviour of the systems have to be programmed into the computer in the form of if-then conditions. ABM of complex systems are not models of the behaviour of variables but of actors acting. Computer simulation models are models of the mechanisms and processes underlying the behaviour of all the actors involved, including animate and inanimate or passive actors, like geographic space, resources and material things. Variables do not exist in the real world, only in the minds and models of researchers. Variables result from our measuring behaviour and can be used as an ancillary to ABM. Just as we measure the behaviour of real business markets, we can also use the same methods to measure the behaviour of our computer simulation models. We use variables based models of our complex system simulations to analyse, test and understand what is going on, as we do in real world models. ABM enable us to study the behaviour of complex systems in ways that we cannot do in the real world. In effect, the real world is a sample size of one; there is one history — life as it has been. But computer models allow us to build models of life as it could be, including types of business and marketing life. One of the founders of complex systems research, Chris Langton, explains it this way: “We trust implicitly that there are lawful regularities at work in the determination of this set [of realized entities], but it is unlikely that we will discover many of

396

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

these regularities by restricting ourselves only to the set of biological entities that nature actually provided us with. Rather, such regularities will be found only by exploring the much larger set of possible biological entities” (Langton, 1996 p. x). The analysis of behaviour and outcomes of computer simulation models requires careful experimental designs in order to tease out and understand the effects of different conditions. With simulations it is easy to perform numerous experiments that would be impossible to do in the real world (and there is no need to get ethics approval). Computer simulations can also be linked to real world experiments in order to inform and test them. This approach is well illustrated in recent experiments to model the evolution of trade and specialisation: “In general, the power of combining agent-based models and experiments is exploited as follows: (1) perform human-subject experiments, (2) design agents that replicate human outcomes, (3) perform additional simulations employing the same agents in a new environment — this yields predictions for human behavior, (4) perform new experiments, (5) test agent-based predictions.” (Kimbrough, 2011 p.491). This kind of science is comparable to the way biologists conduct Petri dish experiments. They bring together the necessary ingredients and starting conditions, including controlled environmental conditions, and observe what happens under different experimental conditions. In the same way in computer simulation experiments the initial conditions, parameter and environmental conditions are set up and simulations are run under different conditions to see the effects they have. Measures are recorded over time of the behaviour and outcomes resulting, which are then subject to statistical analysis. The advantage of simulations is that the researcher has complete access to all the data, and any measures they can think of and achieve a 100%, unbiased response rate. For this reason we need to be careful in comparing the results of computer simulations with real world measures, unless we build in non-response and other biases that affect the results of surveys of real people and organisations. The possibilities of such models are dramatically illustrated in the visionary European Union (EU) Research Framework project FuturICT (www.futurict.eu), which aims to build a Living Earth Simulator of the world. This is the way they introduce the project: “We are surrounded by systems that are hopelessly complex, from the society, a collection of seven billion individuals, to communications systems, integrating billions of devices, from computers to cell phones that form our information and communications technology, or ICT. Many of these systems appear random to the casual observer, but upon closer inspection they are found to display endless signatures of order and self organization. These systems are collectively called complex systems, and the research that to explore and understand them is often referred to as complexity. Given the role they play in our life, their understanding, quantification, prediction and eventual control is the major intellectual scientific challenge of the 21st Century.” (Bishop, Helbing, Lukowicz, & Conte, 2011, p. 34.) Such vision is receiving substantial support from governments – the pilot for this work was funded by an EU research grant of 2 million Euros – and many scientists spanning a variety disciplines. Other examples of large scale complex systems simulation models are: the Los Alamos National Laboratory Epidemiological Forecasting Simulation Model of the USA, EpiCast, which has 300 million agents, one for each person in the USA; and a model of banking networks developed by Lord Robert May and his colleagues that shows, paradoxically, how the banking network as a whole becomes increasingly unstable as each individual bank diversifies its portfolio to spread its risk in similar ways (Haldane & May, 2011). There are only a few examples thus far of business market related complex systems simulation models. These include models of the electricity wholesale markets in the US (Somani & Tesfatsion, 2008);

distribution and the supply chain models developed by Kauffman and others at the BIOS Group for P&G (Seibel & Kellam, 2003); and the firm AntOptima SA, which uses a model based on social insect behaviour to optimize their distribution scheduling (www.antoptima.com/ site/en/index.php). To build realistic computer simulations of business markets, we need to understand and be able to model the various mechanisms and processes that underlie individual and organisational behaviour. We also need to collect information about the behaviour of real business markets that we can use to help validate and test our models. Existing research results obviously have a role to play but additional types of research are also called for, as we describe in later sections. 3. The development and application complex systems thinking in business marketing In this section we describe the development and application of complex systems thinking in business marketing, primarily using our own research to illustrate this. The broad pattern of development reflects Sawyer's (2005) description of the three waves of social systems thinking. The first wave is reflected in structural–functional theories of society and the research of Talcott Parsons. In marketing this is associated with the functionalist approach to marketing theory advocated by Wroe Alderson (Alderson, 1957, 1965; Alderson & Cox, 1948). These ideas had a profound influence on our thinking about the nature of marketing systems and processes (e.g. Dixon & Wilkinson, 1982, 1989). The second wave is derived from general systems theory, which began in the early 1960s, and entered our thinking through the influence of various people, including in particular the writings of Kenneth Boulding (1970, 1978, 1985), Don Dixon (Dixon & Fisk, 1967; Dixon & Wilkinson, 1982), Fred Emery (Ackoff & Emery, 1972; Emery, 1969; Emery & Trist, 1965), George Fisk (1967), Jay Forrester (Forrester, 1958; Forrester, 1961, 1971, 1980) and Roger Layton (1981a,b, 1985, 1986). This directed initial modelling attempts, the direction of which was augmented by chaos theory. Along the way we have been involved in many other types of business market research, including surveys and case studies, which influenced our thinking and the way it developed, eventually leading us to research in complex systems science. Early models of the dynamics of marketing channels were based on Jay Forrester's models of industrial dynamics (Wilkinson, 1989). The pattern of orders, products and advertising over time in a multilevel distribution channel was modelled using simple difference equations. They showed how the equilibria in such models were not necessarily stable and “whiplash” or “bullwhip” effects occurred as firms over and under reacted to orders received due to lags and limited information and as a result of the effects of advertising expenditures, which were based on order levels in the previous period. Such models are at the heart of the dynamics of the Beer Game developed at MIT by John Sterman (1989) which is used to give managers and business students a sense of operating in a complex system in which they are not in charge. Prigogine's (Prigogine & Stengers, 1984) concept of dissipative structures led to a reorientation of thinking about the nature of marketing system structure and structuring (Wilkinson, 1990). Marketing systems can be viewed as dissipative structures, on-going processes of action and interaction among and within firms and households which both shape and are shaped by the networks of relations that exist. This was incorporated with ideas from chaos theory, leading to a model that showed how simple feedback models of competition could result in complex dynamics under plausible parameter settings (Hibbert & Wilkinson, 1994). The model is summarized in Fig. 1. Firms gain a share of market expenditure in each period based on their relative marketing effort and their intrinsic attractiveness. The amount they change their marketing effort in the next period is a function of the profit they earn in the current period, i.e. revenue minus marketing

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

397

Fig. 2. Connected transactions. Fig. 1. A dynamic model of market share.

costs, which in turn affects their subsequent market share. We analysed the case of a single brand, a small brand in a large market and two brands interacting and show how the dynamics varies in response to changes in parameter values within plausible limits. These are examples of complex systems models but the models are extremely simple. They are based on only a few market actors interacting, using spreadsheet programmes and systems dynamics software like Stella and IThink. More complex models were needed that involved interaction among many actors and where the rules of behaviour could evolve in response to the behaviour taking place, as well as in response to exogenous events. This led naturally into the world of complexity and complex systems science. Working with others, our first attempts to model industrial networks as complex systems focused on the generic NK models of networks developed by Stuart Kauffman (1992). Here N refers to the number of actors in a system and K to the number of other actors whose behaviour affects a focal actor. Business networks can be analysed as NK models in which the behaviour and performance of each individual firm is affected by K other firms, such as their suppliers, customers, competitors and complementors. Instead of focusing on the way firms were influenced by other firms in a network we focused on the way the connections between actors are themselves interconnected and change over time. The models make the market transaction the basic unit of analysis and examine how market transactions between actors depend on other market transactions. For example if A sells a product to B in one period, A had to have bought the necessary inputs in previous periods, which involve transactions with competing or complementary suppliers (see Fig. 2). Similarly, B needed to earn money or make goods before it could trade with A. The rules by which transactions are connected to other transactions can be interpreted in terms of various types of Boolean rules (the conditions that dictate whether connections change or stay the same in the next period). As these rules play out over time, different patterns of connections among actors arise and structure can change. In a series of articles, NK models of this type have been used to study the dynamics of network structures in business markets (Easton, Wilkinson, & Georgieva, 1997; Easton et al., 2008; Wilkinson, Wiley, & Lin, 2010). In more recent work we modelled the interactions among groups of actors in a system (network) in terms of iterated games among each pair of actors in the system. The iterated games include prisoners' dilemma as well as other types of games that reflect different types of interdependence among the actors. The agents in these models are the strategies followed by actors, which indicated what they would do depending on the past behaviour of the other player in the iterated game. The strategies evolve over time based on their individual or group level performance, which are ways of assessing their fitness. The technical details are not relevant here. We used the model to explore the effects of individual versus group level selection of

performance i.e. evolving the groups based on the best performing individual strategies in different groups or the best groups of strategies. We were able to show that for many types of games, where there are conflicts of interest, group level selection not only produces better performing groups but also the highest performing individual strategies. Group level selection led to the evolution of productive ecologies of strategies in groups (Ladley, Wilkinson, & Young, 2007; Ladley, Wilkinson, & Young, 2011; Wilkinson, Young, & Ladley, 2007). 4. Implications of complex systems research for business marketing The research described above has helped us to understand the complex dynamics and evolution that characterise business relations and networks. The work highlights that ongoing change is a nearinevitability in highly connected network models and indicates the need to build further theories to explain the underlying drivers of change. In addition this work highlights the need for continuing development of methods that allow meaningful exploration of complex systems to guide the design and validation of models. We have examined these issues in various publications that address the implications of complex systems thinking for marketing theory, research, practice and policy. These are reviewed in this section. 4.1. Implications for theory A general model of the dynamics and evolution of business relations and networks is shown in Fig. 3 in terms of the relation between two organisations, which can be extended to include relations with other organisations. There are coupled feedback loops for each organisation between the ongoing experience and outcomes of the actions and interactions taking place over time and the structure of the relation. Relationship and network structure is co-produced, reproduced and changed over time through actions and interactions taking place. Interactions of various kinds shape relationship and network evolution. These include the co-development in and across relations of: (a) actor bonds such as trust, power-dependence and commitment through the experience of trading and interacting; (b) activity links through operational and strategic co-adaptations in relations over time; (c) resource ties, through mutual and asymmetric investments in relationship specific assets; and (d) schema couplings, as actors adapt their ideas, goals, plans and theories in use regarding themselves and others over time (Håkansson & Östberg, 1975; Håkansson & Snehota, 1995; Welch & Wilkinson, 2002). These relationship interaction processes take place subject to starting conditions such as prior histories and interactions and differences in resources and positions, as well as in the way relations are connected to other relations both directly and indirectly and embedded in more general contexts including markets, technologies and environment.

398

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

Fig. 3. Relationship development processes.

To some degree this is reflective of the models of other researchers, including those within the Industrial Marketing and Purchasing (IMP) group. Largely missing from that work however is consideration of generic processes and drivers of change, i.e. mechanisms of business evolution. In recent work with others we have carried out a comprehensive multi-disciplinary review of previous research regarding business relations and networks and have identified five broad types of mechanisms and processes underlying their dynamics and evolution as well as the way these mechanism and processes can be modelled (Held, 2010; Held, Wilkinson, Marks, & Young, 2010a,b; Wilkinson, Held, Marks, & Young, forthcoming). The review covered disciplines including biology, ecology, physics, engineering, artificial intelligence and the social, economic and business sciences. The results are summarized in Table 1. A full list of references is available in the interim research grant report (www.agsm.edu.au/bobm/ARC/ARC_DP0881799_Interim_Report.pdf). The first group of mechanisms depicted in Table 1, business acting and specialising, refer to the activities actors perform in markets in order to produce, use, buy and sell products and services. They specialise in performing some activities themselves and rely on others to carry out other activities and they access the fruits of others' labour through various types of market and other types of exchange mechanisms. The patterns of specialisation and exchange develop in a society and form the structure of the economy and business system. Firms make decisions on making or buying goods and services (Wilkinson, 2008), they work alone and with others to identify profitable opportunities for trade and specialisation that have to be discovered through research and interacting with others. (Kimbrough, Smith, & Wilson, 2008). Business relations and networks emerge and evolve as result of variations in offers and capabilities and the mutual learning taking place and as opportunities arise due to activity within and beyond the system. Firms also go out of business and the people and resources move elsewhere (Wilkinson, 2008). The actions of firms in producing and consuming things have been modelled in terms of various input–output or production functions. Models of the mechanisms of specialisation are scarce, although economic theories of comparative advantage, negotiation, exchange, scale and scope efficiencies, and transaction-cost theory point to the existence of various factors affecting decisions to trade and specialise. Exceptions are a model of the development of wholesale banking, which arises as a result of the way scale and scope efficiencies and transaction costs affect trading patterns over time among a network of actors (Sapienza, 2000) and a dynamic model of role differentiation in social networks (Eguíluz, Zimmermann, Cela-Conde, & Miguel, 2005). As noted before, experiments have also been carried out to simulate the emergence and evolution of trade and specialisation in artificial economies (Kimbrough et al., 2008), which have led to the development of ABM that are able to reproduce the essential features of the mechanisms and processes involved (Kimbrough, 2011).

In addition to the specific mechanisms related to production and trade, there are more general mechanisms that underlie a wide range of behaviour – including learning, choosing, evaluating, communicating and imitating – that have received far more modelling attention. Business mating mechanisms guide the ways potential trading partners encounter each other and choose, get chosen, accept or refuse to do business and develop longer lasting relations. This can vary from random processes to ones that are influenced by past interactions, predispositions and communication networks. Changing or keeping partners depends on the evaluation, choice and learning processes of each party as well as the recommendations and referrals of third parties. There are many examples of partner search and mating mechanisms in simulations, as it is one of the core mechanisms necessary to design a network simulation. Random pairing can be used to represent well mixed networks (Erdös & Renyi, 1959) or preferential attachment based on the number of existing links can be used, i.e. the rich get richer rule (Barabási & Albert, 1999). Preferential attachment has been extended in many ways. Agents can become less attractive with age (Dorogovtsev & Mendes, 2000), form links based on their activities (Fan & Chen, 2004) or performance (Ren, Wu, Wang, Chen, & Wang, 2006) or prior successful cooperation (Gilbert, Pyka, & Ahrweiler, 2001). The formation of relationships can be linked to actors' mutual preference and expectations (Tesfatsion, 1997), or by copying the relations of others (Vázquez, 2000). Iterated games, such as iterated prisoners' dilemma, with choice and refusal of partners have been used by Tesfatsion and others to model the evolution of trading networks, where actors make and accept, or not, offers to trade based on previous experience (LeBaron & Tesfatsion, 2008; Stanley, Ashlock, & Smucker, 1995). Attraction and gravity models originating from geography are also relevant (Allen & Sanglier, 1979; Clarke & Wilson, 1983), as is Schelling's (1971) classic model of the processes of segregation based on actors' preferences to stay or move in to places where there are a minimum percentage of similar actors. Business Dancing mechanisms guide the nature of actors' interactions and their evolution. When there is no relationship history, actors seek information about potential parties through reputations, stereotypes and observation. Over time the experience and outcomes of business and social interactions and exchanges result in mutual learning and a store of knowledge about others. However the available network positions, resources, attractiveness, perceptions, beliefs and evaluations of all those involved continue to evolve, thus leading to varying amounts of continuing uncertainty and ongoing relational evaluation (Young, 2006). Repeated interactions produce, reproduce and change a relationship atmosphere resulting in the evolution of activity links, actor bonds, resource ties and schema couplings and the relationships between them — as described in the previous section. Models of relationship interactions have focused mostly on various types of iterated games, often using the prisoners' dilemma game (Marks, 1989; Tesfatsion,

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

399

Table 1 Business relations and network mechanisms and models. Source: Based on Wilkinson et al. (2012). Mechanism types

Example mechanisms

Example models

1. Business acting and specialising

Producing, consuming, buying, selling, learning, copying, targeting, in-sourcing, outsourcing, innovating, firm creation and demise

2. Business mating

Finding, being found, attracting, homophily, repelling, choosing, being chosen

3. Business dancing

Interacting, exchanging, cooperating, defecting, responding, initiating, trusting, liking, committing, learning, adapting, terminating

4 Networking: Connecting business relations

Enabling and constraining effects of other relations, comparing, accessing, prioritising

5. Environmental impact

History, enabling and constraining effects of exogenous environment

-Economic models of scale and scope efficiency, transaction cost and production functions (Dixon & Wilkinson, 1986) -Choice and evaluation models (Mosekilde, Larsen, & Sterman, 1991) -Imitation and learning models (Brenner, 2006) -Specialisation (Eguíluz et al., 2005; Sapienza, 2000) -Central place and gravity models (Allen & Sanglier, 1979; Clarke & Wilson, 1983) -Trade and network formation models (Bianconi & Barabási, 2001; Kimbrough, 2011; Watts & Dodds, 2007) -Attraction and gravity models (Allen & Sanglier, 1979; Clarke & Wilson, 1983; Huff, 1964; Huff, 1964) -Matching, alliance models (Gavrilets, Duenez-Guzman, & Vose, 2008) -Partner and trade choice models (Li & Rowley, 2002; Sun & Tesfatsion, 2007; Wilhite, 2001) -Iterated game models (Axelrod & Hamilton, 1981; Tesfatsion, 1997; Zimmermann, Eguíluz, & San Miguel, 2004) -Bargaining models (Debenham & Sierra, 2007; Kimbrough et al., 2008; Kirman & Vriend, 2001) -Trust models (Kim, 2009; Tomassini et al., 2010) -Evolution of cooperation models (Henrich, 2004; Ladley et al., 2007; Pennisi, 2009) -Learning models (Brenner, 2006) -Attraction and loyalty models (Kirman & Vriend, 2001) -Learning models (Brenner, 2006) -Competition and trade models (Axtell, 2005; Stanley et al., 1995) -Diffusion models (Goldenberg et al., 2001; Simoni, Tatarynowicz, & Vagnani, 2006) -Recommendation engines (e.g. Amazon, Google) -Network co-evolution models (Ehrhardt, Marsili, & Vega-Redondo, 2006; Fronczak, Fronczak, & Holyst, 2006; Gross & Blasius, 2008) -Starting conditions -Parameter settings -Location mapping -Environment processes (Axtell et al., 2002)

1997). Iterated game models have also been used to examine the evolution of cooperative groups and network structures under conditions of individual and group selection (Henrich, 2004; Ladley et al., 2007; Pennisi, 2009). Lastly, some models have been developed regarding the nature and development of trust in markets and networks (Kim, 2009; Tomassini, Pestelacci, & Luthi, 2010) and loyalty (Kirman & Vriend, 2001). Networking mechanisms guide the way the outcomes and experience of interactions in a focal relation impact – positively and negatively – upon other relations in which the organisations are also involved (e.g. Anderson, Håkansson, & Johanson, 1994; BlankenburgHolm, Eriksson, & Johanson, 1996; Young, Wiley, & Wilkinson, 2009). What happens in one relation may well affect others over time, through changes to their network connections (Young & Bairstow, 2011, 2012) and more directly through changes to the image and reputation of those involved, and thus will affect subsequent relationship formation. The connections among relations are due to the operation of various kinds of mechanisms, including comparison made across relations and the way exchanges are interconnected positively or competitively in business networks such as in supply and value chains and distribution channels (Easton, Wiley, & Wilkinson, 1999; Easton et al., 1997; Easton et al., 2008). Communication and diffusion processes also lead to the spread of information and ideas through networks, leading to learning and adaption. Models have been developed in which actors change partners based on learning and comparisons (Brenner, 2006; Stanley et al., 1995), network position and competition for trade partners (Axtell, 2005). Diffusion models also model the way information flows across networks (e.g. Goldenberg, Libai, & Muller, 2001). Environmental impact mechanisms constrain and enable what actors embedded in socio-cultural, economic and physical environments can and cannot do. These environmental forces can be included in simulation models in various ways. Parameter values, starting conditions and known environment histories can be imported into the simulation

models. Use can be made of actual geographies and eco-systems. This approach is exemplified in modelling the evolution of the Anasazi tribe over hundreds of years, in which the modellers included known weather, geological, population and production yield patterns (Axtell et al., 2002). The mechanisms operating in the environmental systems also can be directly modelled in the simulation. The mechanisms in Table 1 work together over time, in ways that are not yet well understood, leading to the emergence of macro outcomes, structure and processes in terms of types of relationships, network structures, performance and broad patterns of change and evolution (Wilkinson et al., forthcoming). 4.2. Methodological implications To inform, guide and validate ABM of the behaviour, dynamics and evolution of business markets that will be of use to researchers, practitioners and policymakers we need to develop various types of research, in particular: • systematic case histories to describe the sequences and interconnection of events and mechanisms taking place over time in business relations and networks, including methods that allow analytical transparency, • the development of databases recording the patterns of change in business relations and networks over time. More systematic in-depth case histories of the dynamics and evolution of business relations and networks are required in which the sequence of events taking place and the ways they are interconnected by the operation of different mechanisms and processes are articulated (Buttriss & Wilkinson, 2006; Van de Ven & Engleman, 2004). Such case histories are more than storytelling; they are designed to identify the key actors involved, trace their actions and interactions over time; identify what is going on in the environment and the effects this has; parse the foregoing into sequences of interconnected events and how they are

400 I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

Fig. 4. Event structure map for phase of Australian IT industry. (Source: Bairstow & Young, 2012).

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

linked or not; how they aggregate to collective behaviour and identify the mechanisms connecting events. The explanation of such connections is “one of the most fundamental and pervasive questions in all of science” (Watts, 2003 p. 24). An example of such a case history concerns the evolution of the IT distribution network in Australia over many years (Bairstow & Young, 2012). The narrative map of events for one phase of the evolution is shown in Fig. 4. This provides an overview of the sequence and nature of events over time in the general environment and the Australian IT channel and highlights the sources and level of conflict in relations (in this case conflict and its resolution or lack of resolution are key mechanisms). This depiction of broad patterns can guide a more detailed event articulation and structure analysis. Other examples of systematic case histories in business marketing include: the evolution of an organisation to becoming an e-business (Buttriss, 2009); the development of trust in a business relation over time (Huang, 2010); and the process of discovery of international business opportunities (Chandra, Styles, & Wilkinson, 2012). In addition to in-depth systematic histories of business relations and networks, we need better data on the way business relations and networks change over time. There are a few examples of detailed longitudinal research regarding the development of business relations and networks, or indeed any kinds of networks (Powell, White, Koput, & Owen-Smith, 2005) that would be suitable for validating ABM. There are some detailed snapshots of business networks at points through time in particular industries in terms of types of alliances and patterns of trading that can be used to develop stylized facts to verify ABM (Browning, Beyer, & Shetler, 1995; Konno, 2009; Powell et al., 2005; Rosenkopf & Schilling, 2007). The Japanese, in particular, have been collecting trading data between individual firms over time for particular industries but not always in a form suited to network analysis and modelling, though it can be adapted for this use (Luo, Whitney, Baldwin, & Magee, 2011) There are signs that more research is beginning to be conducted to gather and model data of this type. For example, a more detailed history of the evolution of Norwegian fishing industry networks was used to inform the development of an ABM of its dynamics (Følgesvold & Prenkert, 2009). And a major EU research project is currently underway to measure and model industrial ecologies and innovation networks in different regions, which involves mapping the ways business networks have evolved over time (http://cress.soc. surrey.ac.uk/skin/). There are also commercial data bases that are relevant. For example, we are working with a large commercially developed database of multiinformant measures of advertising agency–client relations in various countries over time by the APRAIS firm, which assists such firms in creating more cooperative relations. More data of this kind needs to be located or collected if we are to have a sound base for validating and verifying our ABM. Lastly, at a more macro level, data regarding the pattern of trade among nations for different types of products and services over time is available and can be used to develop and calibrate models of the evolution of trade networks at the national level. To augment interpretation of these models with their real world data, we are working to develop statistical measurement techniques whose assumptions are compatible with complex phenomena, which most mainstream statistics are not (Thompson & Young, 2012). 4.3. Implications for management and policy The implications of complex systems thinking for management and policy have been explored in a number of articles (Denize & Young, 2007; Ritter, Wilkinson, & Johnston, 2004; Wilkinson, 2006; Wilkinson & Young, 2002; Wilkinson & Young, 2005; Young & Denize, 2008). Here we focus on a few of the main points only. The behaviour and performance of complex systems, including business relations and networks is coproduced through the actions and interactions taking place over time among the actors involved. No one actor is in overall

401

control, though some have more influence than others, and it is impossible to trace outcomes to the acts of individual actors. As James March (1996) describes it: “An organization reacts to the actions of others that are reacting to it. Much of what happens is attributable to those interactions and thus is not easily explicable as the consequence of autonomous action.”(p. 283). This makes the task of management more difficult. Management in complex systems is not about directing and controlling others but more about responding and participating and developing flexible, adaptive strategies. The principle of requisite variety (Ashby, 1958) states that, in order to cope and respond to its environment a system has to be able to match its complexity or variety. The relevant unit of action and response is not the individual firm but the network itself, because this is collectively more intelligent than the individual firms comprising it and has a greater variety of knowledge and potential responses. An important focus of strategy becomes that of developing collaborative advantage (Kanter, 1994) through the development of cooperative relations with other organisations in the business ecosystem of which it is a part and positioning itself in these systems in ways that help underpin competitive advantage (Iansiti & Levien, 2004; Wilkinson & Young, 2002). Firms collaborate to compete – to gain access to key resources and information – and they compete to collaborate with others that are best able to provide needed resources and information (Wilkinson, 2008). The relations and networks in which a firm is embedded, its business ecosystem, can be seen as an extension of the firm, an extended enterprise, by which firms gain access to key resources and information and extend their sensing, action and response potential (Davis & Spekman, 2004; Dyer & Singh, 1998; Håkansson & Snehota, 1995) Their role is exemplified in various management concepts, including: Lane and Maxfield's (1996) concept of generative relations, which shape the way a manager senses their environment; the building of relations with leading edge users to increase innovation potential (von Hippel, 1986; von Hippel, 2005); the concept of transformational outsourcing (Linder, 2004), and in the role played by relations and networks in identifying international market opportunities (Chandra et al., 2012). A way of understanding the kinds of strategies firms need to develop in complex systems is in terms of Clark's (1997) concept of softassembled strategies. Although this was developed to help explain individual behaviour and cognition we believe it can usefully be applied to business. The extended enterprise idea is similar to the concept of embodied mind in cognitive science. The brain, body and local environment are viewed as a form of extended mind and body that works together in sensing, thinking and responding to the problems confronting us. There is no central direction from the brain based on some model of the system that controls and directs all the action. Instead, we sense, think and act with and through our body and local environment by learning from them and using their intrinsic response tendencies. Think of the way a blind person incorporates a white stick as part of the sensing and responding self. In a similar way the manager, firm and its relations and networks form an extended mind and body that work together in sensing, thinking and responding to the problems confronting it. Over time managers learn from them and how to utilise their intrinsic response tendencies to develop and adapt its behaviour. By learning to soft assemble strategies in this way managers do not need to try to take into account, explicitly, all the direct and indirect interactions affecting outcomes. This would be impossible and, even if possible it may well be counterproductive because it would make the system more richly connected, unpredictable and chaotic. This problem has led (Kauffman, 1994) to suggest the idea of optimally myopic strategists; not because of computation costs but because if we become too clever in our strategizing we tend to transform the world in which we are adapting into one that is more chaotic.

402

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

Complex systems and relations and network thinking also have implications for government policy. Working with others we have examined this issue in three main ways. First the role of the political actor in industrial networks has been examined, showing how government agencies and policy makers are not outside the business systems they seek to monitor and control; they are part of the complex system (Welch & Wilkinson, 2004). How they participate and respond, including the rules they develop and the way they are implemented and the relations they have with business and other political actors, affects the way the system behaves and performs — but in ways that are difficult to predict. Second, we examined the implications for trade and industry policy from a business relations and network perspective, as opposed to individual firm characteristics. One part of this research involved developing a way for government to target firms for assistance based on network positions and the international competitiveness of networks in which they operate (Wilkinson, Mattsson, & Easton, 2000). In other research we examined the extent to which governments can facilitate the development of internationally competitive business networks through a different type of market led grouping scheme (Welch, Welch, Wilkinson, & Young, 1996a,b, 1998). Third, the implications of networks and complex systems thinking for antitrust policy have been examined (Wilkinson, 2006). Complex systems theory focuses attention on the structure, dynamics and evolution of business relations and networks and how they affect the competitiveness of firms and industries over time. Studies of the properties of networks and how they evolve challenge some of the assumptions underlying antitrust policy, including the existence and potential benefits of powerful hub firms that can and do emerge in networks like Microsoft and Google. This leads to an additional dynamic focus for antitrust policy — on the evolvability of business systems and the way existing network structures constrain or enable subsequent evolution in productive ways. Lastly, agent based models represent a new and potentially valuable way to help not only researchers, but managers and policy makers to develop and explore the implications of alternative strategies and policies. Realistic “flight simulators” can be developed, validated and calibrated against the known characteristics of existing systems, which can help to sensitise them to the often counterintuitive behaviour of complex systems and to develop and evaluate alternative strategies and scenarios and examine how sensitive they are to various assumptions. 5. Summary and conclusions In some quarters of marketing there is considerable resistance to complexity and associated simulation methods. The problem is two-fold, there is a lack of research capability and research training in these areas and thus there is limited capability to use them or even evaluate the benefits they can bring. Perhaps more worryingly, we fear there is ever-increasing myopia within marketing, that our discipline is focusing on ever-smaller pieces of the big picture of marketing, that is, the markets within which business occurs, using the ever-more-rigorous reductionist methods that have characterised much of our discipline in recent decades. This process closes us off to all but a few new areas and opportunities. It has been argued that this is in part due to traditional methodologies driving the conceptual understanding of business and other social systems, which in turn is limiting the type of problems that are considered (Andriani & McKelvey, 2009; Laughlin, 2005). Instead of theories creating a requirement for appropriate research methods, the current limitations and assumptions of the traditional research approach are restricting the basis for theories about business and other social systems. In such a climate complexity science is rejected by these entrenched interests; they argue that simulation has limited value and/or that this approach is relatively new and/or that the theories are not strong enough. Leading

complexity scientists agree that not all the principles of organisation, patterns and hidden forces that make our world what it is have been solved but this does not mean that work in this area should not proceed and not be published in marketing journals purporting to be at the cutting edge of our discipline. To summarize, we have argued that marketing theory and research needs to break out of the straightjacket of its current dominant forms of methodology and theorising. It needs to move from a linear, comparative static, variables based approach to a non-linear, dynamic, evolutionary, process and mechanism based approach, as reflected in complex systems theory and methods. Such a transformation of the research agenda would not replace existing methods and theories. On the contrary, complex systems research complements, extends and contextualizes existing research. It must of necessity be informed by it, in part be validated by it and make use of existing methods in developing, analysing and testing complex systems models of business markets. Complex systems research will help stimulate, accelerate and necessitate further research of the more familiar type as well as enable new and neglected areas of research. A future research agenda needs to include: • Conducting systematic case histories and qualitative research to identify and better understand the nature and role of different mechanisms and process operating in business markets; • Developing and testing variables based statistical models of complex systems simulations of business markets; • Developing systematic experimental designs to analyse the behaviour and performance of complex computer simulation models under different conditions; • Designing and conducting experiments with real people and organisations that can inform and test the predictions and behaviour of complex system computer simulation models. There is a fundamental shift involved; it involves moving from the assumptions of reductionism to a constructionist view. As Nobel laureate Philip Anderson (1972) notes, “the reductionist hypothesis does not by any means imply a ‘constructionist’ one: The ability to reduce everything to simple fundamental laws does not imply the ability to start from those laws and reconstruct the universe” (p. 393). We urge marketers to at least consider both alternatives in doing science, thus opening the possibilities to do more research and know more about marketing systems. As we move further into the “century of complexity” (as articulated by Stephen Hawking), we urge the continuation and development of our own and others' voyages of scientific discovery via consideration of the complex properties of business networks' evolution and an extension in the ways we research them. References Ackoff, R., & Emery, F. (1972). On purposeful systems. London, UK: Tavistock Publications. Alderson, W. (1957). Marketing behaviour and executive action. Homewood, Ill: Irwin. Alderson, W. (1965). Dynamic marketing behaviour. Irwin. Alderson, W., & Cox, R. (1948). Towards a theory in marketing. Journal of Marketing, 13(2), 137–152. Allen, P., Maguire, S., & McKelvey, B. (2011). The SAGE handbook of complexity and management. London: Sage. Allen, P. M., & Sanglier, M. (1979). A dynamic model of growth in a central place system. Geographical Analysis, 11(3), 256–272. Anderson, P. W. (1972). More is different. Science, 177(4047), 393–396. Anderson, J. C., Håkansson, H., & Johanson, J. (1994). Dyadic business relationships within a business network context. Journal of Marketing, 58(4), 1–15. Andriani, P., & McKelvey, W. (2009). From Gaussian to Paretian thinking: Causes and implications of power laws in organizations. Organization Science, 20(6), 1053–1071. Ashby, W. R. (1958). Requisite variety and its implications for the control of complex systems. Cybernetica, 1(2), 83–99. Axelrod, R., & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211(4489), 1390–1396. Axtell, R. L. (2005). The complexity of exchange. The Economic Journal, 115(504), F193–F210. Axtell, R. L., Epstein, J. M., Dean, J. S., Gumerman, G. J., Swedlund, A. C., Harburger, J., et al. (2002). Population growth and collapse in a multiagent model of the Kayenta Anasazi in Long House Valley. Proceedings of the National Academy of Sciences of the United States of America, 99(Suppl. 3), 7275–7279.

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404 Bairstow, N., & Young, L. (2012). How channels evolve, a historical explanation. Industrial marketing management, 41, 385–393. Barabási, A. -L., & Albert, R. (1999). Emergence of scaling in random networks. Science, 286(5439), 509–512. Bianconi, G., & Barabási, A. L. (2001). Competition and multiscaling in evolving networks. Europhysics Letters, 54(4), 436–442. Bishop, S., Helbing, D., Lukowicz, P., & Conte, R. (2011). FuturICT: Managing and Exploring the Future. Procedia Computer Science, 7, 34–38. Blankenburg-Holm, D., Eriksson, K., & Johanson, J. (1996). Business networks and cooperation in international business relationships. Journal of International Business Studies, 27(5), 1033–1053. Borrill, P. L., & Tesfatsion, L. (2010). Agent-based modeling: The right mathematics for the social sciences? Ames, Iowa. : Iowa State University-Department of Economics. Boulding, K. E. (1970). A primer on social dynamics: A history of dialectics and development. New York: The Free Press. Boulding, K. E. (1978). Ecodynamics: A new theory of societal evolution. Sage. Boulding, K. E. (1985). The world as a total system. : Sage. Brenner, T. (2006). Agent learning representation: Advice on modelling economic learning. In L. Tesfatsion, & K. L. Judd (Eds.), Handbook of computational economics, 2. (pp. 895–947)Amsterdam: Elsevier. Browning, L. D., Beyer, J. M., & Shetler, J. C. (1995). Building cooperation in a competitive industry: SEMATECH and the semi conductor industry. Academy of Management Journal, 38(1), 113–151. Buttriss, G. J. (2009). An analysis of the process of evolution and impact of internet technologies on firm behaviour and performance using narrative sequence methods. Ph.D. Thesis, UNSW. Buttriss, G., & Wilkinson, I. F. (2006). Using narrative sequence methods to advance international entrepreneurship theory. Journal of International Entrepreneurship, 4, 157–174. Campbell, J. L. E. (Ed.). (2005). Where do we stand? Common mechanisms in organizations and social movements research. New York: Cambridge University Press. Chandra, Y., Styles, C., & Wilkinson, I. F. (2012). An opportunity based view (OBV) of rapid internationalisation. Journal of International Marketing, 20(1), 74–102. Clark, A. (1997). Being there: Putting brain, body, and world together again. Cambridge, MA: MIT Press. Clarke, M., & Wilson, A. G. (1983). The dynamics of urban spacial structure: Progress and problems. Journal of Regional Science, 23(1), 1–18. Davis, E., & Spekman, R. (2004). The extended enterprise. : Prentice Hall. Debenham, J. and Sierra, C. (2007). Building Business Relationships with Negotiation. E-Commerce and Web Technologies. ed.: 119–128. Denize, S., & Young, L. (2007). Concerning trust and information. Industrial Marketing Management, 36(7), 968–982. Dixon, D. F., & Fisk, G. (1967). Theories for marketing systems analysis: Selected readings. : Harper & Row. Dixon, D. F., & Wilkinson, I. F. (1982). The marketing system. Melbourne: Longman Cheshire. Dixon, D. F., & Wilkinson, I. F. (1986). Toward a theory of channel structure. Research in Marketing, 8, 27–70. Dixon, D. F., & Wilkinson, I. F. (1989). An alternative paradigm for marketing theory. European Journal of Marketing, 23(8), 59–69. Dorogovtsev, S. N., & Mendes, J. F. F. (2000). Evolution of networks with aging of sites. Physical Review E, 62(2), 1842–1845. Dyer, J., & Singh, H. (1998). The relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review, 23, 660–679. Easton, G., Brooks, R. J., Georgieva, K., & Wilkinson, I. F. (2008). Understanding the dynamics of industrial networks using Kauffman Boolean networks. Advances in Complex Systems, 11(1), 139–164. Easton, G., Wiley, J., & Wilkinson, I. F. (1999). Simulating industrial relationships with evolutionary models. 28th European Marketing Academy annual conference. Berlin: Humboldt University. Easton, G., Wilkinson, I. F., & Georgieva, C. (1997). Towards evolutionary models of industrial networks — A research programme. In H. G. Gemünden, T. Ritter, & A. Walter (Eds.), Relationships and networks in international markets (pp. 273–293) Oxford: Elsevier. Eguíluz, V. M., Zimmermann, M. G., Cela-Conde, C. J., & Miguel, M. S. (2005). Cooperation and the emergence of role differentiation in the dynamics of social networks. The American Journal of Sociology, 110(4), 977–1008. Ehrhardt, G. C. M. A., Marsili, M., & Vega-Redondo, F. (2006). Phenomenological models of socioeconomic network dynamics. Physical Review E, 74(3), 036106–036111. Emery, F. (1969). Systems thinking: Selected readings. London: Penguin Books. Emery, F., & Trist, E. (1965). The causal texture of organizational environments. Human Relations, 18, 21–32. Epstein, J. M. (2006). Generative social science: Studies in agent-based computational modeling. Princeton: Princeton University Press. Erdös, P., & Renyi, A. (1959). On random graphs. Publicationes Mathematicae Debrecen, 6, 290–297. Fan, Z., & Chen, G. (2004). Evolving networks driven by node dynamics. International Journal of Modern Physics B, 18(17–19), 2540–2546. Fisk, G. (1967). Marketing systems. New York: Harper & Row. Følgesvold, A., & Prenkert, F. (2009). Magic pelagic — An agent-based simulation of 20 years of emergent value accumulation in the North Atlantic herring exchange system. Industrial Marketing Management, 38(5), 529–540. Forrester, J. W. (1958). Industrial dynamics — A major breakthrough for decision makers. Harvard Business Review, 36(4), 37–66. Forrester, J. W. (1961). Industrial dynamics. Cambridge, MA: M.I.T. Press.

403

Forrester, J. W. (1971). World dynamics. Cambridge, MA: MIT Press. Forrester, J. W. (1980). Principles of systems. Cambridge, MA: MIT Press. Fronczak, P., Fronczak, A., & Holyst, J. A. (2006). Self-organized criticality and coevolution of network structure and dynamics. Physical Review E, 73(4) (046117-046114). Gavrilets, S., Duenez-Guzman, E. A., & Vose, M. D. (2008). Dynamics of alliance formation and the egalitarian revolution. PLoS One, 3(10), e3293. Gilbert, N., Pyka, A., & Ahrweiler, P. (2001). Innovation networks — A simulation approach. Journal of Artificial Societies and Social Simulation, 4(3). Goldenberg, J., Libai, B., & Muller, E. (2001). Talk of the network: A complex systems look at the underlying process of word-of-mouth. Marketing Letters, 12(3), 211–223. Gross, T., & Blasius, B. (2008). Adaptive coevolutionary networks: A review. Journal of the Royal Society, Interface, 5(20), 259–271. Håkansson, H., & Östberg, C. (1975). Industrial marketing: An organizational problem? Industrial Marketing Management, 4(2–3), 113–123. Håkansson, H., & Snehota, I. (1995). Developing relationships in business networks. London, UK: Routledge. Haldane, A. G., & May, R. M. (2011). Systemic risk in banking ecosystems. Nature, 469(7330), 351–355. Hedström, P. (2005). Dissecting the social: On the principles of analytical sociology. Cambridge, UK: Cambridge University Press. Held, F. (2010). Developing agent-based models of business relations and networks as complex adaptive systems. Sydney. http://www.agsm.edu.au/bobm/ARC/ARC_DP0881799_ Interim_Report.pdf Held, F., Wilkinson, I. F., Marks, R., & Young, L. (2010a). Exploring the dynamics of economic networks: First steps of a research project. 3rd World Congress on Social Simulation (Kassel, Germany). Held, F., Wilkinson, I. F., Marks, R., & Young, L. (2010b). Modelling the dynamics relations and networks in B2B markets Australia New Zealand Marketing Academy conference. New Zealand: Christchurch. Henrich, J. (2004). Cultural group selection, coevolutionary processes and large-scale cooperation. Journal of Economic Behavior and Organization, 53(1), 3–35. Hibbert, B., & Wilkinson, I. F. (1994). Chaos theory and the dynamics of marketing systems. Journal of the Academy of Marketing Science, 22(3), 218–233. Huang, Y. (2010). Understanding Dynamics of Trust in Business Relationships. Ph.D. Thesis, University of New South Wales. Huff, D. L. (1964). Defining and estimating a trading area. Journal of Marketing, 373–378. Iansiti, M., & Levien, R. (2004). The keystone advantage: What the new dynamics of business ecosystems mean for strategy, innovation and sustainability. Boston, MA: Harvard Business School Press. Jorg, T. (2011). New thinking in complexity for the social sciences and humanities. Heidelberg: Springer. Kanter, R. M. (1994). Collaborative advantage: The art of alliances July–August 1994, 96–108. Harvard Business Review(July–August), 96–108. Kauffman, S. A. (1992). Origins of order: Self organisation and selection in evolution. New York: Oxford University Press. Kauffman, S. (1994). Whispers from Carnot: The origins of order and principles of adaptation in complex non-equilibrium systems. In G. A. Cowan, D. Pines, & D. Meltzer (Eds.), Complexity: Metaphors, models and reality (pp. 83–136) Santa Fe, NM: Addison Wesley and Santa Fe Institute. Kim, W. -S. (2009). Effects of a trust mechanism on complex adaptive supply networks: An agent-based simulation study. Journal of Artificial Societies and Social Simulation, 12(3), 4. Kimbrough, E. O. (2011). Heuristic learning and the discovery of specialization and exchange. Journal of Economic Dynamics and Control, 35, 491–511. Kimbrough, E. O., Smith, V. L., & Wilson, B. J. (2008). Historical property rights, sociality, and the emergence of impersonal exchange in long-distance trade. American Economic Review, 98(3), 1009–1039. Kirman, A. P., & Vriend, N. J. (2001). Evolving market structure: An ACE model of price dispersion and loyalty. Journal of Economic Dynamics and Control, 25(3/4), 459–502. Konno, T. (2009). Network structure of Japanese firms — Scale-free, hierarchy, and degree correlation: Analysis from 800,000 Firms. : SSRN eLibrary. Ladley, D., Wilkinson, I. F., & Young, L. C. (2007). Group selection and the evolution of cooperation. 9th European Conference on Artificial Life (Lisbon, Portugal). Ladley, D., Wilkinson, I. F., & Young, L. (2011). The evolution of cooperation in business. The University of Sydney Business School. Lane, D., & Maxfield, R. (1996). Strategy under complexity: Fostering generative relations. Long Range Planning, 29(2), 215–231. Langton, C. G. (Ed.). (1996). Artificial life: An overview. Complex adaptive systems. Boston: MIT Press. Laughlin, R. M. (2005). A different universe: Reinventing physics from the bottom down. New York: Basic Books. Layton, R. A. (1981a). Trade flows in macromarketing systems, part 1: A macro model of trade flows. Journal of Macromarketing, 1(1), 35–48. Layton, R. A. (1981b). Trade flows in macromarketing systems, part 2: Transforming input output tables into trade flow tables. Journal of Macromarketing, 1(2), 48–55. Layton, R. A. (1985). Trade flows in Australia, 1974–75: An assessment of structural change. Journal of Macromarketing, 5(1). Layton, R. A. (1986). Modelling trade flows in a developing region. In G. Fisk (Ed.), Festschrift essays in marketing theory in honor of Reavis Cox. New York: Praeger. LeBaron, B., & Tesfatsion, L. (2008). Modeling macroeconomies as open-ended dynamic systems of interacting agents. American Economic Review, 98(2), 246–250. Leombruni, R., & Richiardi, M. (2005). Why are economists sceptical about agent-based simulations? Physica A, 355(1), 103–109. Li, S. X., & Rowley, T. J. (2002). Inertia and evaluation mechanisms in interorganizational partner selection: Syndicate formation among U.S. investment banks. Academy of Management Journal, 45(6), 1104–1119.

404

I.F. Wilkinson, L.C. Young / Industrial Marketing Management 42 (2013) 394–404

Liao, T. F. (2008). Series Editor's Introduction. In N. Gilbert, Agent Based Models. (pp. ix–x). Los Angeles: Sage. Linder, J. (2004). Transformational outsourcing. Sloan management review(Winter), 52–58. Luo, J., Whitney, D. E., Baldwin, C. Y., & Magee, C. L. (2011). How firm strategies influence the architecture of transaction networks. Boston, MA: Harvard Business School. March, J. G. (1996). Continuity and change in theories of organizational action. Administrative Science Quarterly, 41(2), 278–287. Marks, R. E. (1989). Breeding optimal strategies: Optimal behavior for oligopolists. Proceedings of the Third International Conference on Genetic Algorithms. George Mason University, Morgan Kaufmann Publishers. Mosekilde, E., Larsen, E. R., & Sterman, J. D. (1991). Coping with complexity: Deterministic chaos in human decisionmaking behavior. In J. L. Casti, & A. Karlqvist (Eds.), Beyond belief: Randomness, prediction, and explanation in science (pp. 199–229) Boston, MA: CRC Press. Pennisi, E. (2009). On the origin of cooperation. Science, 325(5945), 1196–1199. Powell, W. W., White, D. R., Koput, K. W., & Owen-Smith, J. (2005). Network dynamics and field evolution: The growth of interorganizational collaboration in the life sciences. The American Journal of Sociology, 110(4), 1132–1205. Prigogine, I., & Stengers, I. (1984). Order out of chaos. New York: Bantam Books. Rand, W., & Rust, R. T. (2011). Agent-based modeling in marketing: Guidelines for rigor. International Journal of Research in Marketing, 23(3), 167–280. Ren, J., Wu, X., Wang, W. X., Chen, G., & Wang, B. H. (2006). Interplay between evolutionary game and network structure: The coevolution of social net, cooperation and wealth. (Arxiv preprint physics/0605250v2). Ritter, T., Wilkinson, I. F., & Johnston, W. (2004). Firms' ability to manage in business networks: A review of concepts. Industrial Marketing Management, 33(3), 175–183. Rosenkopf, L., & Schilling, M. A. (2007). Comparing alliance network structure across industries: Observations and explanations. Strategic Entrepreneurship Journal, 1(3–4), 191–209. Sapienza, M. D. (2000). An experimental approach to the study of banking intermediation: The Banknet simulator. In F. Luna, & B. Stefansson (Eds.), Economic Simulation in Swarm (pp. 159–179) Boston: Kluwer. Sawyer, R. K. (2005). Social emergence: Societies as complex systems. Cambridge, UK: Cambridge University Press. Schelling, T. (1971). Dynamic models of segregation. Journal of Mathematical Sociology, 1, 143–186. Seibel, F., & Kellam, L. (2003). The virtual world of agent-based modeling: Proctor & Gamble's dynamic supply chain. Perspectives on Business Innovation, 9, 22–27. Simon, H. A. (1968). On judging the plausibility of theories. In B. van Rootselaar, & F. Staal (Eds.), Logic, methodology and philosophy of sciences III (pp. 439–459). Amsterdam: North-Holland. Simoni, M., Tatarynowicz, A., & Vagnani, G. (2006). The complex dynamics of innovation diffusion and social structure: A simulation study. WCSS 2006 — The First World Congress on Social Simulation, Kyoto, Japan. Somani, A., & Tesfatsion, L. (2008). An Agent-Based Test Bed Study of Wholesale Power Market Performance Measures. Computational Intelligence Magazine, IEEE, 3(4), 56–72. Stanley, E., Ashlock, D., & Smucker, M. (1995). Iterated prisoner's dilemma with choice and refusal of partners: Evolutionary results. Advances in artificial life (pp. 490–502) Berlin: Springer. Sterman, J. D. (1989). Modeling managerial behavior: Misperceptions of feedback in a dynamic decision making experiment. Management Science, 35(3), 321–339. Sun, J., & Tesfatsion, L. (2007). Dynamic testing of wholesale power market designs: An open-source agent-based framework. Computational Economics, 30(3), 291–327. Tesfatsion, L. (1997). A trade network game with endogenous partner selection. Computational Approaches to Economic Problems, 249–269. Thompson, M., & Young, L. (2012). The complexities of measuring complexity, special session on agent based modelling in business markets, IMP 2012, Industrial Marketing and Purchasing Conference. (Rome). Tomassini, M., Pestelacci, E., & Luthi, L. (2010). Mutual trust and cooperation in the evolutionary hawks–doves game. Bio Systems, 99(1), 50–59. Van de Ven, A. H., & Engleman, R. M. (2004). Event and outcome driven explanations of entrepreneurship. Journal of Business Venturing, 19, 343–358. Vázquez, A. (2000). Knowing a network by walking on it: Emergence of scaling. (arXiv:cond-mat/0006132v4). von Hippel, E. (1986). Lead users: A source of novel product concepts. Management Science, 32(7), 791–805. von Hippel, E. (2005). Democratizing innovation. Cambridge, MA: MIT Press. Watts, D. J. (2003). Six degrees: The science of a connected age. London, UK: William Heinemann. Watts, D. J., & Dodds, P. S. (2007). Influentials, networks, and public opinion formation. Journal of Consumer Research, 34(4), 441–458. Welch, D., Welch, L., Wilkinson, I., & Young, L. (1996a). Network development in international project marketing and the impact of external facilitation. International Business Review, 5(6), 579–602. Welch, D., Welch, L., Wilkinson, I. F., & Young, L. (1996b). Export grouping relationships and networks: Evidence from an Australian scheme. International Journal of Research in Marketing, 13.

Welch, D. E., Welch, L. S., Wilkinson, I. F., & Young, L. C. (1998). The importance of networks in export promotion: Policy issues. Journal of International Marketing, 6(4), 66–82. Welch, C., & Wilkinson, I. (2002). Idea logics and network theory in business marketing. Journal of Business-to-Business Marketing, 9(3), 27–48. Welch, C., & Wilkinson, I. F. (2004). The political embededness of international business networks. International Marketing Review, 21(2), 216–231. Wilhite, A. (2001). Bilateral trade and small-world networks. Computational Economics, 18, 49–64. Wilkinson, I. F. (1989). An illustrative analysis of the stability of simple channel systems' behaviour. Australian Management Educators Conference (Adelaide). Wilkinson, I. F. (1990). Toward a theory of structural change and evolution in marketing channels. Journal of Macromarketing, 10, 18–46. Wilkinson, I. F. (2001). A history of network and channels thinking in marketing in the 20th century. Australasian Marketing Journal, 9(2), 23–52. Wilkinson, I. F. (2006). The evolvability of business and the role of antitrust. Antitrust Bulletin, 51(1), 111–141. Wilkinson, I. F. (2008). Business relating business — Managing organisational relations and networks. Cheltenham, UK: Edward Elgar. Wilkinson, I. F., Held, F., Marks, R. E., & Young, L. (forthcoming). Developing agent-based models of business relations and networks. In R. Stocker, & T. Bossomaier (Eds.), Networks in society: Links and language. Singapore: Pan-Stanford Publishing. Wilkinson, I. F., Mattsson, L. -G., & Easton, G. (2000). International competitiveness and trade promotion policy from a network perspective. Journal of World Business, 35(3), 275–299. Wilkinson, I. F., Wiley, J. B., & Lin, A. (2010). Modelling the Structural Dynamics of Industrial Networks. In A. Minai, D. Braha, & Y. Bar-Yam (Eds.), Unifying Themes in Complex Systems, Vol. IV (pp. 347–365), Part II.New York: Springer. Wilkinson, I. F., & Young, L. (2002). On cooperating firms, relations and networks. Journal of Business Research, 55, 123–132. Wilkinson, I. F., & Young, L. C. (2005). Toward a normative theory of normative marketing theory. Marketing Theory, 5(4), 363–396. Wilkinson, I. F., Young, L., & Ladley, D. (2007). Group selection vs individual selection and the evolution of cooperation in business networks. Manchester UK: IMP Conference. Young, L. (2006). Trust: Looking forward and back. The Journal of Business and Industrial Marketing, 21(7), 439–445. Young, L., & Bairstow, N. (2011). Narrative event methods: Understanding how business market processes equilibrate and change over time. Proceedings, Industrial Marketing and Purchasing Conference, University of Strathclyde. Glasgow, Scotland: IMP Grooup. Young, L., & Bairstow, N. (2012). Explaining Channel Evolution with Historical Methods. Industrial Marketing Management, 41(1), 385–393. Young, L., & Denize, S. (2008). Competing interests: The challange to collaboration in the public sector. International Journal of Sociology and Social Policy, 28(1/2), 46–58. Young, L., Wiley, J., & Wilkinson, I. F. (2009). A comparison of European and Chinese supplier and customer functions and the impact of connected relations. The Journal of Business and Industrial Marketing, 24(1), 35–45. Zimmermann, M. G., Eguíluz, V. M., & San Miguel, M. (2004). Coevolution of dynamical states and interactions in dynamic networks. Physical Review E, 69(6), 065102.

Ian F. Wilkinson (B.Sc., M.Sc. and Ph.D. from University of New South Wales, Australia) is an honorary professor of Marketing at the University of Sydney and a visiting professor in the Department of Entrepreneurship and Relationship Management at the University of Southern Denmark. He has held academic posts at various American, European as well as Australian universities. His research focuses on the development, evolution and management of interfirm relations and networks in domestic and international business, especially from a complexity perspective. His research has been published in leading international journals including the European Journal of Marketing, Industrial Marketing Management, Journal of the Academy of Marketing Science, Journal of Applied Psychology, Journal of Business Research, Journal of International Marketing, Journal of Retailing, and Marketing Theory. His books include Business Relating Business: Managing Organisational Relations and Networks. Louise Young is a professor of Marketing at the University of Western Sydney and visiting professor of Relationship Management and Entrepreneurship at the University of Southern Denmark. She was educated in the USA and Australia and has held academic posts at various Australian and European universities. She has published widely in a range of discipline areas including marketing, complexity science, international business, social policy, management, human resources and information technology. Her current research focuses on the evolution and management of business relationships and networks and in particular on the psychology of the individuals co-creating them, using methods including case study, ethnography, deep qualitative interview, lexicographic semantic study and agent-based modelling. Her research has been funded by grants from the Australian Research Council, the Australian Trade Commission, the CSIRO and various university research grants. She also has worked extensively in applied research areas participating the generating of market research for Australian firms, government instrumentalities and not-for-profit organisations.