« Prior Essay (Commenting still active) | Main | Comment Here on "A Call for Changes in Research Practice" »

July 24, 2008


Dr. Liz Breen

I enjoyed reading this but found it a bit frustrating. I was waiting for the obvious statement to be made early on in the proceedings that a network approach should be adopted, and it did emerge right at the end. So worth waiting for. I do appreciate that we need to start with the foundations of buyer-supplier realtionships but supply chains and globalisation being as it is, most relationships involve a greater number of suppliers linked by their customers and suppliers, directly or indirectly. In order to fully examine elements such as risk for example we need to examine risks across and within a supply chain, as mitigating risk with one supplier can have a direct impact on other customers. The same can be said of supplier development, it has a positive impact on othe r customers. Having spent a substantial amount of research time in the pharmaceutical supply chain in the NHS (UK), there is more need for examination of these buyers and suppliers by focusing on them as interdependent entities as opposed to individual entities. They may choose not to work in a co-operative manner, being competitive and some providing similar products and services e.g. wholesalers, but they are inextricably linked by this very fact. This paper sparked a train of thought, as good papers do, so I look forward to futher discussions focusing on buyer-supplier network/web relationships.

Jamison Day

In this Forum article, Choi and Wu have touched on a fundamental issue that could mark an important turning point in supply chain research. Examining the interactions between relationships, rather than between firms, inherently takes us from a dyadic perspective into, at the very least, a triadic one.

Indeed, relationships are interdependent and influence the behavior and performance of a supply chain network. For example, the tolerance a supply network has for errors or attacks is influenced by the structural topology of relationships between the many participating firms (Albert et al., 2000; Thadakamalla et al., 2004).

Introducing additional relationships may present new research challenges, but it can also open up new opportunities for advancing the state-of-the-art and delving into greater managerial relevance. Consider that there is no known way, short of vertical integration, to fully mitigate the order variability magnification induced between a buyer-supplier dyad (Simchi-Levi et al., 2008). However, if we create a triad by introducing a second buyer, it is not only possible to fully mitigate this bullwhip effect, the supplier's incoming composite order variability can actually be less than the demand variability experienced by either of the buyers (Murray and Day, 2008). So in a triad, a firm that manages multiple buyer relationships can actually do even better than eliminating the bullwhip effect - and no vertical integration is necessary. This level of performance has not been found possible in dyads.

Techniques of theoretical isolation have been and will continue to be important for the advancement of our field. However, very few firms today maintain only one buyer or supplier relationship. Yes, triads are an essential foundational structure of supply chain networks. Choi and Wu have made a strong case that, in addition to inter-firm issues, it imperative for our field to explore inter-relationship phenomena as well.

Albert, R., H. Jeong, and A.L. Barabasi 2000. Error and attack tolerance of complex networks, Nature 406, 378-382.
Murray, M. and J.M. Day 2008. Taming the bullwhip effect in complex supply networks with demand portfolio management, Working Paper, University of Houston.
Simchi-Levi, D., P. Kaminski, and E. Simchi-Levi 2008. Designing and managing the supply chain, 3rd edition, McGraw-Hill/Irwin.
Thadakamaila H.P., U.N. Raghavan, S. Kumara, and R. Albert 2004. Survivability of multiagent-based supply networks: a topological perspective, IEEE Intelligent Systems 5(19), 24-31.

Assoc. Professor Mile Terziovski

This is an interesting essay which argues that dyads are not sufficient to capture “..the interactive nature inherent in a network.” I congratulate the authors for “thinking outside of the box.” I have decided to use an article by Whetten (1989) in order to evaluate the contribution of the essay to the operations management field. Whetten has articulated seven factors that should be considered in judging conceptual papers:

1. What is new? Does the essay make a significant, value-added contribution to current thinking?

The essay provides a strong argument for supply chain networks to be investigated from a triad perspective (three nodes and three links) rather than a dyad (two nodes and one link). The argument is well illustrated and supported from the literature. The authors have pushed the current paradigm of supply chain theory.

2. So What? Will the triad theory likely alter research practice in supply chain management?

The authors have satisfied the so what test by providing a strong explanation of triads answering the what and how questions. What are dyads and triads are and how they are interrelated. I think it is just as important if not more important to address in a more rigorous manner the why question. Why triads? What are the assumptions and the limitations of using triads?

3. Why so? Are the underlying logic and supporting evidence compelling?

I think the authors need to develop propositions to convince us that these make sense and would impact on supply chain management research. For example, a likely proposition could be: “triadic relationships are expected to vary over time as the role of the supplier changes to that of customer along the integrated supply chain.”

4. Well done? Does the essay reflect seasoned thinking?

The essay reflects seasoned thinking. The authors are very well established researchers in the supply chain management field.

5. Done well? Is the essay well written? Does it flow logically?

The essay is very well written and flows logically.

6. and 7. Why now? Who cares? Is this topic of current interest to scholars in this area? How interesting would the triad theory be to operations management researchers?

This topic would be of high interest to scholars in the area of supply chain management and the operations management field generally.


Whetten, , D.A., What Constitutes a Theoretical Contribution? Academy of Management Review, 1989, Vol.14, No.4. 490-495.

Martin Spring, Lancaster University Management School, UK

Choi and Wu make a welcome call for the extension of OM analysis of supply relations beyond the dyad, to include indirect relationships and their emergent properties. This continues a stream of work by Choi and his co-authors, notably Wu and Choi (2005). Of course, OM has been deeply concerned with one form of analysis beyond the dyad for many years – supply chain management (SCM). But this has only treated indirect inter-firm relationships in one structural arrangement – those between (say) a customer and successive tiers of suppliers. The work of Jay Forrester, which was the forerunner of SCM, famously identified the Forrester Effect, which is truly an emergent phenomenon of the larger system i.e. it is meaningless at the level of the dyad.

Choi and Wu trace an interesting lineage of the study of triadic structural relationships in fields outside of OM. These provide a strong foundation. But there was no need to resort so exclusively to work so distant from a concern with business-to-business relationships, for there is a thirty-year tradition of work in this vein in the Industrial Networks literature. This mainly European body of work (e.g. Håkansson, 1982, Ford, 1990, Axelsson and Easton, 1992, Håkansson and Snehota, 1995), has focussed on inter-firm relationships, ostensibly with a primary concern for marketing and technological development but, in many instances, dealing with issues close to OM. Very much more specifically, Smith and Laage-Hellman (1992) study triadic relationship structures, and identify a typology of seven transformation patterns in triads of firms.

The Industrial Networks approach has informed European OM to some extent: indeed, EurOMA, Europe’s premier OM conference, will in 2009 be hosted by Chalmers University, Gothenburg, Sweden, which is closely associated with the Industrial Networks tradition. But the approach also frustrates some in OM for the reluctance of many of its adherents to move from understanding and explanation to a more prescriptive position. Smith and Laage-Hellman is a case in point: the typology is identified based on detailed case studies, but no prescription follows. This, then, is the OM challenge: to build on such work that identifies a rich palette of network phenomena in at least two ways. First, to take the various concepts and develop prescriptive frameworks that, while retaining sensitivity to context, go some way to providing OM researchers and practitioners with useful theory. Second, to develop this line of research ‘with an OM lens’, if you will, keeping in mind core OM concerns. Choi and Wu and many other supply chain researchers have been quick to turn to issues such as trust and embeddedness as their central concerns; while these are undoubtedly important, it seems that issues more central to OM such as capacity, planning and control systems, purchasing contracts, operations capabilities and risk are being neglected. For example, how do capacity, risk and performance interact in networks of outsourced B2B service providers?

Finally, the network perspective creates a new set of challenges for method. I return to a study from another field, and indeed a study concerning dyads, to explore this. Thompson and Walker (1982) examine methodological issues in researching family relationships. They stress the need to identify those characteristics and phenomena that are archetypically emergent in nature: conflict, for example, is necessarily an inter-personal phenomenon. They then discuss the implications of alternative approaches to empirical study of such relationships, noting that such research generally involves asking individuals about aspects of their relationships with third parties. In a relationship between A and B, we can ask A about herself, B about himself, A about B, B about A, and either A or B about the relationship between them. The implications for method as we move to triads and then on to larger nets should be evident. Returning to industrial networks, a typical research approach, if adopting case methods, would be to gain access via one firm, then use that access (a) to gain data about that firm (b) to gain data about the other firms and (perhaps) (c) to then gain access to other firms, all of which leads, we hope, to being able to generate insight about individual firms, about various dyads as well as about combinations of firms in various small networks. This is a long way from classic Operations Strategy research, say, where we gained access to a plant in order to collect data about the plant to then be able to generate insights and prescription about plants: in other words, the unit of analysis is clear-cut and always the same. In a previous OSM Forum Essay, Carter et al. (2008) call for a multi-method approach to SCM research. The study of networks, with its inherent complexity and range of as yet poorly theorized constructs, makes this even more vital.


AXELSSON, B. & EASTON, G. (1992) Industrial Networks: A New View of Reality, London, Routledge.
CARTER, C. R., SANDERS, N. R. & DONG, Y. (2008) Paradigms, Revolutions, and Tipping Points: The Need for Using Multiple Methodologies within the Field of Supply Chain Management. Journal of Operations Management - OSM Forum.
FORD, D. (1990) Understanding Business Markets: Interaction, Relationships, Networks, London, Academic Press.
HÅKANSSON, H. (1982) International Marketing and Purchasing of Industrial Goods, Chichester, John Wiley.
HÅKANSSON, H. & SNEHOTA, I. (1995) Developing Relationships in Business Networks, London, Routledge.
SMITH, P. C. & LAAGE-HELLMAN, J. (1992) Small group analysis in industrial networks. IN AXELSSON, B. & EASTON, G. (Eds.) Industrial Networks: A New View of Reality. London, Routledge.
THOMPSON, L. & WALKER, A. J. (1982) The Dyad as the Unit of Analysis - Conceptual and Methodological Issues. Journal of Marriage and the Family, 44, 889-900.
WU, Z. H. & CHOI, T. Y. (2005) Supplier-supplier relationships in the buyer-supplier triad: Building theories from eight case studies. Journal of Operations Management, 24, 27-52.

Tim Baker

This essay lays an excellent foundation for theory building and testing in supply chain management.

Figure 1 can be a foundation for describing a number of different buyer-supplier relationships, where, specifically, A is the buyer and B and C are the suppliers. The relationships can be summarized by either an adversarial or cooperative relationship on each link (Choi and Wu, 2007). For instance, having all links cooperative would essentially be the parallel sourcing practices used by Japanese automakers (Richardson, 1993; Richardson and Roumasset, 1995). Another triad relationship would the the classic adversarial relationship between the buyer and each supplier, with a cooperative relationship between the two suppliers (Choi and Wu, 2007). Here, the buyer might be trying to force price reductions on the suppliers, and the suppliers band together in sort of a “misery loves company” scenario outlined by Simmel (1950). Interesting research questions include
(1) Can the latter triad relationship ever be a steady-state, and, if so, under what conditions does this occur?
(2) Under what conditions, if any, is a shift from the latter triad relationship to the all-cooperative one likely?

Another possible triad relationship involves a buyer having a cooperative relationship with one supplier but an adversarial relationship toward the other, with an adversarial relationship between the two suppliers (Choi and Wu, 2007). Such a triad can come about by, for instance, a new supplier coming into the market and being favorably viewed by the buyer. The buyer then views the older supplier as now less useful, and exploits them by forcing them to nurture the new supplier in existing business practices (Wu and Choi, 2005). Like before, this reality generates similar research questions—can this triad be a steady-state (why?) and what conditions promulgate a movement to other triads?

Another triad is a buyer having a cooperative relationship with both suppliers, but the suppliers being adversarial toward each other due to the competition (Choi and Wu, 2007). An interesting question would be what triad transitions out of this triad are likely under what conditions. For instance, would a move toward a dyad between the buyer and one supplier be likely, with the other supplier foregoing future interaction between the two?

The structural holes possible through the elimination of a link in Figure 1 also present interesting research possibilities. For instance, suppose that the buyer has a cooperative relationship with one supplier but an adversarial relationship with the other. And no relationship between the suppliers (the hole) (Choi and Wu, 2007). What would be the nature of the link between the two suppliers if they were forced by the buyer to work together?


Choi, T. and Wu, Z. 2007. Triads in Supply Networks: Interpretation through Balance Theory and Structural-Hole Concept. Working Paper.

Richardson, J. 1993. Parallel Sourcing and Supplier Performance in the Japanese Automobile Industry. Strategic Management Journal. 14(5): 339-350.

Richardson, J. and Roumasset, J. 1995. Sole Sourcing, Competitive Sourcing, Parallel Sourcing: Mechanism for Supplier Performance. Managerial and Decision Economics. 16(1): 71-84.

Dr. Raul R. Rodriguez and Prof. Dr. Rainer Leisten

This is an interesting paper that can be used for both research and teaching issues. In our opinion, its main contribution for our field is that it brings the discussion from the dyadic view (which is, on one hand, usually restricted to simple and sometimes very restricted perspectives of material and information flows, and on the other hand, depending of the respective research focus, either to purely operational or to too simple ‘strategic’ issues) to the triadic one (adding the perspective of sociological issues, more real-world related tactical or strategic issues, …).
Then, we propose to include the following issues into the debate of (not only) this forum:
1. When we first read the paper, an immediate and well-known figure came to our minds. It is widely used by the Supply Chain Council when introducing the well-known SCOR-model: Several and unstructured chains enter a company from the supplier side, they are somewhat more clear within the company and they become again rather unstructured on the customer side outside of the company. A true and precise analysis of a, maybe, extended triadic view on the company might yield significant additional insights on the relationships of a company to its relevant environment(s).
2. It is said that the triadic relationships are changing over time and that they are interrelated. Well, this is commonly accepted and true within the Psychology ambit, where the dynamic relationships of components of a network have been deeply analysed (Ghoshal and Moran, 1996; Serva et al., 2005). When applied to the field of supplier-buyer networks we should be able to identify and quantify such relationships in order to improve analysis as well as decisions. This could be done by applying either subjective techniques based on expertise (e.g. multi-criteria decision aid techniques) or objective techniques (such as multivariate analysis) to establish cause-effect relationships between the main factors that affect triadic relationships by analysing the data of performance of the network. Further, it seems to be necessary to take into account the step-thinking nature that takes place within strategic behaviour. Additionally, it could be possible to think of game theoretical approaches, may it be cooperative or non-cooperative.
3. Taking the last point, the paper states that relationships between the parts are traditionally either “collaborative or adversarial”. There are other possibilities between these two as collaboration goes from “transactional collaboration” (enterprises share data with no strategic value and/or strategic intention) to “cooperative collaboration” (where enterprises comprehensively share information on forecasts, inventories or order and delivery status) (Cohen and Roussel, 2003). Further, the scenario should be clarified where one relationship of the triad is collaborative and the other two are adversarial (or some intermediate configuration). In this case, who gets priority? Moreover, is it possible to establish a win/win/loose situation inside the triad? Imagine the scenario where one producer has got two customers, an internal one and an external one, in the event of a bottleneck, who’s getting the priority? (We know real-life examples of companies which either always prefer the internal customer or the external customer.) Who’s managing and deciding this? Is it possible to have two triads inside the same network and they being adversarial?
4. It is affirmed that “the triad is the smallest unit of a network”: Can this be taken as a general standard for all networks? Is not it possible to think of a network in which four five or six actors would constitute the smallest thinkable unit? From a mathematical point of view, the smallest part of a network is trivially one firm or one node. We believe that this affirmation should be taken carefully and that each particular problem should be studied particularly. At least from a business point of view, a company usually has at least several suppliers, several customers and several competitors, which results at least in a six-pole view. Related to this: What is the added value of the triadic approach as compared with or relative to the widely discussed market-based (e.g. Porter) and/or resource-based approaches (e.g. Prahalad/Hamel)? We think that there will be an enormous contribution but it should be clearly identified.
5. How does introducing the triadic view on flow maximisation considerations bottleneck or duality issues?
6. How can the triadic view be interpreted in terms of centralized or decentralized organisations?
7. If the network is interpreted from a systems theory point of view: Is the network an open/closed system?
Cohen, S. and Roussel, J. (2003). Strategic Supply Chain Management: The 5 disciplines for top performance. McGraw-Hill: New York.
Ghoshal, S. and Moran, P. (1996), Bas for practice: A critique of the transaction cost theory. Academy of Management Review, 21, pp. 13-47.
Porter, M.E. (1980). Competitive strategy: Techniques for Analyzing Industries and Competitors. Free Press.
Prahalad, C.K. and Hamel, G. (1990). The core competence of the corporation', Harvard Business Review, May/June, pp. 79-91.
Serva, M.A., Fuller, M.A., Mayer, R.C. (2005). The reciprocal nature of trust: A longitudinal study of interacting teams. Journal of organizational behavior, 26, pp. 625-648.

Tobias Schoenherr, Michigan State University

Choi and Wu provide a refreshing and exciting perspective to examine buyer-supplier relationships. Their call for the investigation of triads, rather than dyads, is important and well-warranted, especially given the environment companies are faced with in today’s world. In my comments below I will focus on two objectives. First, I will add to Choi and Wu’s arguments, and provide further rationales why the focus on triads, rather than dyads, is even so much more important today. And second, I will elaborate on a specific context mentioned by Choi and Wu – the phenomenon of supply chain disintermediation. In a concluding paragraph I will allude to some of the challenges posed by triadic research.

First, I fully concur with Choi and Wu’s call for research within the triad, and provide further substantiation for their argument. Procurement has been experiencing a renaissance since the mid-1980s, moving up in corporate hierarchies and being considered a strategic asset. This went contrary to the view in the past, where purchasing was merely regarded as a “negative function”, making “little positive contribution” (Henderson, 1975). Furthermore, until the late 1990s the focus has been primarily on internal integration within enterprises, and the breaking down of functional silos. Having realized significant benefits from these initiatives, companies also attempted to integrate with their suppliers and customers, to streamline the supply chain via closer collaboration and information sharing. Information technology, such as enterprise resource planning systems, has been a wonderful enabler in this regard. This drive toward closer integration across company boundaries was also fueled by an increased level of global competition, constantly rising customer expectations, the need for greater agility, and the cost and availability of information (Monczka, Handfield, Giunipero and Patterson, 2009). Simultaneous was the trend toward greater outsourcing combined with the focus on core competencies. At the same time, firms also moved towards reduced inventories, just-in-time environments, and vendor-managed-inventories. Inevitably connected with these developments was the heightened dependence of companies on their supply chain partners. In addition, global sourcing and geographically distant suppliers represented heightened risk in terms of transportation and lead time. To offset some of these risks, companies tended to move away from a pure single-source model, to a setting where there may be a primary supplier together with a number of back-up or secondary suppliers for the same commodity. Building-in these redundancies in the supply chain eliminated, or at least reduced, the risk of disruption. These developments have all contributed to the rising importance of triads, and the move away from one-dimensional dyads. As such, in my supply chain classes I talk about supply networks, supply nets or supply webs, rather than a single linear chain of entities. Having this network in place enables the proper flow of materials and information, even if one link fails or is disrupted. It therefore seems natural to focus future research on triads, and I can only stress the call made by Choi and Wu.

And second, Choi and Wu mention the phenomenon of supply chain disintermediation, which has been investigated within the triadic context (Rossetti and Choi, 2005, 2008). The developments described in the prior paragraph fueled this trend toward disintermediation, especially due to the tremendous opportunities offered by information technology and the Internet, as well as the associated ease of information accessibility and their oftentimes minimal cost. Such disintermediation in a triadic context was also explored by Mabert and Schoenherr (2001) and Schoenherr and Mabert (2003). In Mabert and Schoenherr (2001), the authors chronicle the experiences of a buying company with online reverse auctions. At first, the firm relied on the expertise of a third-party reverse auction provider to conduct the online bidding events, representing a triadic context: the focal buying company, the auction provider as intermediary, and the supplier of the product. Over time, the company became “e-market savvy”, and the third-party provider was not able to add value any more to the triadic relationship. The focal firm ultimately developed capabilities in-house to administer the online bidding events, resulting in disintermediation, i.e. the elimination of the middleman. The company was now able to conduct this type of negotiation with the suppliers directly, representing a dyadic relationship again. This experience illustrates the increased possibilities for disintermediation in a triadic context. At the same time, it also stresses the required constant quest of intermediaries to continuously add value to the relationship. If benefits offered are not sustainable, disintermediation may be inevitable (Schoenherr and Mabert, 2003).

While the move toward the study of triadic relationships is important, it provides also certain challenges. The primary challenge may be the accessibility of triads by researchers. This may be especially difficulty in the research environment today, where companies are constantly approached by researchers for input and information. Evidence are the declining survey response rates, often also triggered by company policy not to participate in any outside survey. Research becomes also challenging when responses from all three entities are required, multiplying the difficulty of obtaining complete sets. However, these challenges should not serve as an excuse to not do research in triads. Investigating buyer-supplier relationships in triads is important, as called for in Choi and Wu, in this comment, and in other comments provided for the article.


Henderson, B. D. 1975. The Coming Revolution in Purchasing. Journal of Purchasing and Materials Management, 44(Summer):44-46 (reprint of an article first appearing in 1964).

Mabert, V. A. and Schoenherr, T. 2001. Evolution of Online Auctions in B2B E-Procurement. Practix – Best Practices in Purchasing and Supply Management. 5 (1):15-19.

Monczka, R. M., Handfield, R. B., Giunipero, L. C. and Patterson, J. L. 2009. Purchasing and Supply Chain Management. Fourth edition. South-Western Cengage Learning, Mason, OH.

Rossetti, C. and Choi, T. Y. 2005. On the Dark Side of Strategic Sourcing: Experiences from the Aerospace Industry. Academy of Management Executive. 19 (1): 46-60.

Rossetti, C. and Choi, T. Y. 2008. Supply Management under High Goal Incongruence: An Empirical Examination of Disintermediation in the Aerospace Supply Chain. Decision Sciences. 39 (3):507-540.

Schoenherr, T. and Mabert, V. A. 2003. A Conceptual Study of Developments in B2B Reverse Online Auctions: The Changing Role of the Online Auction Provider / Intermediary. In: A. Chikan (Ed.), Advances in Purchasing and Supply Chain Management, IFPSM Publications, 4.

Dan Krause

Choi’s and Wu’s call for triadic research is certainly valuable and worthwhile. However, in the following paragraphs I attempt to explain why my enthusiasm for their call is not unbridled!

First, as Choi and Wu point out, the notion of triads is not new. Triads are also not new to supply chain research. For example, the purchasing and supply management literature has long-documented cases of triadic interaction in supplier development (e.g., Leenders, 1966). Leenders and Blenkhorn further developed these ideas in a book on Reverse Marketing (1988), in which they described a company’s effort to counter a sole supplier situation by developing a new, alternate source of supply. The end result was a second source of supply, and a more accommodating stance from the original sole source in the form of price reductions. Further, the marketing literature has also emphasized the need to consider the business network context of dyadic relationships. For example, Anderson, Hakansson and Johanson (1994) provide a conceptual framework and two case studies of dyadic supplier-customer relationships, and how they are embedded in a business network. They further differentiate between primary (“composed of a focal dyad’s immediate suppliers and customers” (p. 5)) and secondary functions of these dyadic relationships, noting that the secondary function “means they are also parts of networks” (p. 3) , that is, there are many other companies “indirectly connected to the focal dyad” (p. 5). A paper by Wathne and Heide in 2004 is just one example of subsequent work that has empirically investigated the effect of other supply chain entities on dyadic relationships.

Additionally, I have concerns that are related to recent calls for multiple responses for most supply chain survey research. Choi and Wu note that we know a lot about dyads through the extant literature. However, much of the extant empirical literature on dyads is comprised of data collected from a single respondent in a single firm – and I include some of my own work in that description! In that sense, much of what we know, or think we know about dyads, is open to question and should be subjected to further scrutiny via truly dyadic research, that is, by collecting data from both parties in the dyad. That most of the extant research hasn’t captured the link – that is, hasn’t gathered data from both parties in the dyad, suggests that we may yet have much to learn about constructs such as inter-firm collaboration, inter-organizational learning, and so on.

Supply chains are comprised of buyer-supplier dyads, which are based on market transactions and contractual relationships – these strike me as basic building blocks of a supply base (as opposed to a network -- see Choi and Krause (2006)). That is, these contracts and the transactions they govern, generally are agreements between one customer and one supplier. Further, these transactions are often between self-described partners that represent a significant portion of each other’s businesses. Clearly the unit of analysis, driven by a research question, should influence how the data are collected. That these transactions take place within the context of the marketplace amidst other customers, suppliers and competitors is also clear.

Thus, while I think the triadic perspective is valuable, and hope that it gains additional momentum, I also hope that researchers who are also focused on supply chain dyads work harder to gain a more valid picture of the inner workings of these dyads by gathering data from two sources. One reason that triadic research has not gained momentum over time may be the sheer difficulty of gaining access to reasonably large numbers of triads for data collection. For example, the work by Wathne and Heide (2004), cited above, collects data from only two sources – the parallel to collecting data from one source in dyadic research. If, as Choi and Wu propose, their call is to gather data on these quasi-entities called links, as well as the nodes in triads, researchers will have to gather data from a minimum of three respondents from three different firms. The challenge as I see it, is to find new, innovative ways to collect this data.


Anderson, J. C., H. Hakansson, and J. Johanson. 1994. “Dyadic business relationships within a business network context,” Journal of Marketing, 58, 1-15.

Choi, T and D. Krause. 2006. "The supply base and its complexity: Implications for transactions costs, risks, responsiveness, and innovation," Journal of Operations Management, 24, 637-652.

Leenders, M. R. "Supplier Development." 1966. Journal of Purchasing 2(4), 47-62.

Leenders, M. R., and D. L. Blenkhorn. 1988.Reverse Marketing: The New Buyer-Supplier Relationship. New York: The Free Press.

Wathne, K. H. and J. B. Heide. 2004. “Relationship governance in a supply chain network,” Journal of Marketing, 68, 73-89.

Thomas Choi and Zhaohui Wu

All eight commentators agree that considering triads in supply networks is important for our field to move forward. Day offers strong support when he states that taking the leap from dyads to triads “could mark an important turning point in supply chain research.” Baker comments that the present consideration of triads “lays an excellent foundation for theory building and testing in supply chain management.” Schoenherr echoes this new focus on triads provides “a refreshing and exciting perspective.” Krause is more cautious. He says “Choi and Wu’s call for triadic research is certainly valuable and worthwhile. However…my enthusiasm for their call is not unbridled.”

We will first summarize the key points of each of the eight commentators. Then, we will try to integrate the key points raised by the commentators. We will close with our reactions and thoughts on future research implications.

Summary of Comments

Breen sees value in studying triads because it leads to network considerations. Buyer-supplier relationships extend out to supply chains crisscrossing the globe. With larger number of buyers and suppliers linked up together “directly and indirectly,” we have no choice but to move beyond dyads and consider such relationships as triads in the supply network context. Day points out that in the past, the field of OSM has focused on interactions between firms (i.e. nodes). He then emphasizes the importance of studying interactions between relationships (i.e. links) in supply networks. The smallest unit where interactions between relationships can be studied is the triad. He points out that interdependent relationships affect the behavior and performance of firms within a network. He offers an example of tolerance for errors or attacks in supply networks and says that such tolerance “is influenced by the structural topology of relationships” between firms.

Terziovski subjects the present essay to seven criteria proposed by Whetten (1989) for evaluating a conceptual work. He offers a passing grade and approves that the essay “pushes the current paradigm of supply chain theory” and should be “of high interest to scholars” in our field. Spring introduces the industrial networks literature, largely based in Europe, as an intellectual foundation for considering triads in supply networks. He first points out how this particular genre of the literature has “frustrated some in OM” by being merely descriptive and lacking in prescriptions. He subsequently articulates two challenges—development of prescriptive theories and application of a network perspective to traditional OM issues, such as capacity management, planning and control, and contracting.

Using the cooperative and adversarial dichotomy in buyer-supplier relationships, Baker discusses several potential triadic relationships. Assuming A to be the buyer and B and C to be competing suppliers in Figure 1, having all three cooperative links would be equivalent to a parallel sourcing practice. Baker points out several other triadic relationship combinations, where each reflects real-world relationship dynamics in supply chains. He also extends his ideas by brining in the time element—how could one form of triadic relationships evolve over time into another form? Indeed, triads bring us closer to “more real-world related tactical or strategic issues,” Rodriguez and Leisten assert. They propose several studies that could build on the triadic perspective. For instance, each of the three nodes in a triad has an extended supply network that it is connected to. So researchers might consider an “extended triadic view” as a next step.

Schoenherr argues that global sourcing has led us to less single-sourcing models and more dual or multi-sourcing models, which equates to “the rising importance of triads.” He then describes a reverse-auction service provider as the middleman between the buying company and the part suppliers and also how this triadic arrangement evolved over time. Krause provides more examples from past literature that have considered triads and networks— the reverse marketing concept where a dyadic relationship (i.e. buyer-supplier) is affected by the presence of a new node (i.e. a new supplier) and the studies that have conceptualized dyads as being embedded in larger networks. He sees value in considering triadic perspectives, but issues caution on two fronts. One, collecting true dyadic data has been difficult, and collecting data on triads will be even more difficult. Two, when considering a supply base as opposed to supply networks, dyads appear as the basic building block, and these dyads are then embedded in the larger “marketplace amidst other customers, suppliers, and competitors.”

Integration of Comments

There appear to be four common themes in the comments. Firstly, Breen and Schoenherr bring out the issue of globalization. Globalization necessarily extends our supply networks across different political, cultural, and legal boundaries, which complicates our buyer-supplier relationships and brings salience to triadic considerations. It is not hard to imagine two of the three links in a triad crossing the national boundary, while the other link remaining in the same country. It would certainly be worthwhile to investigate the dynamics of interactions among links in such globalized settings.

Secondly, the structural issues come to the fore in the comments of Day, Baker, and Schoenherr. They call out how different combinations of buyers and suppliers can arise in a triadic structure. The three firms in a triad can be organized across two tiers or even three tiers. Triads are organized across two tiers when one buyer is linked with two competing suppliers or two buyers are linked with one common supplier. Triads span across three tiers when there is a buyer who receives goods from one supplier and delivers products to a customer or when there is a buyer who buys from a second-tier supplier through a first-tier supplier. For instance, Day points out how the structure of one buyer and two competing suppliers would bring an entirely different perspective to the “bullwhip effect.” Schoenherr addresses the case of a triad cutting across three tiers—a buyer buying parts from suppliers (second tier) through an intermediary (top tier) and then eventually eliminating this intermediary. We believe there can be many other fascinating applications of triadic structures in supply networks.

Thirdly, Terziovsky, Baker, Spring, and Schoenherr call for more theory building using triads. Terziovsky ponders what type of propositions might there be in triads in supply networks. Spring says that we need to take these ideas of triads and offer “prescriptive frameworks that … (provide) useful theory.” Baker in particular takes the structural issues one step further and ponders what might happen if we had different combinations of collaborative and adversarial relationships among the three companies in the triad. Also, he pushes the envelope further by urging us to consider the evolutionary dynamics of such relationship combinations in triads—which relationship combinations might signify a steady state and which combinations signify transient states. He even suggests the idea of triads with structural holes. If two players in a triad are not linked together, this presents a triad with a structural hole. For instance, the example of reverse marketing in Krause’s comment actually reflects a triad with a structural hole. Reverse marketing worked because there was a structural hole between two suppliers. Baker’s comments are indeed fascinating, and we would like to offer a recent publication (Choi and Wu, 2009) that considers a buyer-supplier-supplier triad and how different combinations of collaborative and adversarial relationships evolve from transient states to a steady state by applying the balance theory. That publication also considers structural-hole cases of triads and offers a series of propositions.

Lastly, there are methodological issues in investigating triads. Krause correctly points out the difficulty that researchers would inevitably face when attempting to collect data on triads. We are aware of a few recent studies that have attempted to collect triadic data, and based on our understanding, Krause’s concerns are fully justified. However, with open eyes, we move forward. For instance, Rodriguez and Leisten bring our attention to the need for quantitative and empirical studies that would examine the theoretical arguments that are being posed by the present essay and also by commentators. Spring suggests adopting case studies to capture contextually rich dynamics involved in triads. We think behavioral experiments using subjects representing different companies might be applicable in studying triads. Another possibility might be agent-based simulation and modeling studies using empirical data, as suggested in the special issue of the Journal of Operations Management on empirical simulation edited by Shafer and Smunt (2004).

Closing Thoughts: Looking to the Future

Triads are the fundamental building block of a supply network. As such, a triad offers a practical frame of reference and a broader perspective on network relationships. The commentators suggest analysis of other types of triadic relationship contexts that are practical and relevant to management practices. For instance, consider the problem of creating a sustainable supply network. In a recent study by Pullman and Wu (2009) concerning natural beef production, the authors find that, to truly understand the network, one must consider the interactions of at least three entities at a time. In this network, the key members include ranchers, the feedlot, meat packers, natural food retailers such as Whole Foods Market, and the certification organization, etc. All these members must work together to ensure that they follow sustainable ranching practices, humane treatment of the animals, supply chain traceability and transparency requirements, and holistic resource management principles. A dyadic analysis of the relationship between any two entities, while important to capture the specific operations practices, will surely miss the opportunity to understand the value underlying the network. Instead, when we examine the interactions among any of the three entities, we begin to see how the network as a whole makes decisions to reconcile and makes compromises, sets rules to reduce opportunistic behavior, and settles differences.

There are other recent studies that suggest the triadic relationship as a building block on which larger networks are created (Madhavan, Gnyawali and He, 2004; Obstfeld, 2005). For instance, Obstfeld (2005) proposes that the tertius iungens mechanism actually explains how a large network grows out of small triads. Specifically, he proposes that, in the case of a triad with a structure hole, if the middleman introduces the two otherwise disconnected members to each other, the newly connected members will be more likely to reciprocate the middleman by introducing new members into the triad. Thus an initial triad will grow and the reciprocity process will perpetuate to generate new relationships.

We agree with Krause that we need to do a better job of studying dyads and that dyads play a major role in a supply base. As argued in Choi and Krause (2006), a supply base is a subset of a supply network that contains the suppliers that are visible from a focal buying company. However, dyads are not the fundamental building block of a supply base, triads are. On the one hand, if one considers one buyer and one supplier in isolation, then we have a case of a dyad and there is no supply base. On the other hand, as soon as we consider the embedded nature of a supply network and add another supplier to form a supply base, then we immediately form a triad—in this case, a buyer-supplier-supplier triad. In fact, in our recent publication, we offer a theoretical consideration of this particular triad (Choi and Wu, 2009) and suggest areas of future research.

As attested by the commentators, we need to ask relevant questions about network structures and the complexity of network relationships. Such research endeavors must be carried out in light of addressing core OSM issues such as supply chain design, supply uncertainty, risks, and bottleneck management, etc. We need to learn to collect appropriate data or search for large-scale data sets that can be applied to the study of triads and larger supply networks. Furthermore, we need to develop an OSM theory of triadic relationships in supply networks. We are still in the early stage of understanding the triadic supply relationship, and further grounded theory-building effort is needed. The present essay and subsequent comments have noted a few empirical studies that have looked at triads. But we need more. Without being grounded in empirical evidence, we may fall into the trap of reinventing many theories without enough understanding, as cautioned by Schmenner in a separate essay in this OSM forum.


Choi, T. Y. and Krause, D. 2006. The Supply Base and Its Complexity: Implications for Transaction Costs, Risks, Responsiveness, and Innovation. Journal of Operations Management. 24 (5): 637-652.

Choi T. Y. and Wu, Z. 2009. Triads in Supply Networks: Theorizing Buyer-Supplier-Supplier Relationships. Journal of Supply Chain Management. 45 (1):8-25.

Madhavan, R., Gnyawali, D. R. and He, J. 2004. Two’s Company, Three’s A Crowd? Triads in Cooperative-Competitive Networks. Academy of Management Journal. 47(6): 918-927.

Pullman, M. and Wu, Z. 2009. A Value-driven Sustainable Supply Network: The Case of Country Natural Beef Cooperatives. Working paper.

Shafer, S. M. and Smunt, T. L. 2004. Empirical Simulation Studies in Operations Management: Context, Trends, and Research Opportunities. Journal of Operations Management. 22 (4): 345-354.

The comments to this entry are closed.

Blog powered by Typepad