35 
 

  
 
 
 
 
 
 

Foundations of  

Competitive Intelligence System  

to form Business Coalitions 

 
Olivier Mamawi 

 
LARA / ICD 

12, rue Alexandre Parodi, 75010 Paris, France  

omamavi@gmail.com 
 

Received June 3, revised form 10 September, accepted 22 September 2012 
 

 
ABSTRACT: This study shows how a business can identify the networks allowing to form coalitions to obtain 

French procurement contracts. To this end, we have represented, by a graph, the 2008 co-branding system. We 

have detected, in this graph, 1360 strategic networks of which the organization reveals, on the one hand, 

identical characteristics within business networks, and on the other hand, the role of the dominant parties as to 

access to industrial labor. From these results, we propose a network cartography allowing us to consider new 

applications for competitive intelligence.  

 
KEYWORDS: Coalitions, competitive intelligence, procurement contracts, cartography, network analysis. 

 

1. Introduction 

 

French state procurement contracts, regional 

authorities and public corporation contracts add up 

to over 80 billion euros per year. To ease the way 

for small and medium-sized companies, the reform 

of the French state procurement code, initiated in 

January 2006 and applicable since January 2008, 

encourages responses to invitation to tender through 

co-transaction. Co-transacting is the process by 

which tendering companies can ally  with a group 

of companies, to tender a collective offer when they 

cannot, on their own, assume the necessary 

competences and resources. 

Effectively, co-transacting means forming 

coalitions, that is to say, “temporary alliances, 

which are devised, if not negotiated, amongst those 

who participate” (Lemieux, 1998). But organizing a  

 

 

coalition, to share or to distribute resources to 

respond to the needs of a project, is not easy for 

companies with sometimes different interests; So, 

how can a company analyze its (complex) 

environment to form a coalition and obtain 

markets? 

We begin with the following assumptions: the 

network is a business structure, and the coalition is 

a specific structure of the network. We present as 

the main hypothesis that a coalition is formed 

according to the structure of relations. In 

consequence, we consider that a company must be 

able to interpret its strategic environment as to find 

good partners, and to understand how the structure 

of interactions determines the formation of the 

coalition.  

Available for free online at https://ojs.hh.se/ 

 

Journal of Intelligence Studies in Business 2 (2012) 35-41 

mailto:omamavi@gmail.com
https://ojs.hh.se/


36 
 

The Management science literature has shown a lot 

of interest in network alliances. But, studies on the 

networks, up to now, have mostly looked into the 

reasons linked to their formation than an 

understanding of the inter-organizational relations 

that they imply. The developments surrounding 

control (piloting) of the networks have been, for the 

most part, treated as annexes to more general 

problematic, essentially economic, sociological and 

strategic (Gulati & Nohria, 2000). The aim of this 

communication is to suggest a new perspective 

concerning non-cooperative projects (Yi & Shin, 

2000), mainly reached through operational 

research. The issue today is to go beyond the limits 

of game theory, to aim for an actors theory (Massé 

et al., 2000).To this aim, we wish to suggest the 

basis of a competitive intelligence system which 

will permit a reticular perspective of the strategy, 

keeping in account the complex environment (that's 

to say interactive and dynamic). 

We will attempt, in a first stage, to present the 

theoretical basis of emerging networks within the 

markets. Then, we will explain the method used to 

encapsulate and study numerous alliance 

relationships. Finally, we will discuss the results 

obtained through a cartography that enables to 

localize partners of strategic networks. 

 

2. Contributions of economic sociology 

 

Interorganizational relationships can be considered 

as “a coordinated system of heterogeneous parties, 

developing transactions founded on a co-

operational relationship, as to collectively pursue a 

shared aim” (Voisin et al., 2004). This system is the 

result of a process within which the partners bond 

to exact a mutual benefit. That is to say, a network 

of parties which is subjected to business circles, and 

forced to define a balance with its environment. As 

defined by Assens (Assens, 2003), the concept of a 

network puts forward the inter-connection of parties 

able to participate in exchanges. The link between 

the parties gives the nature of the exchanges, their 

periodicity, their strength, their density. The parties 

hold positions that are apt to evolve, but that testify 

to a role or a function held in face of the other 

parties.  

To understand the frame of a network, many 

authors (Calon, 1989), (Cohendet & Diani, 2003), 

prefer the study of relationships between members, 

rather than a focus on the nature of the parties 

themselves. They thus judge an approach to 

networking as a form of transitional organization, 

of a hybrid tendency, between the market and the 

hierarchy (Williamson, 1985).  

Which is why Granovetter (Granovetter, 1985) 

considers that the functioning of a market depends 

on business conditions. He thus shares, with other 

protagonists of economic sociology (White, 1981; 

Baker, 1984), the conviction that economic action is 

a social action, led by various motivations: 

sociability, recognition, social status and power.   

To prove the integration of economic actions within 

business relation systems Granovetter (Granovetter, 

1985) suggests the concept of “embeddedness”. He 

makes a point of proving that business relations and 

institutions allow a market to function. A relation is 

then distinguished by its content (the exchange of a 

resource), its direction and its strength (level of 

exchange).   

By the same perspective, interorganizational 

relationships can be considered as a means to  

control environment uncertainty accessing 

complementary resources. The rating of a company 

becomes clear when examining the relations 

network into which it is encompassed. This 

“strategic network” (Gulati & Nohria, 2000) relates 

to all the links held with partners, whether clients, 

suppliers, competitors or subcontractors.  

As such, preference goes to working with a 

particular partner, depending on the level of 

interactions, sufficient to obtain an optimal 

transaction, through optimal human links rather 

than economic ones (Voisin et al., 2004). Parties 

using “strong links” (Granovetter, 1985) imply 

frequent contacts, a supply of reciprocal services, 

and this justifies a sharing of resources.  

With this approach, the social relationship helps 

regulate opportunistic behavior, ensures data 

sharing, and facilitates collective solutions to 

problems.  

Uzzi (1996) has shown that a relationship based on 

trust can enhance company advantages (shared risk, 

ability to react to markets changes, organizational 

apprenticeship), that purely commercial 

relationships cannot offer. And as explained by 

Gulati and Singh (Gulati & Singh, 1998), the 

economic parties tend to exchange, in priority, with 

known parties, to diminish uncertainty linked to 

transactions.  

When a party knowingly mobilizes its network to 

follow a given strategy, trust becomes a resource of 

the network, that allows to save on the transaction 

costs (Williamson, 1985), to share resources 

(Richardson, 1972), to reduce measures of control 

or incitement (Gérard, 2000), to avoid asymmetrical 

data amongst partners (Akerlof, 1970), to contain 

opportunistic attitudes (Olson, 1971). As such, one 

of the main aims of studies on networks was to put 

to light the cohesion within members as a means to 

access, share or control resources. Beyond 

opportunistic explanations, linked notably to 

proximity (geographical or social), two main 

principles can explain the formation of links and 

interactions (Lin et al., 2001). When parties have 

the same level of resources and that they are moved 

by recognition, exchange or protection of their 



37 
 

business status, we speak of homophile interactions. 

In this case, the aim of links is to maintain a level 

of resources. When, on the contrary, the 

hierarchical levels are different, and the actions are 

motivated by the need to obtain better or 

supplementary resources, this becomes heterophile 

interactions.  

In a more general way, parties interact with other 

parties because they seek backing or safety which 

allows them to control their turbulent environment, 

and find a certain stability. This search for control 

gives rise to regularities (relationships) which form 

the basis of business networks. The structure of 

these regularities can be identified. It's the principal 

of structural equivalence (White, 2001). As 

reminded by Grosseti and Godard (2007), two 

entities are structurally equivalent, within a 

network, if they occupy the same place or the same 

position, that is to say, if they have the same trade 

relations (or relatively similar relations) with given 

parties. The notion of structural equivalence thus 

allows to return to the classic notion of a role (or of 

position), but from a purely structural point of view, 

through a network analysis, without conjecturing on 

the weight of these roles. 

 

3. Analysis 
 

Social Network Analysis considers society as a 

system of business parties linked by relationships. It 

is an adapted method to understand and formalize 

complex phenomenon calling for an interactive 

system of relations. As such, this method allows to 

describe, and reconstruct, a network, in a simplified 

way by a graph. The graph represents the 

interactions between objects related by links. The 

development of a quantitative method, originally 

issuing from sociometry and completed by the help 

of the theory of graphs, allows today to put forward 

a set of properties which form a changeable 

topology. For network analysis (Borgatti et al., 

2009), three main dimensions (table 1) can be used.  

The first aims to identify the networks and to 

describe the manner in which the structures of these 

networks can burden the members. The 

connectivity is an indicator which allows to define 

the limit of the network in a chart. There is a 

network if there is always a link between two 

summits of the whole. A network is thus a related 

component of a graph.  

The second dimension enables to identify the 

position, more or less dominant, of a party in the 

network. It can be assessed through centrality as 

defined by Freeman (1979). 

The degree of centrality shows the popularity of a 

party within the network, that is to say, the number 

of direct connections of one company with the 

others. The centrality of proximity identifies the 

companies closest to the sources of power and 

influence, that is to say, swiftly reached by the other 

members of the network. 

The third dimension aims to define the cohesion of 

homogeneous groups within the network, to 

analyze structural similarities of the network. 2 

parties are thus structurally equivalent if they have 

identical relationships with the other parties of the 

network (White, 1981). Partitioning techniques 

allow to detect groups of parties of structural 

equivalence (Navarro & Cazabet, 2011). 

 

 

 

 

 

 

 

 

 

4. Methodology  
 

To analyze business networks, we have chosen to 

study the responses, through co-transacting, to 

invitation to tender in French public markets. Our 

study is based on the analysis of coalition 

relationships within company groups. Let's be 

reminded that invitation to tender procedures find 

themselves in a system aiming to enhance the 

transparency of deals within the two categories of 

contributors. The contractors, meaning with the 

power to adjudicate, can be: the State (ministries), 

territorial collectivities (administrative districts, 

departments, regions), public establishments linked 

to the State and to collectivities, public 

establishments outside of a business and/or 

commercial character (universities, schools, certain 

museums, etc.). The companies making an offer are 

“submitters”.  

Data used for this study issues from the attribution 

notices from of the French Official Journals 

(BOAMP). The French Official Journals publishes 

the transactions attributed by a French public 

guarantor, for market sums above €4.000 before 

tax.  

From transactions made in 2008, we have selected, 

with key words, the transactions held only by 

groups of companies. Table 2 details the census of 

the observed businesses. It shows the number of 

parties that we identified within the groups. 

 

 

 



38 
 

Obtaining this data goes through a multi-stage 

process: extraction, cleaning, filtering, formatting, 

dedoubling and indexing. The nature of the data 

used to analyze the groups of companies is held by 

three variables. The first is allotted to the 

identification of the parties (business reasons). The 

second is ascribed to the type of beneficiaries of the 

transactions (company groups). The third variable 

concerns the cooperative relation which links the 

parties within a group. We have considered that 

there is a cooperative relationship with two 

companies when they obtain a market within the 

bounds of co-transaction (a belonging to a group of 

companies).  

 

5. Results  
  

We have organized the relational data in the form of 

a list of adjacency. The list sets out, for each 

company, all the companies close to it. Then, we 

used the software GraphViz (www.graphviz.org is a 

series of open source tools created by the research 

labs of AT&T, which allow to represent and analyze 

graphs) to obtain a complete graph of the co-

operation relations within the French public market 

in 2008.  

 

The graph highlights numerous sub-graphs (related 

components) within which there exists a link 

between any two clusters. These sub-graphs are 

strategic networks. The graph is composed of 1360 

strategic networks having between 2 and 2233 

clusters. Figure 1 represents the repartition of 

strategic networks according to their size. The 

visualization of the complete graph shows company  

aggregates which correspond to a concentration of 

links over a limited amount of companies. 

 

 

Figure 1: network distribution according to size 

 

To understand the manner by which the strategic 

networks are constituted, we have isolated them. An 

in-depth study of the largest strategic network 

allows to produce a number of indicators on the 

structure of relations: 

 
  global density of the graph is low 

(0,0004). It reports the number of existing 

links and the the number of possible links;  

 local density, or clustering coefficient, is 
high (0,47). It corresponds to the 

probability of two close members of a 

same party being linked; 

 the average distance between two 
companies is 6,45. This distance 

corresponds to the length of relationship 

links between random members of the 

network.  

 
To evaluate the position of parties within the largest 

related component, we have measured degree 

centrality. The distribution of degrees (figure 2), 

that is to say the number of connections a company 

has in the network, is heterogeneous. The majority 

of companies have a low degree, and only few 

companies have a strong degree. Put more clearly, 

we are close to the Zipf-Paretto law by which 20% 

of companies attract and generate 80% of network 

links. These companies are shown up statistically, 

but also visually in the graph.  

 

Figure 2: distribution of degrees within the network 

 

6. Discussion 

 

The analysis of the largest strategic network in 

French public markets reveals non-trivial 

characteristics common to other business networks, 

such as: acknowledgment networks (two 

individuals are related if they know each other), 

physical contact networks (two individuals are 

related if they have been in physical contact), 

collaboration networks (two individuals are related 

if they have worked together), exchange networks 

(two entities are related if they have exchanged for 

example an email). As such, the great business 

networks all possess a low global density, a strong 

local density, shortcuts to the summits, a 

http://www.graphviz.org/


39 
 

heterogeneity of degrees and a low average degree. 

These characteristics are generally attributed to 

graphs of a large field (Strogatz, 2001), in reference 

to the “small world” in the experience of Travers 

and Milgram (Travers & Milgram, 1969). These 

characteristics thus compose a organizational model 

of inter-organizational networks. Furthermore, the 

distribution of degrees within the network 

highlights the existence of zones more densely 

connected than others. These zones correspond to 

groups of companies more strongly connected to 

each other than to others. They correspond to an 

entity of companies with common points and 

between which the links are naturally stronger.  

 

From a more general point of view, the complete 

graph of the co-joining of companies to a group of 

companies shows us a number of indicators to 

describe the phenomena of strategic networks 

within the public markets. But, to profit from these 

indicators, we need to determine, on an ongoing 

basis, the position occupied by each company, to 

understand its role and importance, or which are the 

affinities allowing a company to acquire or keep a 

central position in the network.  

 

This purpose renders the use of cartography 

essential. The cartography stems from the graph of 

each company with its alliance relationships. The 

chart represents the strategic space within which 

influences are played out, and the topology allows 

to classify the companies according to their 

relationship proprieties. Cartography then becomes 

a reticular lay-out, representing business 

interactions in the public markets. This lay-out is a 

space in which the parties communicate 

information and interact with each other.  

 

In fino, the main asset of cartography is its ability 

to analyze, on an ongoing basis, transactions and 

relationships within the public markets. This 

network analysis allows to: 

 represent the companies and their 
relationships, 

 navigate through the company networks, 

 identify the position of each company on 
the market, 

 measure the strength of the links 
(affinities) amongst the companies, 

 determine the role and the status of each 
company. 

 

From this network cartography, we can recognize 

and act on the strategic environment of a company, 

and use it as tool of competitive intelligence. In 

effect, competitive intelligence can be considered as 

a process with the aim of reducing the part of 

uncertainty in the taking of any strategic decision 

(Revellic, 1998). So, to be intelligent is the ability 

to find a solution in a complex environment (Massé 

et al., 2006). This corresponds to the capacity of 

absorption (Zahra & George, 2002) of information 

to a strategic end, that is to say to its acquisition, its 

assimilation, its processing and its development. 

 

7. Conclusion and future research 

 

This study, carried out through data issued by the 

French Official Journals (BOAMP), is an analysis 

of French public markets throughout 2008. Thus, 

from 54.181 transactions carried out between the 

contractors and the submitting companies, we were 

able to observe 4.203 transactions undertaken by 

coalition companies. The growth of these 

transactions, with relational information, allowed us 

to set up a structural analysis. From this, we 

constitute a data basis on interorganizational 

relationships.  

The main contribution of this study is to offer a 

framework of strategic network analysis within 

public markets. We set up a cartography giving a 

graphic representation of alliance networks. It eases 

their visualization and reveals non-trivial 

characteristics common to other business networks. 

But the aim of this study is to propose the basis of 

an competitive intelligence system. For competitive 

intelligence can be considered as a process aiming 

to reduce the uncertainty factor in the taking of any 

strategic decision (Moinet, 2011). The aim, then, 

would be to provide companies a system of new 

links references, to help them form coalitions 

within public markets.  

To devise an competitive intelligence system, we 

need to keep in account that networks are where 

business interactions are made and undone. They 

can evolve, but also disappear. So we must keep on 

with this study. To this end, our perspective is to 

follow up, in a dynamic way, the evolutions and 

movements of network alliances within public 

markets over a period of several years. The 

longitudinal study will try to put forth the 

emergence and the evolution of strategic networks 

within public markets. The analysis will be based 

on the way in which businesses link themselves up 

and would use the preferential attachment concept 

described by Barabasi and Albert (1999). 

Then, the approach that we are considering, the 

building a system of recommended links would be 

the following: Through a longitudinal study over 

several years, we would be able to predict new 

emerging links, which will connect the companies 

already present in the networks, but which had 

never been linked previously. Supervised training 



40 
 

techniques (Benchettara et al., 2011) could, then, be 

applied to build a prediction pattern of new links 

within the networks to help businesses form 

winning coalitions. 

 

Making the decision to form a coalition within 

public markets means constructing a particular 

structure of links within a complex system of 

relations. Building a coalition can be considered as 

a plan of strategic decisions, in which the parties 

(the companies) can group to obtain earnings 

(transactions) through their choices (temporary 

alliances) and keep within the rules (implicit or 

explicit) which frame or curb their their 

performances. The earnings depend upon the 

decisions of parties exterior to the plan, and of 

which the distribution does not respond to a known 

probabilist law (uncertain environment). This 

decision will be defined by the taking into 

consideration of multiple criteria linked to the 

scores of relational properties. It will imply, 

amongst others, the diversity of considered options 

and the in-depth assessment of each of them, in 

considering possible profits. 

Acknowledgment: The author wishes to thank 

Frédéric AIDOUNI for his collaboration in 

collecting data. 

 
References 
 

Akerlof, G. 1970. “The Market for ‘Lemons’: 

Quality Uncertainty and the Market 

Mechanism.” Quarterly Journal of 

Economics 84: 353–374. 

Assens, C. 2003. “Le Réseau D’entreprises : Vers 

Une Synthèse Des Connaissances.” 

Management International 7 (4): 49–59. 

Baker, W.E. 1984. “The Social Structure of 

National Securities Market.” American 

Journal of Sociology 96 (3): 589–625. 

Barabàsi, A.L., and R Albert. 1999. “Emergence of 

Scaling in Random Networks.” Science 286: 

509–512.  

Benchettara, N., R. Kanawati, and C. Rouveirol. 

2010. “Apprentissage Supervisé Pour La 

Prédiction De Nouveaux Liens Dans Des 

Réseaux Sociaux Bipartie.” In Actes de la 

17iéme rencontre de la société francophone 

de classification SFC2010, 63–66. 

Borgatti, S., A. Mehra, D. Brass, and G. Labianca. 

2009. “Network Analysis in the Social 

Sciences.” Science (New York, N.Y.) 323 (13): 

892–5.  

Callon, M. 1989. La Science et ses réseaux: genèse 

et circulation des faits scientifiques. éditions 

la Découverte. 

Charreaux, G. 2000. “La Théorie Positive De 

L’agence : Positionnement et apports.” Revue 

d’Économie Industrielle 92: 193–214. 

Cohendet, P., and M. Diani. 2003. “L’organisation 

comme une communauté de communautés 

croyances collectives et culture d'entreprise,” 

Revue D’économie Politique 113: 697–720. 

Freeman, LC. 1979. “Centrality in Social Networks 

Conceptual Clarification.” Social Networks.  

Granovetter, M. 1985. “Economic Action and 

Social Structure: The Problem of 

Embeddedness.” American Journal of 

Sociology 91 (3): 481–510.  

Grossetti, M., and F. Godart. 2007. “Harrison 

White : des réseaux sociaux à une théorie 

structurale de l’action.” SociologieS (17).  

Gulati, R, and N Nohria. 2000. “Strategic 

Networks.” Strategic Management Journal 

(4): 1–34.  

Gulati, R., and H. Singh. 1998. “The Architecture 

of Cooperation: Managing Coordination, 

Uncertainly and Interdependence in Strategic 

Alliance.” Adminitrative Science Quartely. 

Lemieux, V. 1998. Les Coalitions : Liens, 

Transactions et Contrôles. Presses 

Universitaires de France - PUF.  

Lin, N., K. Cook, and R. Burt. 2001. Social 

Capital: Theory and Research.  

Massé, G., C. Marcon, and N. Moinet. 2006. “Les 

Fondements De L’intelligence Économique : 

Réseaux & Jeu D'influence.” Market 

Management 6 (1): 84.  

Moinet, N. 2011. Intelligence économique : mythes 

et réalités. CNRS. 

Navarro, E., and R. Cazabet. 2011. “Détection de 

communautés, étude comparative sur graphes 

réels.” Information Interaction Intelligence 11 

(1): 77–93. 

Olson, M. 1971. Logic of collective action: public 

goods and the theory of groups. Cambridge, 

Harvard University Press. 

Revellic, C. 1998. Intelligence Stratégique sur 

Internet. Ed. Dunod. 

Richardson, G. 1972. “The Organization of 

Industry.” Economic Journal 82: 883–896. 

Strogatz, S H. 2001. “Exploring Complex 

Networks.” Nature 410 (8): 268–76. 

Travers, J., and S. Milgram. 1969. “An 

Experimental Study of the Small World 

Problem.” Sociometry 32 (4).  

Uzzi, B. 1996. “ The sources and consequences of 

embeddedness for the economic performance 

of organizations: the network effect,” 

American Sociological Review 61 (4): 674–

698. 

Voisin, C., S. Ben Mahmoud-Jouini, and S. 

Edouard. 2004. Les Réseaux : Dimensions 

Organisationnelles Et Stratégiques. 

Economica.  

White, H. C. 1981. “Where Do Markets Come 

From?” American Journal of Sociology 87 

(3): 517–58.  

 



41 
 

Williamson, O. 1985. The Economic Institutions of 

Capitalism. Free Press-Collier Macmillan. 

Yi, S.-S., and H. Shin. 2000. “Endogenous 

formation of research coalitions with 

spillovers.” International Journal of 

Industrial Organization 18 (2): 229–256. 

Zahra, S.A., and G. George. 2002. “Absorptive 

Capacity: A Review, Reconceptualization, 

and Extension.” Academy of Management 

Review 27 (2): 185–203.