Stakeholder analysis and engagement in projects: From ... · PDF fileStakeholder analysis and engagement in projects: From stakeholder relational perspective to stakeholder relational

Available online at www.sciencedirect.com

ScienceDirect

www.elsevier.com/locate/ijpromanInternational Journal of Project Management 32 (2014) 1108–1122

Stakeholder analysis and engagement in projects:From stakeholder relational perspective to stakeholder

relational ontology

Stephanie Missonier a,⁎, Sabrina Loufrani-Fedida b

a HEC Lausanne-UNIL, Faculty of Business and Economics, Dorigny ISI, Bat Internef 131, 1015 Lausanne, Switzerlandb University of Nice-Sophia Antipolis, Research Center GREDEG UMR 7321 UNS-CNRS, 250 rue Albert Einstein, 06560 Sophia-Antipolis, France

Received 21 October 2013; received in revised form 21 February 2014; accepted 25 February 2014Available online 3 April 2014

Abstract

This paper investigates the stakeholder analysis and engagement in the field of project management. In response to the limits of prior studies,we propose a relevant conceptual approach by moving from a stakeholder relational perspective, anchored in recent studies on Social NetworkTheory, to a stakeholder relational ontology, anchored in Actor–Network Theory (ANT). We apply our approach to read and understand alongitudinal case study of an Information System (IS) project. Our most important findings suggest that this approach based on ANT improvesstakeholders' analysis of and engagement in a project by shedding light on the dynamic and emergent nature of the relationships, since wedemonstrate that the nature, roles, and relations between stakeholders co-evolve with the project's definition and trajectory. Consequently, we canprovide project managers with a relevant approach that informs them about what to observe in stakeholder project networks, as well as how andwhen to observe them.© 2014 Elsevier Ltd. APM and IPMA. All rights reserved.

Keywords: Managing stakeholder; Actor–Network Theory (ANT); Relational ontology; Information System (IS) project

1. Introduction

In the field of project management, many authors have clearlyhighlighted the extraordinary importance of stakeholders inprojects (Beringer et al., 2013; Cleland, 1986; Freeman, 1984;Littau et al., 2010; Savage et al., 1991; Winter et al., 2006). Themanagement of a project's stakeholder means that the project isexplicitly described in terms of the individuals and institutionsthat have a stake or an interest in the project. Whatever the natureof the project, various researchers (Achterkamp and Vos, 2008;

⁎ Corresponding author. Tel.: +41 21 692 35 95.E-mail addresses: [email protected] (S. Missonier),

[email protected] (S. Loufrani-Fedida).

http://dx.doi.org/10.1016/j.ijproman.2014.02.0100263-7863/00/© 2014 Elsevier Ltd. APM and IPMA. All rights reserved.

Brown and Jones, 1998) have acknowledged that project failureis generally not the result of lacking or ineffective projectmanagement practices, but of inappropriate social interactionsbetween the project stakeholders.

In this study, our focus is on the stakeholder analysis of andengagement in projects. Stakeholder analysis is an essential partof stakeholder management (Aaltonen, 2011; Freeman, 1984;Jepsen and Eskerod, 2009; Magness, 2008; Mitchell et al.,1997). In turn, stakeholder engagement includes communicat-ing with, involving and developing relationships with stake-holders (Chinyio and Akintoye, 2008; Greenwood, 2007). Asnoted by Yang et al. (2009a,b), stakeholders should be engagedas early as possible and this engagement is essential forstakeholder analysis and decision-making.

These matters seem clear and sufficiently addressed. Inreality, this perception is unfortunately a pipe dream. Despite

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijproman.2014.02.010&domain=pdf

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.ijproman.2014.02.010




1109S. Missonier, S. Loufrani-Fedida / International Journal of Project Management 32 (2014) 1108–1122

its popularity, the literature on stakeholder analysis of andengagement in projects suffers various limitations (Jepsenand Eskerod, 2009; Littau et al., 2010; Pacheco and Garcia,2012; Yang et al., 2009a,b). We classify these limitationsinto three main improvement areas: the relevance, thedynamic, and the emergence of stakeholder analysis andengagement.

Firstly, prior studies have a provided rich and solid basis foridentifying, classifying, and categorising stakeholders as wellas understanding their behaviour to manage them better(Crawford, 2005; Cummings and Doh, 2000; Mitchell et al.,1997; Savage et al., 1991). Despite these rich classifications,various studies call for improvements in these approaches, asthey do not provide methods and tools with which to identifyall stakeholders and their interests (Pouloudi and Whitley,1997; Yang et al., 2009a,b). In particular, Jensen andSandström (2011) underline that stakeholder theory may befailing to reach one of its central aims: being useful tomanagers.

Secondly, Parmar et al. (2010) call for improved descrip-tions of how firms manage their relationships with stakeholdersover time, i.e. for longitudinal analyses. According to Eskerodand Vaagaasar (2012), to date, the majority of projects do notregard stakeholder management as a dynamic and on-goingprocess. Stakeholder analysis is therefore not only a front-end,but also a longitudinal process.

Thirdly, our literature review reveals that prior studies havenot addressed the emergent nature of relationships. Inparticular, prior studies on stakeholder analysis from the SocialNetwork Theory anchored in a relational perspective haveactually shown developments by studying the “resultant”effects of stakeholder relationships (Bourne and Walker,2005; Rowley, 1997). Nevertheless, the “emergent” effectshave been understudied. The studies overlook the importanceof the emergent nature of stakeholder networks, i.e. theco-evolution of the stakeholder identity and the project overtime.

Consequently, our literature review points out two mainimplications and necessities. The first implication is toconsider developing a relevant framework to identify the keystakeholders and maintain good relationships throughout theproject. The second implication is that this framework shouldintegrate stakeholder relationships' dynamic and emergentnature and, thus, project management stakeholders' timeperspective.

In the face of these implications, we propose to develop arelevant conceptual approach based on a theory rarely used inthe understanding of the stakeholder analysis and engagementin the field of project management. This theory, called theActor–Network Theory (ANT), is anchored in a relationalontology. Accordingly, our research question is: How can arelevant approach based on ANT provide a dynamic andemergent stakeholder analysis of as well as engagement in aproject throughout its duration? In other words, our maincontribution is to demonstrate how ANT can be used as arelevant conceptual approach that allows stakeholder analysisand engagement during the project.

The remainder of the paper is organised as follows: webegin the article with an overview of the stakeholderliterature, taking the movement from a dyadic perspective toa relational perspective into consideration (1). We argue that arelational ontology anchored in ANT offers a worthwhileperspective for stakeholder theory advancement (2). Conse-quently, from our understanding of ANT concepts, wepropose an approach to stakeholder analysis of and engage-ment in projects (3). We then briefly describe our researchmethodology on the basis of a longitudinal case study of anInformation System project, called “Pupitre Virtuel”, to whichwe apply our method to read and understand the evolution ofthe project. (4). Thereafter, we report our conceptual method'sfindings (5). Finally, we discuss and comment on theimplications of our research (6).

2. Theoretical background

2.1. From a stakeholder relational perspective

Initially, most stakeholder literature concentrated on thedyadic relationships between individual stakeholders anda focal organisation (Freeman, 1984), considering theorganisational interactions with stakeholders as independentrelationships. One of the most significant contributions tothis dyadic perspective is by the stakeholder identificationand salience framework of Mitchell et al. (1997). The stake-holder salience framework allows classifying stakeholdersaccording to their power, legitimacy and their claim'surgency. According to Rowley (1997), this dyadic perspectiveis appropriate for classifying different types of stakeholders.However, this analysis is limited to explaining how organisa-tions react to stakeholder influence because they do notrespond to each stakeholder individually; instead, theyrespond to the “interaction of multiple influences from theentire stakeholder set” (Rowley, 1997, p. 890). Accordingly,the analysis should integrate the complex arrangement ofmultiple and interdependent relationships in stakeholderenvironments. Which is why the stakeholder theory movedfrom a dyadic perspective (Freeman, 1984) to a relationalperspective based on a network perspective (Pouloudi andWhitley, 1997; Rowley, 1997).

Applying the Social Network Theory (Granovetter, 1973) tostakeholder management, Rowley (1997) examined anddescribed how aspects of an organisation's stakeholdernetwork, namely the network density and focal organisation'scentrality, affect the organisation's response to stakeholderdemands. Density is a characteristic of the entire network. Itmeasures the relative number of ties in a network that linkactors. As the network density increases and the number of tiesbetween the network members grows, communication acrossthe network becomes more efficient, facilitating the voluntarydiffusion of norms, values, and shared expectations. Concur-rently, centrality refers to an individual actor's position in thenetwork relative to others. From this analysis, Rowley (1997)argued that a firm can resist stakeholder pressures better if it is acentral actor in its stakeholder networks and if these are less

1110 S. Missonier, S. Loufrani-Fedida / International Journal of Project Management 32 (2014) 1108–1122

densely interconnected. Given the behaviours related toresisting stakeholder pressures, the author identified fourtypes of firms: a commander, a compromiser, a subordinate,and a solitarian.

On the basis of Rowley's (1997) research, numerousscholars have acknowledged that stakeholder relationships donot occur in a vacuum of dyadic ties, but in a complex networkof intertwining relationships. Various scholars used suchanalysis, for instance, in the context of IS projects (e.g.,Coakes and Elliman, 1999; Walsham, 1993) or in construction(e.g., Bourne and Walker, 2005; Cova and Salle, 2006). Ratherthan focusing on stakeholder attributes, the network perspectiveviews stakeholder characteristics and behaviours as arisingfrom the social structural environment. It also emphasises theinteractions between stakeholders for a better understanding ofthe decision-making process.

2.2. To stakeholder relational ontology

Despite the richness and advances of these previous studieson the stakeholder relational perspective, we argue that thisapproach should be adapted to integrate the dynamic andemergent nature of stakeholder networks.

Research anchored in a stakeholder relational perspectivefocuses only on the “resultant” effects of relationships,ignoring the “emergent” effects of the stakeholder networkover time. This research considers pre-existing actors whomeet in a network and establishes the relationships betweenthem. For instance, in their systematic literature review ofstakeholder methods, Pacheco and Garcia (2012) show thatsuch studies suggest that once we have an idea of who themain stakeholders are, the basic interactions between themshould be identified. Consequently, how stakeholders evolvethroughout a project due to the interactions between them andthe project over time, is understudied. Although Rowley(1997, p. 890) underlined the “interdependent relationship instakeholder environments”, the effects of this interdependencyon the stakeholders and the project over time remainunderstudied (Yang et al., 2009a,b). Rowley (1997) tookstakeholder and project characteristics for granted by presum-ing that networks' a priori existence was a manifestation ofstakeholder dynamics (Papadopoulos and Merali, 2008). Oneof the main reasons for this was that this research tends toignore time in stakeholder analysis and engagement, reducingit to a log effect. Consequently, research on a stakeholderrelational perspective considers the network, its elements andcharacteristics as stable and innate properties, thusoverlooking that the stakeholder and the project propertiesco-evolve over time.

We propose a relevant approach that integrates dynamic andemergence mechanisms in order to thoroughly improve stake-holder analysis of and engagement in projects. We thus base ouranalysis on the Actor–Network Theory (ANT) (Callon, 1986;Latour, 2005). ANT views a network as a process that shapes andreshapes relationships. This thinking invites us to think aboutindividuals, organisations, and entities in terms of “ceaselesschange, emergence and self-transformation” (Nayak and Chia,

2011, p. 282). It implies a switch from a relational perspective(“weak relationality”) of stakeholder analysis and engagementto a “strong relational ontology” (Slife, 2004). Slife (2004)thus regards everything, including each person, as first andforemost a nexus of relationships. Instead of recognising astakeholder network and “organisation” as stabilised entities,we believe that people's actions are always locally defined andemergent, and that this local emergence includes the materialas well as the social structures and processes (Orlikowski andScott, 2008). This perspective emphasises the understandingof project management through unpredictable and emergencyproject features and integrating human and non-human(materiality, artefacts) in the analysis. As noted by Pouloudiet al. (2004), the application of stakeholder analysis has beenpredominantly restricted to human stakeholders. Vidgen andMcMaster (1996, p. 225) boldly defined stakeholders as any“human or non-human organisation unit that can affect as wellas be affected by a human or non-human organisation unit'spolicy or policies”. Anchored in this relational ontology, wesuggest the transposition of this mode of thinking to thestakeholder theory in project management by mobilisingANT.

3. ANT and its application to stakeholder theory

3.1. ANT and stakeholders

The basic idea of ANT (Akrich et al., 2002a,b; Callon, 1986;Latour, 2005) is that to achieve a goal, a network or“assemblage”, requires a faithful alliance between humansand non-humans and, thus, the social and the technical. Thistheory seems to fit stakeholder management perfectly.

ANT has been used in many different and sometimescontradictory ways, and in a variety of disciplines. Organisa-tion Studies and Information Systems are its two mainapplication areas. ANT is increasingly criticised and consid-ered (Czarniawska and Hernes, 2005; Dery et al., 2013; Faikand Walsham, 2012; Walsham, 1997). Although ANT hasbeen largely used to study projects especially IS projects(Dery et al., 2013; Elbanna, 2010; Meier and Missonier, 2012;Ramiller and Wagner, 2009; Wagner et al., 2010), few studies(Blackburn, 2002; Pollack et al., 2013) have mobilised thistheory in the field of project management and, a few authorshave underlined its usefulness for stakeholder management(Luoma-aho and Paloviita, 2010; Pouloudi et al., 2004;Vidgen and McMaster, 1996). We assume that ANT wouldbe particularly useful in this area for the four followingreasons. Firstly, it provides a framework for conceptualising aproject as an emerging network (associations, relations) thatextends and transforms over time. This network is social, butalso includes non-human actors. Indeed, one specific aspect ofthis approach is anchored in the principle of “symmetry”which asserts that the concept of “actor” should also beextended to non-human actors. A non-human can be anyobject or more precisely, as Latour (2005) explains, a series ofheterogeneous inanimate actors called “actants”. It could be atechnology, its components, a contract, or even a knowledge.


Indeed, intentions cannot be traced to a purely human origin,and they are also effects of a network. For instance, Callonand Law (1995) showed how the manager of a company isable to see, decide, and act because a network of actants suchas reports, telephones, the internet, paperwork, desks, oradministrators create the capacity for choice and intentions.The possibility to act is not what an actor does, but “whatactants provide with their actions, with their subjectivity, withtheir intentionality, and with their morality” (Latour, 1999, p.18). Secondly, ANT proponents refuse to pre-empt the actoridentity, regardless of their relationships or effects within anetwork (Latour, 1988). Thirdly, from this perspective, aproject's success is considered to be dependent on the activeparticipation of those who are determined to advance and,thus, form a convergent network (Callon, 1991). Fourthly,mainstream stakeholder analysis usually only operates at aproject's front-end as a basis for on-going decisions andplanning (Jepsen and Eskerod, 2009), whereas ANT offers thepotential for a dynamic and thorough examination of theon-going, influential interactions between actors about theproject.

Contrary to stakeholder studies anchored in a relationalperspective, an actor network does not mean that pre-existingactors come together in a network and establish relationships.Entities stand in relation to all other entities that constitutethem. The network (or the “collective”) is not a collection ofpre-existing things, but a property emerging from relationships(Callon and Law, 1995). As Callon and Law (1995, p. 486)explained, relationships are prior to essences, or are “things inthemselves”.

Table 1Conceptual approach.

Stakeholder analysis

Stages Contents

First step:Morphological stakeholdernetwork analysis(front-end)

(1) Identify stakeholdersand analyse stakeholderrelationships

- Poles (status and roles)- Intermediaries (what iswill produce and put inby actors): nature, diveand frequency

(2) Identify stakeholderinterests

- Interessement (identifyiinterests of members ofthe project and devices

(3) Assess stakeholderinfluence

- Identifying degree of coof the network (degreeof interests and goals, dof coordination)

Second step:Dynamic stakeholderanalysis(over the project)

(4) Identify controversies - Nature, i.e. the subjectcontroversy and its stak

- Actants involved- Stabilisation whether acompromise seems to hbeen reached or not

- Redefinition of the techobject

(5) Analyse effects ofcontroversies onstakeholder network

Effects on the network

3.2. Proposition of a relevant conceptual approach basedon ANT

In response to our research question, we need a relevantapproach with which project managers can analyse stakeholderengagement, considering the dynamic and emergent nature ofstakeholder networks. On the basis of ANT, we propose aconceptual approach divided into two interrelated steps andwhich considers temporally evolving phenomena. The first stepaims to analyse the morphology of the stakeholder network(the project) and allows the network's degree of convergence tobe qualified (Section 3.2.1). The second step aims to follow andvisualise the network evolution over time (Section 3.2.2). Wealso suggest that the “process of translation” (Callon, 1986),which is divided into four stages labelled problematisation,interessement, enrolment, and mobilisation, can be used toguide project managers' stakeholder engagement strategy.This process can be understood as a type of consensus-seeking process, a multifaceted interaction in which oneentity assigns others a role. In Table 1, we offer an overviewof our conceptual approach. Subsequently, we detail thedifferent stages and their contents. Later in the text, wereturn to the stages of stakeholder analysis by the numbers1 to 5, and to the stages of stakeholder engagement by usingA–B–C.

3.2.1. Morphological stakeholder network analysis (front-end)

(1). Identify stakeholders and analyse stakeholder relation-ships. Anchored in a relational ontology, the identification of

Stakeholder engagement

Stages Contents

produced,to circulationrsity, amount,

(A) Problematisation Framing the problems, identifyingother relevant actors, andhighlighting how the problemaffects the other actors

ng factorspoles in)

(B) Interessement andenrolment

Ability of an actor to arouse theinterest of others for his own projectAssign a role to each pole

nvergenceof alignmentegree

(C) Mobilisation Stabilisation of the stakeholders

of thees

ave

nical

If necessary reengagethe process of translation(A–B–C)

Problematisation–interessement–enrolment–mobilisation


a project's heterogeneous stakeholders, their roles and theirdefinitions cannot be separated from their relationships.They take the form of different subsets called “poles”(Callon, 1991, 1992). Consequently, like traditional methodsof stakeholder identification that use a list of specific stake-holders in a given context, firstly, we use Callon's (1992)list of “poles”: “Science”, “Technology”, “Market”, etc.Then, to deeper observe their relationships, Callon (1991)recommended observing the “intermediaries”, i.e. “anythingpassing between actors which defines the relationshipsbetween them” (Callon, 1991, p. 134). He defines actorsaccording to their interactions with others, through theintermediaries that they put into circulation. The nature ofthe intermediaries that the stakeholders produced or willhave to produce during the project informs their identity andhelps identify the “poles”. Intermediaries may be material(documents, contracts, technical artefacts, money, technicaldrawings, and schedules), or intangible (skills, knowledge,informal exchanges). The diversity of intermediaries con-firms the existence of several coordination methods (Callon,1992), and the amount and frequency of movement ofintermediaries reflect the intensity of interactions. Moreintermediaries mean more and varied interactions and,therefore, strong links.

(A). Problematisation. With problematisation, the focalactor (e.g., the project manager or a firm) frames theproblem to be solved in the project in its own terms,identifies other relevant actors, and highlights the problem'sinfluence on the other actors. The focal actor outlines broadstrategies for addressing the problem at hand. In thisprocess, as underlined by Luoma-aho and Paloviita (2010),firms mark their terrain by mapping the issue of interest andthe relevant actors.

(2). Identify stakeholder interests. The notion of interessementallows for identifying and understanding stakeholder interestsin a project. The analysis is here similar to traditionalstakeholder analysis which recommends identifying stake-holders' area of interest (their stake). This identification islinked to interessement as a strategy of engagement, i.e. apersuasive process that makes an identified actor interested in aproject implies this actor's transformation into an ally (Akrichet al., 2002b).

(B). Interessement and enrolment. Interessement focuseson the persuasive processes, which, in stakeholder terms, areaddressed as stakeholder management or stakeholder align-ment. According to Callon (1986), when an actor agrees toplay the role it has been assigned, this implies a successfulinteressement, which suggests that the “enrolment” can takeplace. Enrolment includes actors accepting the roles definedfor them and explained during the previous phases.The enrolment means assigning each representative stake-holder group (pole) a role. An actor's ability is mainlyresponsible for arousing the interest of others in his project.Akrich et al. (2002b) explained that to enrol another actor,an actor has access to devices that can take variousnon-rhetorical (seduction, coercion, solicitation) or rhetorical

(texts, conversations, etc.) forms. The interessement cantherefore realise the network of considered alliances. Thus, anetwork of actors whose interests are aligned is a convergednetwork (Callon, 1991).

(3). Assess stakeholder influence. Based on the studies byCallon (1991), we examine the structure of a project'sstakeholder network and its influence on the project manager'sresponse strategies. Callon (1991) believes that a network candevelop in two directions: convergence to or divergence fromits stakeholders.

Convergence measures the extent to which the process oftranslation and its circulation of intermediaries lead to anagreement (Callon, 1991). In a network, convergence doesnot mean that each element acts or becomes the same, butthat the actor activities fit together despite their heterogene-ity. As a whole, the network should be able to concentrate itsefforts on a single point. Conversely, in a divergent project(a weakly convergent project), actors find that their status isconstantly questioned and that mobilising other networkparts is difficult. Actors resist the role that the networkassigns them and respond unpredictably to instruction.Consequently, key actors may first pursue their interestselsewhere and, if they grow too divergent, becomevulnerable to collapse (Alderman and Ivory, 2011). Toobserve the trajectories of convergence or divergence, Callon(1991) proposed two dimensions: the degree of alignmentand the degree of coordination. (1) The degree of alignmentof interests and goals concerns individuals' motivations,which align their interests and goals. This occurs during thetranslation process by means of interessement's active work.A network is aligned when each stakeholder's respectiveinterests are aligned with the project's global interest(Callon, 1991). For instance, Callon (1991) explained thatif a network includes three actors “X–Y–Z”, and the relationbetween X and Z necessarily implies Y, then the network'sdegree of alignment is strong. Intermediaries are passedbetween the actors to assure a certain degree of convergenceamong them. (2) The degree of coordination concernsdifferent forms of coordination (hierarchy, trust, knowledgesharing, and contract) in the network. Callon (1992) referredto weak coordination when a network has no specific localrules, and to strong coordination when local and generalrules shape a network.

(C). Mobilisation. During mobilisation, the initiator orfocal actor uses methods to ensure that allies act according totheir agreement and do not betray the initiator's interests. Withthe stakeholders and, thus, the allies mobilised, an actornetwork achieves stability. This stability means that the actornetwork and its underlying ideas have become institutionalisedand they are no longer seen as controversial (Mähring et al.,2004).

3.2.2. Dynamic stakeholder analysis (over the project)

(4). Identify controversies. In an on-going project with time,cost, and quality pressures, it is essential that the project


actors are not only aware of, but also know preciselywhere and when to identify and observe these dimensions.Here, in terms of ANT, the concept of “controversy”appears to be very helpful. A controversy can be anything(argument, idea, ideology, etc.) that challenges thenetwork's status quo and thus affects and changes thestakeholder interactions (Latour, 2005). A controversyemerges when issues that were taken for granted arequestioned and discussed. This thus implies requiring verystrong interactions to redefine the relationship. Conse-quently, researchers and project actors have to examine thebasics of each controversy (its components) in a projectnetwork because, firstly, controversies reveal interactions,which emphasise that the relationships between humansand non-humans are never fixed. Secondly, the effects ofcontroversies in the network reveal new definitions and,thus, the evolution of the trajectories. The aim is toobserve controversies that occur within a network (oroutside) and their effects. We explain the details of theidentification of controversies in our methodology (seeSection 4.3).

(5). Analyse effects of controversies on stakeholdernetwork. At the end of each controversy, the researcher orproject manager should again focus on the intermediaries,count them, and identify their nature to determine their effectson the stakeholder network and on the degree of convergence.Consequently, the network depends on the changes occurringafter the controversy, but the focal actor sometimes needsto re-engage the process of translation (problematisation,interessement, enrolment, and mobilisation) to reinforce orbuild another stakeholder network. Finally, this approach issensitive to the project's time.

4. Methodology

4.1. Research design

To illustrate our conceptual approach based on ANT, ourfindings were based on a real-time qualitative longitudinalstudy. This illustrative case study comprised an InformationSystem (IS) project. The main objective of the project wasto generalise a technology (called the Pupitre Virtuel) in allFrench schools. Eight people were members of the projectteam. A longitudinal study was needed to observe theprocess of the stakeholder network, the dynamics andemergence of its assemblages, and its evolution over time.The case was selected because it was exemplary, i.e. itsconditions allowed us to apply our framework (Yin, 2008).Indeed, firstly, Pupitre Virtuel was a rich opportunity tounderstand the evolution of a project because it featuredvarious stakeholders (project managers, a team developer,technology, users, national institution, and private partners).Secondly, this project also provided us with interestingopportunities to live the (everyday) events as well as toidentify and understand the interactions between stake-holders (actors and objects) belonging to different spheres

and, thus, in real-time. This project started in March 2003and finished in December 2005. One of the authors hadfirst-hand knowledge of the project because s/he hadparticipated in it from the beginning to the end. As aninternal expert, s/he evaluated specific aspects of the projectand received regular project updates. Despite her/hisinvolvement in the case, the researcher undertook partici-pant observation, utilising the case study approach. Asparticipant observers, s/he conducted the study withoutinforming the project actors of the research goals. One ofthe authors was present throughout the project as a“moderate” participant observer, a role which enables abalance between involvement and detachment: the projectmembers were told that s/he was researching how usersappropriated the new system and s/he did not act orintervene on any aspect of the project. By regarding theproject as an actor network, our analysis considers anindividual level involving associations and practices —both of which the project formed — between humans andnon-humans within the project.

4.2. Data collection

We used four sources of data: observation, interviews, dataarchival, and e-mails. These mixed methods ensure therichness of the findings and the purpose of triangulation(Yin, 2008), as well as allowing us an in-depth examination ofthe process (Langley et al., 2013). (1) We benefited from242 days of participatory observation of the various projectactors in the different project groups. We transcribed ourobservation in a daily journal. (2) We conducted 35 one-on-one interviews (26 open interviews and 9 semi-structuredinterviews), each lasting an average of 1 h, with actors indifferent functions and positions within the project. Weconducted interviews during three periods: the exploratoryperiod — open interviews, the in-depth period, and thecontrol period. We summarised each interview on an indexcard after recording and transcribing it in detail on acomputer. In total, 38 h of interviews were recorded, and336 pages were transcribed. (3) We collected 85 public andofficial documents (the partnership contract, presentations ofthe technology and the project), internal and privatedocuments (e.g., the seven versions of the new partnershipcontract, legal documents, etc.). The documents wereannotated, and systematically listed under the themes theyaddressed. (4) We had the opportunity to gather more than110 e-mails shared between the actors of the two firms duringthe entire observation period. We had access to the project'sintranet, and one of the main stakeholder of the project (ERI'smanager) agreed to transfer all mails he received from otherstakeholders to us and those he sent to all the project actors.Finally, a report of approximately fifty pages describing thecase study was submitted to the key actors for theiragreement, which validated our interpretations, and thusincreased the construct validity and the internal researchvalidity (Yin, 2008).


4.3. Data analysis

Since time is central to our analysis, and as recommendedby Langley et al. (2013), we combined different methods toexamine the project evolution (the network) and, thus, thestakeholder analysis and engagement in depth. Real-timeobservation was useful to observe how these processesunfold over time. Interviews and data collections areparticularly suitable for tracing event chronologies andmeaning over long project periods. The data analysis wasbased on a chronological database (Van de Ven and Poole,1995), which seemed the most appropriate approach fordescribing and understanding the project from start to finish.A chronological database aims to highlight the course of astudied phenomenon over time. Based on our conceptualapproach, all the data were coded: the daily journal,interviews, documents, and e-mails. To ensure the stabilityand reliability of our codes, we used the data analysissoftware ATLAS/Ti and, at different intervals, coded the dataseveral times following the guidelines by Miles andHuberman (1994).

To identify the nature of the network link, as recom-mended by Callon (1992), we made an inventory of theintermediaries exchanged between each project group byqualifying and quantifying these. We then developed a

Fig. 1. Visualisation of front-

matrix code to “map” the actors. To identify the intensity ofthe network link, we constructed an adjacency matrix toquantify the ties (symmetrical matrix). The accumulation ofthe intermediaries between the actors was identified usingthe multiplicity network, which acted as a measure ofintensity or frequency of interaction and thus of networkties. Callon (1991) advocated a method of counting thenumber of intermediaries circulating among the actors andgroups of actors to calculate the degree of alignmentnumerically. This operation was repeated during the analysisafter all network controversies had been identified. It wasthus possible to provide a visual map of the network'semergence and evolution that shows all the stakeholdersconstituting the network, their intermediaries and links.

We relied on the principles of Latour (2005) andrecommendations by Venturini (2010) to identify controver-sies. We created a summary table for each controversy thatincludes the following markers: (1) its nature, i.e. the subjectof the controversy and its stakes; (2) the actants (humans andnon-humans) involved; (3) its stabilisation, mentioningwhether a compromise was reached; (4) a redefinition ofthe technical object; and (5) its effects on the network withinthe convergence dimensions. The aim was to observecontroversies when they occur within the network (oroutside) and their effects.

end stakeholder network.


5. Case study findings

5.1. Morphological analysis: an initial convergentstakeholder network

In keeping with our conceptual approach (see Table 1), ourmorphological analysis allowed us to identify and classify themain stakeholders into seven “poles” and their relationships atthe front-end of the project (see Appendix A).

(1). Identify stakeholders and analyse stakeholder relationshipsand (A) problematisation

A French region (which we call Alpha) and a team ofuniversity researchers and designers initially invented anddeveloped Pupitre Virtuel in 1999, which the local government(LG) supported financially. Concretely, this technology, anENT (Espace Numérique de Travail — numerical space ofwork) called the “Pupitre Virtuel”, should make information,resources, and services universally available in real-timeto students using a login and a password. This local projectwas part of a broad national project that the French Ministryof National Education (MEN) devised to develop and establishan ENT in all French schools. In addition, the MEN, aninfluential stakeholder in the project, believed that this PupitreVirtuel was a very innovative technology. Since 2000, thisinnovation has been tested by 11, pilot-user schools in theAlpha region.

The project manager (the LG), who aimed to generalisePupitre Virtuel for most French learners, conductedproblematisation as a strategy of engagement. In 2003, theproject team (the LG and some university researchers) wantedto generalise its use across the country. The project teamtherefore needed to improve the software to ensure that a largenumber of students could use the programme at the same time.To do so, the LG identified the actors who needed togeneralise Pupitre Virtuel (e.g., private actors with expertiseand funding, the Ministry, the researchers and their universi-ties, the pilot users and the experimental technology) and gotthem interested in the project. The LG subsequently decided toresort to private funding for the generalisation. In October2003, a private and (very) small company (which we call ERI)met the requirement and the LG chose it to integrate theproject.

(2). Identify stakeholder interests and (B) interessement andenrolment

The actor interessement in the project occurred over twolong months of on-going negotiations between the identifiedactors. After several compromises by each actor, a partnershipcontract was signed in February 2004. At this time, thepartners aligned their respective interests and goals for theirmutual collaboration. The inventor of Pupitre Virtuel and histeam (university researchers and designers) wanted to spread itand offer a structure for deployment throughout the country.This deployment could not occur without a committed partner.This partner was ERI, a company recognised for innovativetechnology projects and for providing funds. The profitability

of ERI's investment in the project depended on the technologythat the team's researchers had invented and on theirknowledge. The technology, Pupitre Virtuel, required re-searchers' technical competencies and ERI's market compe-tencies to be generalised. The experiments with PupitreVirtuel technology in certain schools were included in theproject to test the technical evolution of Pupitre Virtuel.Lastly, for the LG (project manager), the project's successwould lead to the Alpha region receiving high visibility as aprecursor of an innovative national education project. Thissuccess would depend on the technical qualities (provided bythe designers) and commercial qualities (provided by ERI).The interessement was successful and each of the entities'roles in the newly created actor network were defined andcoordinated (enrolment). ERI therefore had an exclusiveexploitation licence for Pupitre Virtuel while at the sametime, it was responsible for the commercialisation andtechnical developments of Pupitre Virtuel. ERI's managerhad very strong assumptions about Pupitre Virtuel technology,which pilot users had tested for four years and which theFrench Ministry recognised.

(3). Assess stakeholder influence and (C) mobilisationAccording to the definition of the density given by Rowley

(1997), we conclude that this project's stakeholder network isdense: the number of ties that link actors in this network ishigh. Indeed, at this time, ERI had to order technicaldevelopments from Pupitre Virtuel inventor and designerswho now worked in a new company, Centile engineering, adata-processing company. The inventor of Pupitre Virtueldirected this company. The LG was responsible for the projectin his region. The university continued to work with thedesigners and used Pupitre Virtuel. The French EducationalMinistry published recommendations for its technical andcommercial development. The developers improved thetechnology and the region's users tested it. Therefore, atthis step of the project in February 2004, we observed aconvergent network. The actors' interests were aligned; eachentity had a strong interest in attempting the project.Furthermore, the roles and tasks were clearly assigned andwell defined in the partnership contract. Each actor'scompetencies complemented and were essential to the projectgoal. Fig. 1 shows the visualisation of the front-endstakeholder network Fig. 1.

At that time, the project appeared ideal as all thestakeholders were very involved in and trusted the project.Additionally, no stakeholder appears to have a more centralposition in the network relative to others. Nevertheless, themobilisation stage became very complicated.

5.2. Dynamic analysis: from convergence to collapse

(4). Identify controversies and analyse their effects onstakeholder network

During the Pupitre Virtuel project, we identified fivecontroversies.


First controversy: coordination of project stakeholders. InMarch 2004, the first controversy appeared. This controversyconcerned the project's internal stakeholders and involved thetechnical developments' specificity, which required thePupitre Virtuel components to be generalised, and thesedevelopment costs. ERI's manager requested more detailsabout the transactions between the two firms. He felt thatCentile's invoices for Pupitre Virtuel developments werenot justified. The technical development costs appeared to behigh.

Centile's manager felt that the Pupitre Virtuel contentdid not allow for detailed development specifications andthe components required specific developments. OnlyCentile interacted directly with the technology and notifiedERI about its assumptions regarding the technology, whichwas supposed to allow effortless improvements. PupitreVirtuel revealed the complexity of its components' merger.The stakeholders thus compromised as they had tocollaborate to generalise the technology. A mail exchangebetween the two managers expressed their willingness towork together:

“The difficulties we encounter are from my perspectivebasically due to a lack of common practices and we musteach make the necessary effort to iron out these differencesin opinion and work methods. We will soon forget our initial‘hiccups’ and the only thing that matters is the will to worktogether and to succeed in our projects to the best of ourrespective interests”.

[ERI's manager (May 2004)]

“We would like to thank M.B. [ERI director] for hisinitiative and to reaffirm our total support for the aim tosucceed, which he points out, as well as our confidence inthe prospects. The strong ties that unite our two companiesshould not make us forget the need to formalise ourrelations, and the perspective in M.B.'s email is a majorand encouraging contribution to the strengthening of theserelations”.

[Centile's manager (May 2004)]

Effects on the network: the compromise strengthened theconvergence, even after the controversy had redefined onepart of the relationship between ERI, Centile and PupitreVirtuel. However, the alignment of the resources wasweakening from the initial complementarity roles of ERI,Centile and Pupitre Virtuel to ERI's dependency on Centileand Pupitre Virtuel. However, their strong vision for theproject and the need to reinforce the convergence allowedthem to compromise.

Second controversy: questioning the quality and reliability ofthe Pupitre Virtuel components. In July 2004, a newcontroversy emerged between ERI, Centile and Pupitre

Virtuel. The quality and reliability of the Pupitre Virtuelcomponents were questioned and their improvement sug-gested. The project actors were also informed that they hadlost two new Pupitre Virtuel project implementations inother schools, which had been necessary for the project'svisibility and generalisation due to doubts about PupitreVirtuel's quality and reliability. Interactions between PupitreVirtuel and the external actors had led to questions about itsreliability. Moreover, the technologies chosen for the newprojects had not yet been tested. These choices called thereasons for ERI's interest in the project into question:Pupitre Virtuel's reputation, experience and quality. Thetechnology would not be able to keep its promises. ERIwanted to build new Pupitre Virtuel foundations:

“While the main advantage of Pupitre Virtuel is its threeyears' experimentation and our technical know-how, theyare not convinced of the scalability capacity of our solution.Our growing business skills with Pupitre Virtuel, improvedover three years of experimentation, are useless”.

[ERI's manager, meeting (June 2004)]

Centile's manager was opposed to this decision:

“Pupitre Virtuel components do not need to be called intoquestion. They have proven their reliability and arerecognised by a strong and large community of developers.Actors involved in invitations to tender know nothing”.

[Centile's manager, meeting (June 2004)]

Effects on the network: the convergence of the threestakeholders — ERI, Centile, and the Pupitre Virtuel — wasweakening. ERI's manager wanted more transparency regard-ing the technology's technical functioning. He wanted to openthe Pupitre Virtuel “black box”. Consequently, ERI askedCentile for the technology's code source. The firm thenrecruited a new computer developer to work exclusively forit. At the same time, ERI decided that Centile was no longer acredible technical spokesperson for Pupitre Virtuel. ERI thendecided that its new computer developer would be the technicalspokesperson for the technology.

Third controversy: assignment of new responsibilities. InNovember 2004, Pupitre Virtuel encountered technical diffi-culties when learners could not access Pupitre Virtuel'sservices. The LG was worried and wanted a quick solution:

“Everything is going wrong! Many users cannot access theirPupitre Virtuel. Everyone is complaining!”

[Member of the LG (5 December 2004)]

ERI and Centile blamed each other for the technicalquestions. During this controversy, the technology Pupitre


Virtuel was looking for a credible spokesperson. Theinvolved actors could not compromise. Faced by thetechnical status quo and the users' distress, the projectmanager entrusted the exploitation of the technology to anexternal company.

Effects on the network: this third controversy weakened theconvergence and strengthened the divergence. It revealed thedissolution of the project network's initial organisationalstructure.

Fourth controversy: questioning of the technology components.In December 2004, the French Ministry (MEN), an influentialproject stakeholder, delivered a formal report that revealedthe national recommendations and norms that had to befollowed in such projects. The French Ministry argued thatthe Pupitre Virtuel components were not sufficiently reliableand sustainable. Centiles' manager believed that the com-ponents were reliable and did not require modification.Conversely, ERI's manager believed that Pupitre Virtuel'sfuture was unthinkable without Ministry support. He wantedto rebuild Pupitre Virtuel to meet the Ministry recommenda-tions. Centile was opposed to this decision, arguing that thereinforcement of the components would be sufficient. Thecontroversy concerning the virtual desk's component reli-ability was essentially redefined to consider the componentsand their reliability.

Effects on the network: this controversy reinforced theweakening of the convergence and the divergence of thenetwork. ERI searched for another stakeholder, a new partneroutside the Pupitre Virtuel network, to help develop a newPupitre Virtuel platform, while Centile worked on the initialPupitre Virtuel. The entities in the network shared the samecommon goal (the generalisation of the Pupitre Virtuel),although their visions differed.

Fifth controversy: questioning of the partnership contract.When ERI received the Pupitre Virtuel source code inDecember 2004, the ERI's computer developer realised thatthe software technology included open source software,covered by a general public licence, and a private licencesoftware. Alerted to this discovery, ERI was worried aboutthe partnership contract's legitimacy. A controversy aroseamong all the project members about the nature of thelicences covering the Pupitre Virtuel components.

According to ERI, the presence of free software in thetechnology called for the reconsideration of the partnershipcontract; it casts doubt on the validity of the exclusiveexploitation licence. Conversely, for the LG, the universityand Centile, the partnership contract was still relevant. Eachcompany requested technical and legal experts to detail thecontents of Pupitre Virtuel and to assess the lawfulness ofthe contract. Pupitre Virtuel, which was essentially definedby the nature of its components and their development, waslooking for a reliable spokesperson.

In January 2005, the respective experts' reports disagreed onthe validity of the exclusive licence exploitation and reinforced

the controversy. Therefore, the project members decided tonegotiate a redefinition of the relevant clauses in theirpartnership contract. Fundamentally, ERI's manager was notagainst the use of open source licences, as they could be acompetitive advantage against property licences in the publicsector. The aim was to find a compensation for PupitreVirtuel's exclusive licence exploitation and to redefine themodalities of their collaboration.

“I don't want the partnership contract to be broken; in thenational context, we have no interest in revealing thesetensions about our partnership. The main issue is that wehave to reach compromises without breaking the partner-ship to pursue the project”.

[ERI's manager (March 2005)]

However, despite seven new versions of the contract, thestakeholders could not reach an agreement. They decided toresolve the issue through a new actor in the network: a court.Finally, in December 2005, after two and half years, the projecthad only one remaining major supporter: ERI. Consequently,the LG and the University of the Alpha region cancelled thePupitre Virtuel project.

Effects on the network: this latter controversy stressed thecollapse of the Pupitre Virtuel network. The stakeholders wereplaced in a “controversial situation” that was irreversible,because they were unable to compromise.

The longitudinal follow-up of the five controversies revealsthe trajectory of the Pupitre Virtuel stakeholder network along aconvergent/divergent continuum. This type of observationallows the visualisation of its trajectory from its convergenceto its progressive divergence and finally to its collapse (cf.Fig. 2).

To conclude our case study findings, we note that at thefront-end, the Pupitre Virtuel network was convergent becausethrough the process of translation, the key stakeholders and theirinterests were well identified and aligned, the resources werecomplementary, and the coordination was strong. Nevertheless,along the string of the five controversies, the network facedweakening convergence, which led to extreme divergence.Consequently, the continued weakening of the alignments ofinterests, resources and coordination led to the network'scollapse. More precisely, we found that two main stakeholders(the private actor ERI and the Pupitre Virtuel technology)resisted the role that the network had assigned them andresponded unpredictably, even though they had accepted theirroles at the front-end. Thus, although the interests had beenaligned at the beginning, these key stakeholders later began topursue interests elsewhere. When it grew too divergent, thePupitre Virtuel network became vulnerable to collapse. Thecollapse occurred when the important actors ceased to bemobilised by the network's intermediaries, that is, when they nolonger believed that the project served their interests and theywithdrew.

Fig. 2. Visualisation of stakeholder network trajectory over the project.


6. Discussion, managerial implications and further research

6.1. Theoretical contributions

The primary purpose of this study was to contribute tostakeholder research on project management by proposing arelevant approach based on ANT to provide a dynamic andemergent stakeholder analysis and engagement throughout aproject. Consequently, the main theoretical contribution shedslight on the relevance of switching from a stakeholder relationalperspective to a stakeholder relational ontology aimed ateffectively improving the stakeholder analysis of and engage-ment in projects. Despite the richness of Rowley's works (1997),his relational approach, anchored in the Social Network Theory,is “moment-to-moment” and misses the flow of interactions withdynamic and emergent dimensions. Consequently, the stakehold-er decision-making is viewed as the result of the networkstructure (its density and centrality) at a specific moment in theproject. Instead, the conceptual approach that we proposedconsiders the effects of the evolution of the network andstakeholders' influences on the stakeholders, on the networkand on the project. We are thus equipped to analyse, follow andunderstand the network and its members' influences. As such,decision-making and response to this influence can be betteradapted. For instance, our case study reveals that the interactionsbetween the Pupitre Virtuel network stakeholders and theirevolution had consequences for the network. More specifically,each entity as well as its place, role and boundaries wereredefined and transformed inside and outside the network andacross the project. These interactions changed the alignment ofinterests, resources and coordination, and partially weakened theconvergence in the Pupitre Virtuel project network. Therefore,the project also developed due to the connections between theheterogeneous stakeholders.

Our findings moreover indicate that, at the beginning ofthe project, the stakeholder network's density was high. Forinstance, ERI's centrality was relatively high. Thus, thestudy by Rowley (1997) improved our understanding ofERI's response after the first controversy, suggesting that itadopted a “compromiser” role, because it attempted todiscuss and negotiate with its stakeholders (Centile andPupitre Virtuel). Nevertheless, our study revealed that thisrole changed throughout the project. This evolutionemerged from the evolution of the interactions betweenthe stakeholders and the project. After the fourth

controversy, ERI decided to adopt a “solitarian” role bydeveloping a new platform and searching for a newstakeholder outside the network. This was not only anetwork configuration response at a moment. Instead, itemerged from relations and events that ERI had experi-enced in the network from the project's start. From theANT view, anything that exists emerges from relations withsomething else (Latour, 2005). If the fourth controversyhad emerged early in the network, ERI's response mighthave been different.

Hence, our mode of thinking resolves a problem thatJepsen and Eskerod (2009), and Pouloudi and Whitley (1997)underline: the false assumption that a coalition of stakeholdersis stable across the course of a project. In particular, Pouloudiand Whitley (1997) revealed that stakeholders might interactand influence each other. By considering relational ontology,we integrated the coalition of stakeholders' inadequatedelimitation, which changes over a project's duration in sofar as different stakeholders may be important at the differentprocess stages, into our stakeholder analysis. We thus sharethe conclusions by Luoma-aho and Paloviita (2010), whoexplain that categorisation (such as stakeholders and non-stakeholders) risks have excessively strict boundaries betweena subject and an object, and do not adequately describevarious networks.

6.2. Implications for practice

The empirical evidence of the study has value in itself,since empirical investigations in the field are still ratherscarce, of which a comprehensive review of existing projectstakeholder analysis supports (Aaltonen, 2011). The mainpractical contribution lies in demonstrating the usefulnessof our conceptual approach as a useful method for proj-ect managers that informs them about what to observe ina stakeholder network in a project and how and when to doso.

• What to observe concerns firstly the associations betweenactors and non-humans in the project. ANT provides aunique approach to integrate and define previously ignorednon-human entities and their influence. We also suggest thatstakeholder analysis (divided into 5 stages: relationships,interests, influences, controversies, and effects) is linked tostakeholder engagement where the “process of translation”

image of Fig.�2


(Callon, 1986) (problematisation, interessement, enrolment,and mobilisation) can be used to guide project managers.If the process of translation is successful, a network ofaligned interests is formed. Secondly, in order to identifystakeholders and understand their evolution throughout theproject, practitioners have to observe not only human butalso non-human actors that they produce or put intocirculation and their associations or influence during theproject. Such observation allows us to deeply understand thenature and evolution of stakeholders during the project. Forinstance, practitioners have to focus on human stakeholders,such as the users or the project's partners, as explained intraditional stakeholder analysis, as well as on the technologythat they used or developed in the project, the softwarecomponents, or the partnership contract. In our case study,the role and place of the stakeholder ERI in the net-work change according to its relation with not only otherhuman's stakeholders (Centile), but also non-human'sstakeholders (the technology when they discover its opensource software; the partnership contract, etc.). This kind ofobservation is very important because every entity (e.g.,project manager, technology, users, software components) isthe result of the actions that it can and does form with bothhuman, such as members of the Ministry, and non-human,such as computer systems, contracts, or open sourcesoftware.

• How to observe the stakeholder network with a stakeholderrelational ontology implies focusing on the relationships toidentify the stakeholders. Here, we observe the intermedi-aries in the project; that is, what is produced and put intocirculation to deeply identify the stakeholders all alongthe project. Our observation concerns the reworking ofthe traditional notion of causality, where the future isdetermined in terms of a linear causality. We suggestthat no predetermined, unchanging agents can cause some-thing to happen. The causes and effects emerge throughinteractions. Consequently, our type of observation of aproject network focuses on the ways in which an actor isdefined according to his or her relationships. Here, thethoughts, feelings, actions, and even identity of any actor(such as project managers) are the result of the actions thatthe actor can and does form with other entities, which maybe human, such as team members and the French Ministry,or non-human, such as computer systems, contracts, oropen source software.

• When: at the beginning of the project (morphologicalanalysis) and at each controversy (dynamic analysis of theproject). By focusing on controversies, the analysis isdynamic because it is not only a front-end, but also alongitudinal process as recommended by Eskerod andVaagaasar (2012). Indeed, we propose to follow thecontroversies throughout the entire project and take intoaccount the effects of each controversy on the network andthus on the stakeholders. As such, we consider the evolutionof stakeholders' roles and natures after each controversythroughout the project. Consequently, the analysis isdynamic in so far as it allows a project to be perceived as

an emerging network (associations, relations) that extendsand transforms over time. Controversy appears thus to be arelevant marker of the stakeholder network's transformationand evolution throughout the project. We assume that theproject manager does not have to necessarily observe andforesee everything at all times, but s/he should be awareof the relevant markers, such as the associations andcontroversies.

6.3. Limitations and suggestion for further research

All research studies have limitations, and this study is noexception. Firstly, beyond the theoretical and managerialcontributions of our work, our study has a methodologicallimitation regarding the external validity of the mentionedresults and, thus, their scope. This study provides a single casestudy of a failed French education IT project. Accordingly,our proposed approach requires more case studies to confirmthe scope of our framework. Although this framework wasapplied to a second IT project that did succeed, for the sake ofbrevity, we have chosen to present only one case study.However, at this stage of our research, we cannot claim thatour results are broadly applicable. Nevertheless, the initial aimof our work was to provide analytically rather than statisticallygeneralisable results (Yin, 2008), with the aim of enriching themost recent work on stakeholder management in projects.Secondly, project actors did not use our conceptual approachin practice. We proposed this framework based on ANTconcepts and used it to read and understand the evolution ofstakeholders' network in the “Pupitre Virtuel” project.Thirdly, the application of ANT to stakeholder theory haslimitations. It is important to understand that ANT regards thenetwork as a concept rather than as a thing; it is a tool thathelps describe something (Latour, 2005). Thus, the project is aproposed emerging network that extends and transforms overtime.

Finally, these limits lead us to emphasise two avenues ofresearch, which seem promising. One perspective would be toextend the qualitative approach to new areas of investigationand other project contexts and thus improve external validity,because we acknowledge that our findings, related to our singlecase study, have a strong internal validity to the detriment oftheir external validity. A second promising research perspec-tive could apply our framework in real time and observe how ithelps actors manage the project. Ultimately, we hope that ourwork will contribute to a deeper understanding of stakeholderanalysis of and engagement in projects. More generally, thisresearch aims to be a step towards improving our understand-ing of stakeholder management in temporary and project-basedorganisations.

Conflict of Interest

The authors have no conflict of interest.

Appendix A. Matrix of stakeholders' identification and their relationships

Features Groups actors

The local government(LG)

French Ministry ofNational Education

The University in thedepartment

The users inexperimentation inthe department

The company ofengineering Centile

The Pupitre Virtuel The private societyERI

Aims in theproject

Generalise IT's useon the nationalterritory and generaliseit to ensurea large number ofusers, have return oninvestment

Develop and establishthe ENT in all theFrench schools by2010

Improve IT andpedagogy

Participation on theexperimentation toimprove learning

Improve the technology Let be improved andtested

Profitability of itsinvestments on theproject by itsgeneralisation

Status in theproject

Project manager in thelocal project

Coordinator in theglobal project

Research/developmentand exploitation licenceof the technology inuniversities

Responsible ofpedagogy

Engineering anddevelopment of the IT

Numerical platform intest and generalisation

Responsible of commercialand technical developmentof the technology

Activity in theproject

Planification andexecution of time, cost,and resources

Supervise local projectand help projectmanagement byrecommendationsPut label on theproject

Follow experimentationsand research on thisCreation of a foundationof the usages of this ITExploitation of the IT inuniversities

Test and evaluate thetechnologyApprove/disapproveconcepts

Technical developmentsof the technology

Access to students totheir numerical content

Order technical andcommercial developmentto CentileResearch and supervise newmarkets and experimentationoutside the department

Intermediariesproduced

• Funds necessary forIT development

• A person manages localexperimentations in thedepartment

• Recommendations fromthe choices of the software, the functionalitiesor project

• Management• Label for good projects

• Research papers• Analyses

• Returns on problemsduring technologyusages

• Utilisation guide andframeworks

• Technical developmentsof the IT andfunctionalities

• Operation anddysfunction

• Order to Centile• Analyses of

functionalities

Poles Financial pole Institution Scientific pole Market pole Technical development pole Technological pole Commercial pole

1120S.M

issonier,S.L

oufrani-Fedida

/International

Journalof

Project

Managem

ent32

(2014)1108–1122


References

Aaltonen, K., 2011. Project stakeholder analysis as an environmentalinterpretation process. Int. J. Proj. Manag. 29, 165–183.

Achterkamp, M.C., Vos, J.F.J., 2008. Investigating the use of the stakeholder notionin project management literature, a meta-analysis. Int. J. Proj. Manag. 26,749–757.

Akrich, M., Callon, M., Latour, B., 2002a. The key to success in innovation partI: the art of interessement. Int. J. Innov. Manag. 6, 187–206.

Akrich,M., Callon,M., Latour, B., 2002b. The key to success in innovation part II:The art of choosing good spokespersons. Int. J. Innov. Manag. 6, 207–225.

Alderman, N., Ivory, C., 2011. Translation and convergence in projects: anorganizational perspective on project success. Proj. Manag. J. 42, 17–30.

Beringer, C., Jonas, D., Kock, A., 2013. Behavior of internal stakeholders in projectportfolio management and its impact on success. Int. J. Proj. Manag. 31,830–846.

Blackburn, S., 2002. The project manager and the project-network. Int. J. Proj.Manag. 20, 199–204.

Bourne, L., Walker, D.H.T., 2005. Visualising and mapping stakeholderinfluence. Manag. Decis. 43, 649–660.

Brown, A.D., Jones, M.R., 1998. Doomed to failure: narratives of inevitabilityand conspiracy in a failed IS project. Organ. Stud. 19, 73–88.

Callon, M., 1986. Some elements of a sociology of translation: domestication ofthe scallops and the fishermen. In: Law, J. (Ed.), Power, Action and Belief:A New Sociology of Knowledge. Routledge & Kegan, London.

Callon, M., 1991. Techno-economic networks and irreversibility. A Sociologyof Monsters: Essays on Power, Technology and Domination, 38 132–161.

Callon, M., 1992. The dynamics of techno-economic networks. In: Coombs, R.,Saviotti, P., Walsh, V. (Eds.), Technical Change and Company Strategies.Academic Press, London.

Callon, M., Law, J., 1995. Agency and the hybrid “Collectif”. The SouthAtlantic Quarterly 94, 481–507.

Chinyio, E.A., Akintoye, A., 2008. Practical approaches for engagingstakeholders: findings from the UK. Constr. Manag. Econ. 26, 591–599.

Cleland, D.I., 1986. Project stakeholder management, Project ManagementHandbookSecond edition. 275–301.

Coakes, E., Elliman, T., 1999. Focus issue on legacy information systems andbusiness process engineering: the role of stakeholders in managing change.Commun. AIS 2, 4.

Cova, B., Salle, R., 2006. Communications and stakeholders. In: Pryke, S.,Smyth, H. (Eds.), The Management of Complex Projects: A RelationshipApproach, pp. 131–146 (UK).

Crawford, L., 2005. Senior management perceptions of project managementcompetence. Int. J. Proj. Manag. 23, 7–16.

Cummings, J.L., Doh, J., 2000. Identifying who matters: mapping key playersin multiple environments. Calif. Manag. Rev. 42, 83–104.

Czarniawska, B., Hernes, T., 2005. Actor–Network Theory and Organizing.Liber and Copenhagen Business School Press, Copenhagen, Denmark (ed).

Dery, K., Hall, R., Wailes, N., Wiblen, S., 2013, 225-237. Lost in translation?An actor–network approach to HRIS implementation. J. Strateg. Inf. Syst.22 (3), 225–237.

Elbanna, A., 2010. Rethinking IS project boundaries in practice: a multiple-projects perspective. J. Strateg. Inf. Syst. 19, 39–51.

Eskerod, P., Vaagaasar, A.L., 2012. Stakeholder management strategies andpractices during a project course, In the 28th EGOS Colloquium.

Faik, I., Walsham, G., 2012. Modernisation through ICTs: towards a networkontology of technological change. Inf. Syst. J. 23 (4), 351–370.

Freeman, R.E., 1984. Strategic Management: A Stakeholder Approach.Ballinger, Cambridge, Mass.

Granovetter, M.S., 1973. The strength of weak ties. American Journal ofSociology 1360–1380.

Greenwood, M., 2007. Stakeholder engagement: beyond the myth of corporateresponsibility. J. Bus. Ethics 74, 315–327.

Jensen, T., Sandström, J., 2011. Stakeholder theory and globalization: thechallenges of power and responsibility. Organ. Stud. 32, 473–488.

Jepsen, A.L., Eskerod, P., 2009. Stakeholder analysis in projects: challenges inusing current guidelines in the real world. Int. J. Proj. Manag. 27, 335–343.

Langley, A., Smallman, C., Tsoukas, H., Van de Ven, A.H., 2013. Processstudies of change in organization and management: unveiling temporality,activity, and flow. Acad. Manag. J. 56, 1–13.

Latour, B., 1988. In: Sheridan, Alan, Law, John (Eds.), The Pasteurization ofFrance, Trans.

Latour, B., 1999. On recalling ANT. In: Law, J., Hassard, J. (Eds.), ActorNetwork Theory and After. Oxford.

Latour, B., 2005. Reassembling the social — an introduction to actor–network-theory. Reassembling the Social — An Introduction to Actor–Network-Theory, by Bruno Latour. Foreword by Bruno Latour. Oxford University Press,p. 316 (Sep 2005. ISBN-10: 0199256047. ISBN-13: 9780199256044 1).

Littau, P., Jujagiri, N.J., Adlbrecht, G., 2010. 25 years of stakeholder theory inproject management literature (1984–2009). Proj. Manag. J. 41, 17–29.

Luoma-aho, V., Paloviita, A., 2010. Actor–networking stakeholder theory fortoday's corporate communications. Corp. Commun. Int. J. 15, 49–67.

Magness, V., 2008. Who are the stakeholders now? An empirical examinationof the Mitchell, Agle, and Wood theory of stakeholder salience. J. Bus.Ethics 83, 177–192.

Mähring, M., Holmström, J., Keil, M., Montealegre, R., 2004. Trojan actor–networks and swift translation: bringing actor–network theory to IT projectescalation studies. Inf. Technol. People 17, 210–238.

Meier, O., Missonier, S., 2012. Trajectory of an IT project network:convergence, divergence and adjustment process. European Conference onInformation Systems (ECIS) Proceeding. Barcelona, Spanish, pp. 1–15.

Miles, M.B., Huberman, A.M., 1994. Qualitative Data Analysis: An ExpandedSourcebook. Sage Publications, Incorporated.

Mitchell, R.K., Agle, B.R., Wood, D.J., 1997. Toward a theory of stakeholderidentification and salience: defining the principle of who and what reallycounts. Acad. Manag. Rev. 853–886.

Nayak, A., Chia, R., 2011. Thinking becoming and emergence: processphilosophy and organization studies. Res. Sociol. Organ. 32, 281–309.

Orlikowski, W.J., Scott, S.V., 2008. 10 Sociomateriality: challenging theseparation of technology, work and organization. Acad. Manag. Ann. 2,433–474.

Pacheco, C., Garcia, I., 2012. A systematic literature review of stakeholderidentification methods in requirements elicitation. J. Syst. Softw. 85 (9),2171–2181.

Papadopoulos, T., Merali, Y., 2008. Stakeholder network dynamics andemergent trajectories of Lean implementation projects: a study in the UKNational Health Service. Public Money Manag. 28, 41–48.

Parmar, B.L., Freeman, R.E., Harrison, J.S., Wicks, A.C., Purnell, L., de Colle,S., 2010. Stakeholder theory: the state of the art. Acad. Manag. Ann. 4,403–445.

Pollack, J., Costello, K., Sankaran, S., 2013. Applying Actor-Network Theoryas a sensemaking framework for complex organisational change programs.Int. J. Proj. Manag. 13, 1118–1128.

Pouloudi, A., Whitley, E.A., 1997. Stakeholder identification in inter-organizational systems: gaining insights for drug use management systems.Eur. J. Inf. Syst. 6, 1–14.

Pouloudi, A., Gandecha, R., Atkinson, C., Papazafeiropoulou, A., 2004. Howstakeholder analysis can be mobilized with actor–network theory to identifyactors. Inf. Syst. Res. 705–711.

Ramiller, N.C., Wagner, E.L., 2009. The element of surprise: appreciating theunexpected in (and through) actor networks. Inf. Technol. People 22,36–50.

Rowley, T.J., 1997. Moving beyond dyadic ties: a network theory ofstakeholder influences. Acad. Manag. Rev. 22, 887–910.

Savage, G.T., Nix, T.W., Whitehead, C.J., Blair, J.D., 1991. Strategies forassessing and managing organizational stakeholders. The Executive 5, 61–75.

Slife, B.D., 2004. Taking practice seriously: toward a relational ontology. J.Theor. Philos. Psychol. 24, 157.

Van de Ven, A.H., Poole, M.S., 1995. Explaining development and change inorganizations. Acad. Manag. Rev. 510–540.

Venturini, T., 2010. Diving in magma: how to explore controversies withactor–network theory. Public Underst. Sci. 19, 258–273.

Vidgen, R., McMaster, T., 1996. Black boxes, non-human stakeholders and thetranslation of IT through mediation. Information Technology and Changesin Organizational Work 250–271.

http://refhub.elsevier.com/S0263-7863(14)00032-5/rf0005





































































































































Wagner, E.L., Newell, S., Piccoli, G., 2010. Understanding project survival inan ES environment: a sociomaterial practice perspective. J. Assoc. Inf. Syst.11, 276–297.

Walsham, G., 1993. Interpreting Information Systems in Organizations. JohnWiley & Sons, Inc.

Walsham, G., 1997. Actor–network theory and IS research: current status andfuture prospects. Information Systems and Qualitative Research. Springer,pp. 466–480.

Winter, M., Smith, C., Morris, P., Cicmil, S., 2006. Directions for futureresearch in project management: the main findings of a UK government-funded research network. Int. J. Proj. Manag. 24, 638–649.

Yang, J., Shen, G.Q., Ho, M., Drew, D.S., Chan, A.P.C., 2009a. Exploringcritical success factors for stakeholder management in construction projects.J. Civ. Eng. Manag. 15, 337–348.

Yang, J., Shen, Q., Ho, M., 2009b. An overview of previous studies instakeholder management and its implications for the construction industry.J. Facil. Manag. 7, 159–175.

Yin, R.K., 2008. Case Study Research: Design and Methods. SAGEPublications, Incorporated.




















Documents

Stakeholder analysis and engagement in projects: From ... · PDF fileStakeholder analysis and engagement in projects: From stakeholder relational perspective to stakeholder relational