Search

The complexity of business markets, resulting from different levels of organization and the ways social constructs combine across levels, means no clear theory of relational coordination has been developed. The relationship and network framework of the industrial marketing and purchasing (IMP) group provides a means of handling this complexity. This paper proposes a collaborative interest model (CIM) of relational coordination, which partially overcomes these problems by examining the coordination process within a context that accounts for levels of organization. This conceptualization of coordination processes resulting from firm intentions to achieve relationship performance presents a new way to empirically examine relational coordination. An empirical test of the model using structural equation methodology shows relational coordination (i.e., commitment and trust) explains relationship performance better than market coordination mechanisms.

The literature proposes a number of models explaining ethical behaviours but these are seldom of the kind which can be used by marketers in their day-to-day decision making. In this study based on data collected from 166 firms operating in overseas markets, a concomitant clusterwise regression approach is used to define clusters that display good homogeneity both in traits and in models of ethical tolerance, thus allowing an ‘ethical diagnostic’ of firms. Based on readily available and objective variables, namely size, dependence on overseas markets and overseas experience, the paper demonstrates that it is possible to cluster firms into groups of which the ethical tolerance can be predicted. The managerial implications of these findings for international marketers and directions for future research are also discussed.

Business ethics is an area difficult to operationalise. The literature proposes a number of models explaining ethical behaviours but these are seldom of the kind which can be used by marketers in their day-to-day decision making. In this study, based on data collected from 166 firms operating on overseas market, a conjoint clusterwise regression approach is used to predict firms’ ethical tolerance while at the same time homogenising the profiles of firms within clusters. The analysis is based on simple and visible organisational variables, allowing an ’ethical scoring’ of firms and the identification of meaningful models which may be used for present as well as future dealings with other firms. The resulting clusters are interpreted both in terms of the model which they follow as well as in terms of their profiles.

In business science, the studied objects are often groups of partners rather than independent firms. Extending classical segmentation to these polyads raises conceptual problems, principally: defining what should be consid- ered as common or specific at the partners' and at the segment levels. The general approaches consist either in merging partners characteristics and performances into a single macro-object, thus loosing their specific contributions to each partner's performance, or in modelling partners' performance as if their models were inde- pendent. As a step to understanding, how partnership influences firms' perform- ance, the dyadic (i.e. two partners') case is studied. First, some theoretical issues concerning the degrees of individual and contributive interest in a dyadic popula- tion are discussed. Next, partnership's conceptualisation is based upon two models for each firm: a "self-model" that reflects how the firm's characteristics explain its own performance, and a "contributive-model" model that reflects how these characteristics influence the partner's performance. This allows definition of three relationship modes: merging, teaming and sharing. Subsequently, dyad segmenta- tion strategies are discussed according to their capacity to reflect the modes of part- nership and a dyadic clusterwise regression method, based on a genetic algorithm, is presented. Finally, the method is illustrated empirically using actual data of busi- ness partners in the software market.

An approach is proposed for the determination of a population's dominant means-end chains that relies on two criteria: the number of the represented ladders and the measure of their semantic representatives. These measures are provided from a multidimensional analysis, fitted to the semantic characteristics of means-end data collection and concepts. Distances in the semantic space are used to identify the dominant chains through an objective “ chain-constrained ” clustering method. This is a versatile method for analyzing means-end processes and conceptualizing the components of advertising strategy.

Cet article propose une méthode de mesure du halo affectif en marketing. Le concept de halo affectif et l'intérêt de sa mesure sont d'abord exposés. Les principales difficultés rencontrées par les travaux antérieurs sur le filtrage et la mesure du halo en marketing sont ensuite discutées. Puis une méthode originale de mesure est proposée, avec une illustration pratique de l'influence du halo affectif sur les attitudes.

The validity of means-end data is discussed from a conceptual and empirical perspective. A dynamic programming approach (Markov chain) is then proposed that enables to measure the validity and reliability of any group of means-end data (item, hierarchical level, whole data basis), thus enabling to purify it and to have empirical insight on debated topics such as the number of means-end levels to be considered and the validity of the means-end data collection methods.

A bio-mimetic approach of multiple goal problem solving is proposed. The approach relies on the stochastic ope-rators (reproduction, selection) of the Genetic Algorithms (GA). The method is presented in the case of a three-goal management problem: market segmen-tation and compared empirically with the two-goal bio-mimetic algorithm (TYPREN) from which it is derived and with the statistical Regression Mixture Model (GLIMMIX).

"Scheme rearrangement" was initially proposed to improve the fit and the robustness of a genetic algorithm. It was hypothesized that, in a problem comprising non linear or epistatic interactions, a greater efficiency would result from bringing closer the parameters of the interacting variables. The method was assessed, with mild success, in the specific case of the travelling salesman but not in multivariate models. It was not tested neither with real-coded parameters, though they are more adapted for identifying meaningful schemes. In the paper, we check whether scheme rearrangement provides better fit and whether it may help identifying latent interactions. These issues are studied and empirically tested with a typical problem of market segmentation: the clusterwise regression. The study supports the hypothesis that scheme rearrangement provides better solutions but not that it results from latent interactions

Foreword. 1. The Potential of Neural Networks Evaluated within a Taxonomy of Marketing Applications S. Pandelidaki, A.N. Burgess. 2. Complexity Control and Generalization in Multilayer Perceptrons P. Gallinari, T. Cibas. 3. Application of Neural Networks to Bond Rating and House Pricing H. Daniels, et al. 4. Connexionist Approach and Corporate Distress Diagnosis: A Contribution to the Processing of Incomplete Information J.-F. Casta, B. Prat. 5. Comparison of the Predictivity of a Neural Network with Backpropagation with Those Using Linear Regression, Logistic and A.I.D. Methods for Direct Marketing Scoring P. Desmet. 6. Comparison of Discriminant Analysis and Neural Networks Application for the Detection of Company Failures M. Bardos, Wenhua Zhu. 7. Using Artificial Neural Nets to Specify and Estimate Aggregate Reference Price Models M. Natter, H. Hruschka. 8. Towards Connectionist Merchandising: A Conceptual and Operational Definition Ph.R. Demontrond, D. Thiel. 9. A Statistical Methodology for Specifying Neural Network Models: Application to the Identification of Cross-Selling Opportunities A.N. Burgess, S. Pandelidaki. 10. Optimization by a Genetic Algorithm of Stochastic Linear Models of Time Series R. Bone, et al. 11. A Bio-Mimetic Clusterwise Regression Algorithm for Consumer Segmentation J.-M. Aurifeille. 12. Hybrid Genetic Learning of Hidden Markov Models for Time Series Prediction M. Slimane, et al. 13. Learning How to Regulate a Polluter with Unknown Characteristics: An Application of Genetic Algorithms to a Game of Dynamic Pollution Control T. Vallee, C.Deissenberg.

A clusterwise regression method is proposed in the case of binary variables. Its objective is to improve the marketing operationality of the procedure by optimizing clusters’ predictivity under constraints of homogeneity and differentiation. The algorithm is a hybrid of genetic algorithm (GA) and of learning vector quantization (LVQ). GA is used for optimizing predictivity under constraints of cluster homogeneity. Specific LVQ rules are implemented for speeding up the process. In an empirical test, based on actual marketing data, the hybrid LVQ/GA compared satisfactorily with Kohonen’s original LVQ (OLVQ).

To communicate effectively, marketers need to know the process according to which consumers perceive and choose brands. Within these processes, consumer certainty plays an important role, particularly in the determinacy of product items. In this thesis, this role is analyzed through a variable that globalization contributes both to distort and amplify: brand's nationality. Data analysis and statistics indicates: The « mediating » effect of attitude toward the brand, which indicates "how" the country's image and certainty can influence consumer purchase intentions, The « moderating » effect of certainty of belief, which indicates "When" consumers can use an image of the « country of origin » in a decision process. Managerial and theoretical implications are discussed, so that managers can monitor the nationality associated with their brands and avoid negative effects for the company.; Pour communiquer efficacement, les responsables marketing ont grand besoin de connaître les processus selon lesquels les consommateurs perçoivent et choisissent les marques. Au sein de ces processus, la certitude que le consommateur accorde à ses croyances joue un rôle important, notamment dans la déterminance des caractéristiques du produit. Dans la présente thèse, ce rôle est analysé à travers une variable que la mondialisation contribue à la fois à déformer et à amplifier : la nationalité de la marque. Il s'avère que la certitude dans la croyance envers la nationalité d'une marque est un élément capital du succès d'un positionnement stratégique sur les marchés internationaux. L'analyse des données et les statistiques effectuées ont permis d'identifier : • L'effet médiateur de l'attitude envers la marque, qui indique « comment » l'image du pays et la certitude qui s'y attache peuvent influencer les intentions d'achat de la marque, • L'effet modérateur de la certitude de la croyance envers la nationalité de la marque, qui indique « quand » les consommateurs risquent d'utiliser l'image du pays d'origine dans le processus de choix. Les implications théoriques et managériales de ces résultats sont discutées, de sorte que les managers puissent surveiller les origines nationales associées à leurs marques et éviter qu'elles soient défavorables ou en contradiction avec la stratégie marketing de la firme.

International audience

International audience

International audience

Nous montrons comment il est possible d'utiliser l'algorithme d'auto organisation de Kohonen pour traiter des donn\'ees avec valeurs manquantes et estimer ces derni\`eres. Apr\`es un rappel m\'ethodologique, nous illustrons notre propos \`a partir de trois applications \`a des donn\'ees r\'eelles. ----- We show how it is possible to use the Kohonen self-organizing algorithm to deal with data which contain missing values and to estimate them. After a methodological recall, we illustrate our purpose from three real databases applications.

La thèse consiste à étudier les interactions entre les composantes cognitive, affective et conative des attitudes des consommateurs. L'hypothèse d'une congruence des trois composantes attitudinales dans un espace commun est testée au moyen d'une analyse multidimensionnelle des dissimilarités individuelles (INDSCAL). Les résultats obtenus font ressortir, après analyses typologiques complémentaires, l'existence de profils attitudinaux caractéristiques des individus et des produits. Ces profils conduisent à des propositions managériales en matière de positionnement et de persuasion., no abstract

The research investigates and emphasizes the importance of boycotting which is an innovative, 'bottom-up' form of communication between the consumer and the firm which indicates that the needs of some consumers have not been understood. The work is largely conceptual in nature and consists of researching and modelling the various dimensions of boycotting. A definition of consumer boycotting is suggested and a predictive model is supplied to demonstrate the potential impact of boycotting on a targeted organization over the short, medium and long terms. A structural equation model was used to analyse the link between the motivations of those involved in various forms of boycotting as compared to its organizational impact. A twofold approach was used recognizing the cognitive and affective dimensions of boycotting at each stage of the model. In this manner management should be able to implement strategies to either avoid a potential boycott or to disentangle themselves from an existing one at low cost., La recherche attire l'attention sur le thème du boycott et met en évidence son intérêt. Le boycott constitue un mode de communication original entre le consommateur et l'entreprise. Il s'agit, pour une fois, d'une communication ascendante du consommateur vers l'entreprise. Celle-ci est la preuve que certains besoins de consommateurs n'ont pas été compris. Le travail étant essentiellement conceptuel, il a consisté à rechercher les dimensions du boycott et les principaux éléments d'un modèle de comportement de boycott. Il conduit à proposer une définition du boycott et à élaborer un modèle prédictif des effets du boycott à court, moyen et long terme sur l'organisation visée. L'analyse des structures de covariance permet de tester un modèle liant les motivations des boycotteurs aux effets des boycotts par l'intermédiaire des formes de boycott choisies. L'approche duale retenue tente d'appréhender les dimensions cognitives et affectives du boycott à chaque étape du modèle. Les relations observées permettent de faire des propositions managériales visant un boycott ou à s'en dégager à moindres frais.

This research aims at explaining the determinants of sponsorship in the context of sports events. Approval, the dependant variable, is an indicator of sponsorship effectiveness. The independant variables are : values and time orientation. The study leads to : - a new segmentation of sponsorship's targets, using the spectators' values ; - a new measurement scale of sponsorship approval ; besides, it demonstrates the positive role of spectators' values on sponsorship approval, L'objet de la recherche est d'étudier l'agrément retenu comme indicateur affectif de l'efficacité d'une action de parrainage sportif. Plus précisément, nous évaluons le rôle de deux variables de segmentation : les valeurs et l'orientation temporelle des spectateurs de l'événement sportif. L'étude met en avant : - une nouvelle segmentation des cibles du parrainage sportif à partir des valeurs des spectateurs ; - une nouvelle échelle de mesure de l'agrément propore au parrainage sportif, cette dernière étant opérationnelle ; - l'incidence positive des valeurs des spectateurs sur l'agrément du parrainage sportif.

The purpose of this research is to demonstrate that the conception of time is a partial mediator in the causal process by which personal goals (values) have an influence in the frequency of consumption. This hypothesis is tested within the consumption of theater because of its timing characteristics. The analysis of structural covariances confirm the hypothesis : the frequency of consumption is determined by the conception of time, which is itself influenced by personal goals. The optimistic vision of the future (a dimension accounted for by proverbs) as well as the constancy leads to a high frequency of consumption. The lack of planification, a rational conception of time and (persévérance) are the main impediments. The theatre subscription is particulary determined by the locus of control. To measure the concepts of the tested model, two scales were developed combining exploratory and confirmatory factorial analysis : personal goals (values) and the conception of time. Out of an enriched definition of the latent dimensions, the specificity of the scales is to be validated after having been tested for their predictive capacity. The managerial applications, limits and implications for the future conclude the research., L'objectif de la recherche est de montrer que la conception du temps est un médiateur partiel dans le processus causal par lequel les buts personnels (valeurs) exercent une influence sur la fréquence de consommation. Cette hypothèse est testée dans le cadre de la consommation théâtrale du fait de ses caractéristiques liées au temps. Les analyses de structures de covariances permettent de confirmer l'hypothèse émise : la fréquence de consommation est déterminée par la conception du temps, elle-même influencée par les buts personnels. La vision optimiste de l'avenir (dimension mesurée par des proverbes) ainsi que la constance conduisent à une fréquence élevée de consommation théâtrale. L'absence de planification, la conception rationnelle du temps et la persévérance en sont les principaux freins. L'abonnement est particulièrement déterminé par l'origine du contrôle. Afin de mesurer les concepts du modèle testé, deux échelles ont été développées en combinant analyses factorielles exploratoires et confirmatoires : les buts personnels (valeurs) et la conception du temps. Outre une définition enrichie des dimensions latentes, leur spécificité est de n'être validées qu'après avoir été soumises à leurs capacités prédictives. Les implications managériales, limites et prolongements concluent la recherche.

Many researches have focused on the person of the SME's owner-manager, but few research has been made upon VSME (less than 10 employees), whereas they form the greater part of the economic world. One of the major problems met by those VSME is the problem of debts. The basic idea of this research is that the values of the owner-manager, and his mental representations of bank, have an effect on the amount and the nature of debts. A two stage methodology has been used. First, a structural model has been achieved to etimate the causality. This model failed. Then a predictive clustering using a neural network has been developped, the optimization being controlled by a genetic algorithm. This procedure gave positive results. For each cluster of owner-manager, the link between values, mental representations and debts has been shown., De nombreux travaux ont été menés pour cerner la personnalité du dirigeant-propriétaire de P.M.E. La catégorie particulière des très petites entreprises (moins de 10 salariés) a été l'objet de moins d'attention alors qu'elles représentent la majorité du tissu industriel, commercial et artisanal. L'un des problèmes majeurs rencontrés par ces entreprises est le problème du financement. Le postulat de ce travail de recherche est que les valeurs du dirigeant-propriétaire de T.P.E. et les représentations mentales qu'il se fait de la banque influencent ses choix financiers. Une méthodologie en deux temps a été retenue. Un modèle d'équations structurelles a d'abord été élaboré pour faire apparaitre les liens de causalité entre les variables. Ce modèle n'ayant pas donné totalement satisfaction, une typologie prédictive par réseaux de neurones formels optimisés par algorithme génétique a été développée et a donné les résultats attendus. Pour chaque classe de dirigeant-propriétaire le lien a été établi entre les valeurs, les représentations mentales et la structure financière.