How to Generate Affective Reactions Through Social and Spatial Immersion on a Merchant Website&Nbsp;: Proposal of an Integrative Model

ABSTRACT - The objective of this paper is to theoretically clarify, and to make propositions, on the role and impacts of immersion in the consumer behaviour on a merchant website. Through the results of focus group interviews, we explain how virtual-social interactions between consumers and Internet virtual agents affect major Web behaviours such as visit duration, repeated visit, as well as generation of word of mouth.

Citation:

Brice Pablo de Diesbach and Anne-Cecile Jeandrain (2005) ,"How to Generate Affective Reactions Through Social and Spatial Immersion on a Merchant Website&Nbsp;: Proposal of an Integrative Model", in AP - Asia Pacific Advances in Consumer Research Volume 6, eds. Yong-Uon Ha and Youjae Yi, Duluth, MN : Association for Consumer Research, Pages: 244-251.

Asia Pacific Advances in Consumer Research Volume 6, 2005      Pages 244-251

HOW TO GENERATE AFFECTIVE REACTIONS THROUGH SOCIAL AND SPATIAL IMMERSION ON A MERCHANT WEBSITE : PROPOSAL OF AN INTEGRATIVE MODEL

Brice Pablo de Diesbach, ESSEC / IAE Aix-en-Provence, France

Anne-Cecile Jeandrain, Catholic University of Louvain, Belgium

ABSTRACT -

The objective of this paper is to theoretically clarify, and to make propositions, on the role and impacts of immersion in the consumer behaviour on a merchant website. Through the results of focus group interviews, we explain how virtual-social interactions between consumers and Internet virtual agents affect major Web behaviours such as visit duration, repeated visit, as well as generation of word of mouth.

INTRODUCTION

Since one decade the world is facing a new and revolutionary manner in which people shop: Internet shopping. For consumers, it is clearly recognized that internet offers many opportunities to facilitate shopping, mainly utilitarian shopping. Indeed, searching information and purchasing product are considered as easier due to greater product information availability and access, possibility of multi-attribute product comparisons, shopping at home 24h/24 and 7d/7 and so on (Alba et al., 1997; Szymanski and Hise, 2000; Coupey, 2001).

What about experiential shopping ? Even if internet could have a dual personality (Childers et al., 2001)Bi.e., a "cold" media due to its informational architecture (Boulaire and Balloffet, 1999) as well as a potential hedonic environment due to some hedonistic aspects (Korgaonkar and Wolin, 1999), today technological advances in interactive techniques (for instance, 3-D product manipulation, personification in avatar) considerably broaden the extent to which internet could be also exploited for experiential shopping (HelmT-Guizon, 2001, Jeandrain and Limbourg, 2002; Diesbach, 2003).

The relevance of such hedonistic dimension in understanding the consumer behaviour is highly important in a very large variety of consumption contexts, as the seminal work of Holbrook & Hirschman 1982, Holbrook (1986, 1994), Csikszentmihalyi 2000a, has shown.

More specifically in the internet context, for consumers, these advances open the door to using virtual reality (and thus immersion [Immersive virtual reality implies the replacement of the individual=s physical environment with the virtual environment (Potter, 1996), that means an inseparable perception of the individual=s from his perception of the virtual environment.]) for their "everyday" online experience. [Beside technological advances in interactive techniques, the decreasing cost of some immersive virtual reality devices (such as data gloves and head-mounted display), the availability of 3D software, etc., progressively allow consumers to use virtual reality (Picard, 1998; Pichard and Klein, 2001).] As a consequence, the more immersive, hedonic aspects of internet could be found to play at least an equal role than its instrumental ones in predicting online shopping attitudes (Childers et al., 2001; Diesbach 2001). [Effectively, now Consumers want experience that dazzles their senses, touches their hearts, and stimulates their minds (Schmitt, 1999; Pine and Gilmore, 1999). An immersive interactive website, i.e. that generates a feeling of immersion (see further in this article) provides ideal forum for many companies to create such compelling experience (Novak et al., 2000).] However, to our knowledge only few pieces of research have investigated in an experiential perspective how the consumer deals with interactive interface [For a rationale perspective, see the seminal work from Ariely (2000).] (e.g., Meuter et al., 2000), and more specifically with immersive interactive interface (e.g., Grigorivici, 2003). Given the wide variety of website types (i.e., utilitarian- or experiential-oriented websites) and purposes (e.g., purchase, image), it is critical to understand how consumers feel about them (Meuter et al., 2000).

Our research aims at building a model integrating consumer’s immersion in an online consumer behavior perspective. For doing so, our objective is twofold. First, we attempt to study the contribution of spatial and of social immersion in the specific case of merchant website. Second, the notion of autonomous agent is introduced as an ambient cue in order to deeper analyze the social counterpart of immersion in our perspective. Practically, after epistemologically positioning our research, we will precise the notions of merchant website, spatial and social immersion and autonomous agent. Then, an integrative model will be explained and justified. We will conclude by proposing insights for future research.

EPISTEMOLOGICAL PRELIMINARY

It appears that the huge expectations of the late 1990’s in term of internet-based service, commerce and general business development were not met: it is proposed that a number of professionals and academics may have not considered the internet phenomena properly, sometimes ignoring that consumer still needed online stimuli, services, feelings they were to experience off-line (Diesbach 2002). The internet as a tool has been and still is bringing important paradigmatic and theoretical questions (e.g., Hoffman & Novak, 1996; Helme-Guizon, 2001).

Some authors may propose to totally renew a number of basic postulates of the traditional consumer behaviour approach to understand the "new" consumer and hence propose a paradigmatic revolution under a number of aspects (e.g., Hoffman et Novak, 1997). Some authors also strongly focus in their analysis on a reduced set of characteristics, specific to electronic interfaces, which may ignore important variables for having a full picture of how satisfaction, retention, may be generated online. For example a number of scholars in Marketing, and ergonomists, in the late 1990’s do focus mainly on downloading speedness and navigation "ease" (e.g. Nilsen, 1997). Some authors nevertheless highlight, without detailing them, the possible importance of other (eg affective), cues (Chen & Wells 1999, Lynch & al 2001, Ladwein 2001). A renewed vision of Internet as a tool may make affective reactions relevant, particularly in considering the shopping experience online, as proposed by Helme-Guizon, 2001.

The key question is hence: * do we properly consider, approach internet as a tool, a technology, and as a social phenomena ? +.

It can be assumed that such was not the case, due to the failure of so many e-business and particularly e-commerce operators in the 2001-2002. It is proposed by Wind & Mahajan (2002) in their seminal book that the internet user or consumer should be considered as altogether "the same" and "a different being". In some instances, adopting a totally renewed look for capturing, understanding a situation may be necessary, as old schemes sometimes blur our mind and bias our perception: then proposing a drastically new paradigm may be necessary and more operational, Hirschman (1980). Such is the new focus proposed by Wind et Mahajan (2002), for whom the consumer on internet partly follows traditional pattern, but the behaviour of whom also requires considering that he partly obeys totally different patternsBhence the metaphor of the centaur. The centaur is a Greek mythical being, both human (ie the traditional being) and horse-like (the e-consumer), i.e. a sort of hybrid being.

Applying only the traditional theoretical and measurement tools is therefore unsufficient to describe and understand him. Some of those traditional perspectives are useful, even necessary: for instance as to the role of ambiance factors (Diesbach, 2001), the role of affective reactions in general (Chen et Wells, 1999 ; Korgaonkar et Wolin, 1999). But additionally some other concepts not considered so far may be necessary, such as online immersion and the role of anthropological agents, in relation with the fact that the concepts of virtuality and reality are not necessarily opposed, or separated. In that sense, the e-consumer would touch both the old world and the new world.

CONCEPTUAL FRAMEWORK

Our conceptual framework is mainly based on two research fields: the literature on service marketing and a new definition of the concept of a merchant website, and on the man-machine (and man-virtual being) interaction, in order to explicit the notions of spatial immersion, and of virtual agents.

A theoretical revision on the concept of merchant website: towards a multi-functionality approach

First it appears necessary to clarify the notion of merchant site. Actually a website could be considered as having a precise role, function, inheriting so the traditional approach of media, distribution channels and service encounter (Hoffman & al. 1995; Hoffman & Novak 1996). The typology of websites proposed by Hoffman & al. (1995) helps understanding their uses, but what is new is the fact a lot of websites do belong to several of the commented categories.

Moreover we a number of important authors do not give an accurate signification, nor define, the concepts of merchant website, or advertising websites, therefore they do not address the question of differentiating between sites that sell online, image or content sites among advertising sites, and so on (e.g., Korgaonkar et Wolin, 1999 ; Chen et Wells, 1999 ; Mathwick a al. 2001). Bruner et Kumar (1999), Choi & al. (2001) analyse the consumer behaviour on an image website, whereas such concept is not clearly defined: this does not reveal a weakness, but rather that such concept is both assumed to be clearly understood, but maybe still needs to be refined. Internet world is moving, the frontier between "real" and "virtual" world is both widely accepted but unclear For instance a virtual agent with a synthetical, i.e. a fully-pre-programmed-base agent, voice, can be considered as a part of the virtual world. But what if an online brand, or an online service operator, decides to create a virtual agent managed from a server by a sales clerk, which speaks with the operator’s voice: are we still in the virtual or the in real world? Technology allows designing such tools, their development is hence probably just a matter of time, and market maturity. Our concepts needs therefore to be thought in a different, more comprehensive, updated manner.

Also, a detailed review of academic literature, but also a number of interviews conducted those last 2 years with site designers, marketing and sales managers in different products or service categories suggest that a website may be an excellent marketing and sales oriented tool even if it does achieve, or even propose, selling online. In this sense the function of such site would nevertheless be captured by the definition of the e-marketing as conceptualised by Vernette & Dubois (2001).

We propose that a website may be considered as a * merchant website + even if it does not sell, as soon as it helps making the consumer enter in a process that will prepare him/she to a possible, future purchaseBsuch purchase can therefore take place on-line, or off-lineBor even if it takes part into a process of information search, products comparison, or loyalty enhancement, post-purchase problem resolution, etc.., that is any functionality that could normally be performed by employees and contact persons in a traditional outlet or service encounter (Babin & al. 1995, Babin & al. 1999; Bitner 1990, 1992). The fact a transaction is not proposed, or not realized on-line, does not exclude the electronic interface from the purchase process. For instance if giving advices for making a present may be considered as a first step towards purchasing (Babin et al. 1994 ; Sirieix et Dubois, 1999), so is it in an online context (Diesbach, 2003).

Actually consumer may visit websites, in a pre-purchase process, for three main reasons: 1) surfing for utilitarian motives; 2) navigating for hedonistic motives with a purchase intention; and 3) navigating for hedonistic motives without a purchase intention. Those three behaviours are called utilitarian navigationBbrowsingBand surfing, by Lombart et Jeandrain (2003).

We believe that the frontier is not always very clear between such different kinds of navigations, as was shown by several focus groups organized in 2002 and 2003, and therefore the proposed enlarged (not definitive as we will see) definition of a merchant website makes sense. Is there also a clear frontier between an outlet and a media?

If such distinction may be clear enough in traditional commerce and communication, it may not hold true on-line: Ducoffe (1996) shows that a number of website content and design characteristics (eg text, graphics, even musics) are actually considered as advertising elements by most consumers. Therefore an informational or content site with a strongly promotional orientation may be viewed as an advertising site. In our vision such a site may be considered as a "merchant website" even if it is more content, image or relation enhancement-oriented because it prepares purchases.

Last, in order to propose a typology of "merchant websites" we will consider two variables: the degree of integration of the site into the actual or future purchase process, and the cognitive/affective nature of information.

A typology by Daras & Diesbach (2002) records the different degrees into which a website may integrates itself into an actual purchase process (figure 1). First, websites may establish a relationship with the consumer at many different steps of the company-customer relationship. Second, information may be more cognitive or affective by nature, and may require more or less cognitive processingBand processing awareness (Clore & al. 1994; Forgas 1995; Pham Cohen & Pracejus, 2001). A number of works related to the Affect-as-information framework show the relevance of such approach and the different impact strength and speedness of such information, on attitude and purchase intention. A merchant website may provide both kinds of information through textual, graphical (animated or fixed) musical, or even in the future through olfactive or tactile information. In such approach, the quantity of information, its complexity, its more affective or cognitive nature as proposed by Pham & al. (2001), are important cues that will determine the way such information is processed through the interactive media (Stevenson & al. 1999; Bruner & Kumar, 2000; Ariely, 2000; Pham & al. 2001). It must be recorded that our use of internet and other electronic interfaces through their emotional potential is still limited, but is dramatically increasing. Technology makes it possible to propose far more sensorial and/or affective cues in online communication than have been done so far.

FIGURE 1

INTERNET AND PURCHASE: WHAT IS E-COMMERCE? FROM EXPLORING TO THE 100% ELECTRONIC TRANSACTION, DARAS & DIESBACH (2002)

Second, a site may have an important role in the product or service marketing and commercial policy (Kapferer, 2002 ; Diesbach, 2003). Our second dimension will therefore take into consideration how deeply the site is integrated in the purchase process preparation and/or realization. On an extremity we may find pure content or image sites with no possible online purchase functionalities (eg. Sites like Shiseido.com for cosmetics, Hermes.com for luxury goods, or Traserfrance.com for diving watches, Porsche.com and Alfaromeo.com for sport cars); on the other extremity we may find purely selling sites, pure-players (exclusively online selling dedicated corporations) or online channel of multi-channel companies, such as Chateauonline.com or Idealwine.com for wines, Amazon.com for books, music and other consumer goods, for which on-line selling is really the key reason for existing.

We then propose a typology of merchant websites, that is not to be opposed to the 6 mono-functional categories of Hoffman & al. (1995), [The six proposed functional categories are: the online storefront, presence site, content site, mall, incentive site, and the search agent.] but that rather highlights a complementary, bi-dimensional typology (figure 2), along a two dimensional continuum. It consists of:

- the nature of the offered information, highlighting its affective vs its cognitive characteristic;

- the level of integration of the website into the purchase preparation or realization.

Towards a final, global definition of a merchant website: relation with the concept of service

The frontier between those different website missions in a merchant context may sometimes be rather problematic to underline. Internet may be seen as a mean of providing data, emotions, an experiential or social experience, a space for building a relationship with a brand, with a product/service provider, or with other customers. For example tourism sites such as Virtualtourist.com as well as Worldisaround.com provide information for preparing a travel, a real experience of discovery in itself (they are a content, i.e. an utilitarian, and also experiential websites). They also provide the option of informational and social, friendly contacts with other tourists who may answer questions, but even establish personal contacts with the user, as was tested. Therefore, if we consider the navigation on the site as a period of time in which the customer directly interacts with the firm or the provided service, through its capacity to deliver information and generate emotions in a marketing perspective, a merchant site also works like a service encounter in the meaning conceptualised by Bitner (1990), and deepened by Mohr & Bitner 1990, Bitner (1992) and Meuter & al. (2000).

For example let us consider another major field of application for internet uses such as banking. In 2003, 36% of the North Americans use internet for banking operations: in Europe the use rate varies from 11% (Italy) to Norwy (45%) at the end of 2002 (JDNet Solutions). Where to set up a frontier in term of website service functionalities, between service offerings such as accounts consulting, transfer or payment orders? Those banks sites may not be seen as merchant sites in the traditional meaning; they nevertheless allow a company to provide a (paid or free) service, and probably to increase customer loyalty (even for non internet-based operations if the customer prefers for any reason, a man-to-man contact for a number of decisions). Such sites then deliver services, participate into building loyalty, increasing operations (off-line and online), hence preparing a possible purchase, even for services the sites do not sell.

FIGURE 2

A TYPOLOGY OF MERCHANT WEBSITES ALONG TWO DIMENSIONS

In a traditional context of service encounter, space conception as well as social interactions are crucial for building the very service itself, building satisfaction and generating approach behaviours (Booms & Bitner, 1982; Mohr & Bitner, 1991; Bitner, 1992). We can logically assume that similarly in a service encounter online, design elements, particularly their social components, and more specifically such elements as virtual agents, will be of a crucial importance in creating approach behaviours, through the emotional reactions and the feeling of a social relationship.

As such the model of Bitner (1992) [Elaborated in particular integrating in a service encounter the framework proposed by Mehrabian et Russel (1974), and conceptualizing in such context a revised proposition of approaches & avoidance behaviours.] seems to be applicable to the online context: design factors and any elements participating to the social interaction, more specifically a virtual agent, should be of a major importance in constructing the service encounter and in driving customer approach behaviours. We therefore propose a new definition and will call "merchant website": "any site that may be used either as a shopping space in itself or a supporting space for a future shopping process (off-line), as an advertising/informational media, and/or as a service encounter". We study the role of virtual agents in such merchant sites, in such large understanding.

The relevance of online immersion: spatial and social immersions

Subjective immersion refers to a psychological state in which the individual using a communication medium (for instance, internet) fails to perceive or acknowledge the existence of the technology and/or the interface [Objective immersion has to be differentiated from subjective one. Objective immersion could be defined as "the degree to which a virtual environment submerges the perceptual system of the user in virtual stimuli" (Biocca, 1992, P. 10) and thus characterizes a property of the technology. The more the system captivates the senses and blocks out stimuli from the physical world, the more the system is considered immersive. An ideal immersive virtual reality is one where the totality of inputs to the participant=s senses is continually supplied by the computer-generated displays (Slater and Usoh, 1995). However, in our research, immersion has to be understood only in the sense of subjective immersion.] (ISPR, [ISPR is the International Society for Presence Research (http://www.temple.edu.ispr.index.htm] 2003). Two kinds of immersions may be experienced:

- spatial immersion, also called "telepresence",

- and social immersion called "social presence".

In being telepresent, first, individuals feel their presence in the virtual environment and no more (or less) in the immediate physical one (Steuer, 1992). Without developing this point we highlight the similarity of such concept with the one of Flow, proposed by Csikszentmihalyi (1996, 1975-2000b, 1990), and integrated in the model of consumer behaviour elaborated by Hoffman & Novak (1996).

Second, individuals respond to a virtual stimuli just as they would do in the (non "virtual") reality (Lombard and Ditton, 1997). For instance, they could have the impression of * really + manipulating a product in a virtual store and would thus adopt the same behaviors as in physical stores. On the other hand, if individuals could be self-included in a virtual environment in being spatially present in there (Witmer and Singer, 1998), they could also perceive this immersion when communicating with other beings (living or synthetic ones). Those in turn appear to also exist in the mediated world and seem to "really" react with individuals. The individual does not anymore differentiate between human-human interaction and human-virtual being interaction. The virtual "being" is perceived as part of the mediated world, and as being able to interact with humans, which leads to the creation of ("as if" social, or) para-social relationships (Lombard and Ditton, 1997; Donath, 2001a).

Spatial immersion and service encounter: a positive relationship?

The concept of telepresence becomes crucial for understanding the individuals’ behaviour in a mediated environment, such as Internet (Lombard and Ditton, 1997; Steuer, 1995) in the sense that this notion captures a global subjective experience. Moreover, if facilitating the experience of immersion is an explicit goalBfor instanceBin the entertaining sector (Cassel & al. 2000), it is also the case for commercial uses of such experience (e.g., a mall in a MUD’s [MUD: acronym for "Multi-User Dungeon".]).

In order to be immersed in a "space", an environmentBa real space as well as a virtual oneBthe user has to appropriate such space, i.e. to transform and personalize it (Belk, 1988; Boulaire and Mathieu, 2000; Car· and Cova, 2003). For example the empirical study by Cassel and colleagues (2000) is a good illustration of spatial appropriation by children in the case of games.

For doing so, three steps have to be followed (Car· and Cova, 2003):

(1) nesting: the user makes his nest in isolating a part of the space which results more familiar;

(2) exploring: the user locates more places and thus increases his territory;

(3) and, finally marking: the user personalizes the space in symbolically labelling it.

After these three steps, the individual has phenomenologically extended her/his self to this new space [This process is also close to the one that occurs when the user extends her/his self for integrating in it product and object (see the seminal work from Belk, 1988).] in controlling it (Belk, 1988).

Intuitively, consumers experimenting telepresence are more likely to consider such experience as entertaining, and sometimes very compelling (Green, 1998), which could lead to more exploration of the virtual space, that means more time spent in the virtual store (Lombart and Jeandrain, 2003; Diesbach 2003). The link between telepresence and the notion of service encounter should thus be obvious. The more the individual is telepresent on a website, the more s/he is likely to stay and to desire to be in contact with the company and/or the service, which is particularly interesting in considering the telepresence affective impacts on the consumer behaviour. For instance, Boulaire and Mathieu identify the concept of telepresence as one of the hedonic dimensions of affective engagement to a website (2000). E-surfing (i.e. to navigate for hedonic purposes without intention to buy online, Lombart and Jeandrain, 2003), often associated to immersion in a website (Helme-Guizon, 2001), is also linked to positive affect, and more specifically to surprise and arousal (Wolfinbarger and Gilly, 2001).

Social immersionBthe concept, virtual agents and service encounter: a positive relationship?

First let us approach the concept of virtual agent. A number of definitions have appeared, mainly in the literature in Artificial Intelligence (AI) and in Computing sciences (e.g., Blumberg 1996; Anastassakis, Panayio & Ritchings, 2001; Burgoon & al., 2000; Cooke & al. 2002; Cassel & Bickmore, 2000; Cassel & Nakano 2001; Bickmore 2003, etc.). A virtual agent is first, a piece of software, that performs tasks, with a varying level of autonomy (such concept would deserve in itself a number of pages to be explained). Most of the authors highlight some important specificities of the agent, such as: the function of supporting the interface (PC, website, etc.) user, anthropomorphic appearance i.e. visibility. Such visibility refers to the fact that the software is embodied in a mediated interface, e.g. a PC screen, in order to give the user the feeling e/she is interacting with a being, living or synthetic one. We focus on the cases of agents being used for an experiential, more social, and pleasant, navigation and therefore think the embodied agent is more relevant for us. A synthetic definition is therefore proposed for a new concept: an Internet Virtual Agent or IVA is a virtual agent, that is a piece of software, accessible via a website on internet, that is embodied (i.e. it is made visible), and that speaks to the user. It can make movements (e.g. illustrative movements), move inside the site, listen to the instructions of the user (written ones, clicks, or sometimes oral instructions), in order to perform a number of predetermined tasks that helps the user during his/her navigation."

An interesting literature in marketing shows that the nature and quality of the interaction with a sales or a contact person in a service encounter or an outlet, has a very important impact on consumers beliefs, on image formation (for the brand or the institution), and on attitude or behaviour (Mohr & Bitner, 1990; Bitner, 1992 ; Baker & al. 1992 ; Babin & al. 1994 ; Babin & al. 1999). For instance Baker & al. show that social factors impacts the service and merchandise perceived quality, and through them, the image of an outlet. Bitner (1992) proposes that the social interaction with a contact person impacts the consumer approach/avoidance (i.e. to stay, explore, come back, recommendation) behaviours.

It is suggested that most basic patterns of behaviour in a traditional merchant or service context may also exist on internet (Helme-Guizon 2001, Wind & Mahajan 2002), and therefore that we may find such kinds of behaviours on internet when a user is exposed to an IVA.

Do human being react in front of an agent as they would do in the reality?

We would certainly believe not. But actually a number of major findings in robotics, i.e. machine-man interaction, and in computing sciences, show that people largely behave as they would do with a human being when they interact with a virtual being. Reeve & Nass (1996) show that a human user attributes human qualities to an electronic tools he/she interacts with. Burgoon & al. (2000) even find out that he/she can happen to joke with it; they also show how interaction impacts the beliefs, i.e. the fact an IVA has a real power of persuasion, just as a human being. Donath (2001a, 2001b) shows that most of the modalities in human-human communication should be valid in human-virtual agent communication. Empirical results show how the interface is perceived as more attractive and persuasive when inhabited by an agent (Cassel & al. 2000; Takeuchi & Naito, 2002; Diesbach 2003). That is, the same way as it occurs with a man-man interaction, an IVA is likely to generate more (or less) involvement, positive (or negative) emotional reactions, which may in turn strongly impact stickiness or positive emotion + time spent (Cassel & al. 2000; Diesbach 2003) on the interface, and even influence the choice (Takashi & al. 2000; Bengtsson & al. 2001). Such phenomena, observed in a non merchant context, are very likely to also occur in a merchant contextBAmerchant" in the wide meaning previously defined. As a conclusion all those theoretical and empirical results lead us to consider as most likely that an IVA will impact the user’s emotional reaction, particularly his/her feeling of immersion, his/her beliefs and behaviours.

FIGURE 3

PROPOSED MODEL, THE IMPACT OF VIRTUAL AGENTS ON IMMERSION, APPROPRIATION AND APPROACH BEHAVIOURS

RESEARCH MODEL

Consistently with the proposed definition of a merchant website we have proposed, with the conception of a service encounter as proposed by several researches by Bitner, considering that a media, a service encounter or a virtual outlet may be merchant websites; keeping in mind the bi-dimensionality of the information that a site may provide, and its impact on consumer behaviour; last, applying such framework to the case of IVAs (internet virtua agents) the following research propositions are formulated (figure 3):

- P1: online navigation may generate an intense level of immersion;

- P2: emotional, particularly surprising or stimulating design elements of a website may reinforce such feeling of immersion;

- P3: a virtual agent is a particularly adapted case of design element (of a site) for enhancing feelings of immersion (in its two dimensions, that is, social presence, as well as tele-presence);

- P4: the feelings of immersion should positively influence the appropriation behaviours in a merchant context;

- P5: feelings of immersion should positively influence approach behaviours towards a merchant site (see hereafter), either directly (P5a) or via the appropriation behaviours (P5b);

- P6: the global design of a virtual agent (including its voice (audio design), non-verbal behaviour, movements, and not only its "shape" or visual fixed design) is likely to impact the positive reactions of the internet user;

- P7: Such emotional reactions in turn may act as moderators between the feelings of immersion and the appropriation;

- P8: in turn, the emotional reactions to the IVA may impact the feelings of immersion on the site (loop effects);

- P9: emotional reactions should influence approach/avoidance behaviours as conceptualised by Bitner (1992: i.e. stay, affiliate, return, recommend), as we adapt them to the Internet context, that is:

* The duration of the visit on the website;

* The intention to interact with the virtual agent (contact person) of the website;

* The intention to interact with a real contact person of the company represented by the website;

* The intention to return to the website and to recommend it to other users.

We summarize hereafter the expected relationships in a research model.

DISCUSSION

The objective of our paper was to demonstrate the relevance of immersion for understanding the consumer behaviour in an online, merchant context. First we have defined the concept of a merchant website, taking into account the actual existing sites, and observed online behaviours.

Second we have proposed and illustrated a typology of merchant websites that takes into account the nature and variety of provided information, and the degree of integration of the navigation on the site, into the final (possible) act of purchase.

Third we have clarified the concepts of immersion, and shown its relevance in an online context.

Last, the concept of virtual agent, and of internet virtual agent (IVA) was introduced, redefined, and specified to the internet, merchant (in the large adopted acceptation) context. All those concepts have then been related into a global framework. It is based on the model of consumer behavior proposed by Bitner (1992). Finallya model is proposed, together with a number of research propositions.

This paper helps clarifying the complex interplay between affect and consumer behaviour, and why/how observed results related to the presence of virtual agents might be explained. It also deepens into the conceptualisation of a merchant website, proposing a more rigorous and wider definition. Last, it proposes a typology of merchant websites that, we believe, helps better describing the multi-functionality of merchant sites as observed in the reality.

The next steps of our research will consist of testing, step by step, all the proposed hypothesis and research proposals.

BIBLIOGRAPHY

Alba, J., Lynch, J., Weitz, B., Janiszewski, Ch., Lutz, R., Sawyer, A., Wood, S. (1997), "Interactive Home Shopping, Journal of Marketing, 61(3), pp. 38B53.

Ariely, D. (2000), "Controlling the information flow: Effects on consumers’ decision-making and preference", Journal of Consumer Research, vol. 27, pp. 233-248.

Babin B., Boles J., Darden W. (1995), "Salesperson stereotypes, consumer emotions and their impact on information processing", Journal of the Academy of Marketing Science, 23, pp. 94-105

Babin, B., Darden, W., Griffin, M. (1994) "Work and/or fun: measuring hedonic and utilitarian shopping value", Journal of Consumer Research, March 1994.

Babin L., Babin B., Boles J. (1999), "The effects of consumer perceptions of the salesperson, product, dealer on purchase intentions", Journal of Retailing and Consumer Services, 6, pp. 91-97.

Baker J., Grewal D., Levy M. (1992) "An experimental approach to making retail store environmental decisions", Journal of Retailing, 68, 4, pp. 445-461.

Bengtsson B., Burgoon J., Cederberg C., Bonito J., Lundeberg M. (2001), "The impact of Anthropomorphic interfaces on influence, understanding, and credibility", 14pp., Journal of Consumer Research, 2001

Biocca F. (1992), "Virtual Reality Technology: A Tutorial", Journal of Communication, 42(4), 23-72

Bitner M.J. (1990) "Evaluating service encounters: The effects of physical surroundings and employee responses, Journal of Marketing, 54, pp. 69-92.

Bitner M.J. (1992) "Servicescapes: The impact of physical surroundings on customers and employees", Journal of Marketing, 58, pp.57-71, April 1992.

Burgoon, Bonito, Bengtsson, Cederberg, Lundeberg, Allspach (2000), Interactivity in human-computer interaction, Computers in Human Behavior, 16, pp. 553-574

Cassell J., Bickmore T., Campbell L., Vilhjalmsson H., Yan H., (2000) "More than just a pretty face: Conversational protocols and the affordances of embodiment", Working Paper, MIT, Medialab,20 pp.

Cassell J., Nakano Y., Bickmore T., Sidner C., Rich C. (2001), "Non-verbal cues for discourse structure", Working Paper, MIT, Medialab

Chen Q., Wells W. (1999), "Attitude toward the site", Journal of Advertising Research, pp. 27-37.

Childers T.L., Carr C. L., Peck J. C. (2001), "Hedonic and utilitarian motivations for online retail shopping behavior", Journal of Retailing, vol. 77, PP. 511-535.

Choi Y., Miracle G., Biocca F. (2001), "The effects of Anthropomorphic Agents on advertising effectiveness and the mediating role of presence", Journal of Interactive Advertising, 2, 1, 14pp.

Coupey, E. (2001), Marketing and the Interet: Conceptual Foundations, NJ: Prentice Hall

Csikszentmihalyi Mihaly, (1975-2000b), Beyond boredom and anxiety: The experience of play in works and games, SF: Jossey-Bass, 231 pp.

Csikszentmihalyi M. (1990), Flow: The psychology of optimal experience, NY: H. & Row

Csikszentmihalyi Mihaly, The costs and benefits of consuming, Journal of consumer Research, 27, pp. 267-272, Sept. 2000a

Diesbach Brice Pablo, (2003), Web3D 2003, 8^th International Conference on 3D Web Technology, Saint-Malo (FR), Virtual agents and stickiness online, proposition of an integrative model, 20 pp.

Diesbach Brice Pablo (2002), International Symposium on Atmospherics & Consumer Behavior, CERAM-Sofia-Antipolis, AFM, June 3rd, 2002, Nice (FR), Ambiance factors, emotions, and web user behavior: a model integrating an affective and symbolical approach, 19 pp.

Diesbach Brice Pablo (2001), Experiential E-commerce conference, Michigan State University, Sept. 27-29^th 2001, Ambiance factors and approach behavior in the e-commerce: a theoretical framework for capturing the effects of music & color, 14 pp.

Donath, J. (2001a), "Mediated faces", Proceeding of the CHI 2001, 18 pp.

Donath, J. (2001b), "Being real: Questions of Tele-Identity", in: The Robot in the Garden: Telerobotics & Telepistemology in the age of the Internet, by Ken Goldberg, MIT Press, 296-311

Ducoffe R. (1996), "Advertising value & advertising on the Web", Journal of Advertising Research, 21-35

Grigorovici, D. (2003), "Experiential product placements in Virtual Environments: Presence, information processing and advertising effectiveness", in the Presence panel of the Information Systems Division, International Communication Association Annual Conference, San Diego.

Helme-Guizon Agnes (2001) * Le comportement du consommateur sur un site marchand est-il fondamentalement diffTrent de son comportement en magasin ? +, RAM, 16, 3, 25-37

Hirschman E. (1980), Comprehending symbolic consumption: Three theoretical issues, In Symposium on Symbolic Consumption, NYU, ed. Morris Holbrook & Elisabeth Hirschman, pp. 4-6.

Hoffman, D. L, Novak, T.P. Chatterjee, P. (1995), "Commercial Scenarios for the Web: Opportunities and Challenges", Journal of Computer-Mediated Communication, 1(3)

Hoffman Donna, Novak Thomas (1996), Marketing in Hypermedia computer-mediated environments: Conceptual foundations, Journal of Marketing, 60, July, pp. 50-68

Holbrook M. (1986) "Emotion in the consumption experience: Toward a new model of the human consumer", In The Role of Affect in Consumer Behavior, by Peterson R., Hoyer W., Wilson W., 17-52.

Holbrook M. (1994), "The nature of customer value: An axiology of services in the consumption experience", pp. 21-71 ; in Service Quality: New directions in theory and practice, ed. Rust & Richard.

Jeandrain, A.C., Limbourg, Q. (2002), "On application of virtual environments to marketing", Proceedings of Virtual Reality Concept 2003 Symposium, Biarritz, pp. 325-332, France, Nov. 2003.

Lombard, M., Ditton, T. B. (1997), "At the heart of it all: The concept of presence", Journal of Computer-Mediated Communication, 3(2), http://www.ascusc.org/jcmc/vol3/issue2/lombard.html

Lombart, C., Jeandrain AC (2003) "Contrasting browsing in retail stores and e-surfing: Antecedents and consequences", The 30^th International Research Seminar in Marketing, Lalonde les Maures.

Lynch Patrick, Kent Robert, Srinivasan Srini (2001), The global Internet shopper: Evidence from shopping tasks in twelve countries, Journal of Advertising Research, pp. 15-23, (June).

Meuter M.L., Ostrom A.L., Roundtree R.I., Bitner M.J. (2000), "Self-service technologies: Understanding customer satisfaction with technology-based service encounters", Journal of Marketing, 64(3), pp. 50-64.

Novak, T.P., Hoffman, D.L., Yung, Y.F. (2000), "Measuring the Customer Experience in Online Environments: A Structural Modeling Approach", Marketing Science, 19(1), pp. 22B42.

Pham M., Cohen J., Pracejus J., Hughes D. (2001) "Affect-monitoring: deriving information from feelings", Journal of Consumer Research, 2001

Picard R. (1998), Affective computing, MIT Press, 292 pp., 1998

Pine II, B., Gilmore, J.H. (1999). The Experience Economy: Work Is Theatre & Every Business a Stage. Boston: Harvard Business School Press.

Potter S.M.(1986), Charting presence in virtual environments and its effects on performance, unpublished doctoral dissertation, Virginia Polytechnic Institute and State University.

Reeves, B., Nass, C. (1996), The media equation: How people treat computers, television, and new media like real people and places, Stanford, CA: CSLI Publications.

Slater, M., Usoh, M. (1995), "Modeling in Immersive Virtual Environments: A Case for the Science of VR", in Earnshaw, Vince & Jones (eds), Virtual Reality Applications, Academic Press, pp.53-70.

Steuer J. (1992), "Defining Virtual Reality: Dimensions Determining Telepresence", Journal of Communication, 42(4), pp.73-93.

Szymanski S., Hise R.(2000), "E-Satisfaction: Initial Examination",Journal of Retailing, 76, 3, 309-322

Takahashi Toru, Takeuchi Yugo, Katagiri Yasuhiro (2002), Working Paper, ATR Media Integration & Communications Research Laboratories, Kyoto, 4 pp.

Witmer, B. G., Singer, M. J. (1998), "Measuring presence in virtual environments: A presence questionnaire", in Presence: Teleoperators and Virtual Environments, 7(3), pp. 225-240.

----------------------------------------