Memoire Online - Evolution du comportement du consommateur français dans l'e-commerce B2C

The perceived risk is a complex multidimensional concept. It is defined by the total risk which gathers the whole of the risks perceived by the individual.

As presented in the first part of the first paragraph, Cox (1967) was the first to detect various aspects of the risk : « The risk is not perceived solely compared to the purely economic or functional aspects of the purchase, but also compared to psychosocial satisfactions which result from this »^12(*).

At these first risks, authors will add some like Roselius (1971) with the waste of time or like Jacoby and Kaplan (1972) which define 5 dimensions of the perceived risk : economic or financial, functional or performance (principal dimension for these authors), physics, then dissociate psychological and social. Lastly, Zikmund and Scott (1973) add the risk of opportunity.

Thus, the risk of waste of time takes again the time spent for the purchase and repair in the event of problems. The economic or financial risk represents the risk that the money spent, for the product, is excessive compared to waitings of the consumer, that the same product perhaps less expensive elsewhere or that, when the product does not answer our waitings, the company the exchange or refund it. The risk of performance or functional evokes the quality of the product : it is bad ? You it answers waitings ? Which will be the losses in the event of repair ? The physical risk symbolize the dangers of the product for health, the environment... The psychological risk takes again the disappointment of the person compared to the product and the modification of its image in its environment. The social risk represent the impact of the decision of purchase of a product on the opinion and the judgment that the others have us and of our purchase. Which will be the projected image ? The risk of opportunity is the risk of loss associated with a purchase whose quality ratio/price evolves/moves quickly.

Stone and Grönhaug (1993) measured the importance of the various risks financial, social, of waste of time, of performance, physics and psychological. In spite of their variation according to situations' of purchases, financial and psychological dimensions dominate these 6 risks. These same 6 risks, explain 88% of the perceived risk (tests for the intention of purchase of micro computers).

The approach of dimensions of the perceived risk rests on the idea that the risk perceived by the consumer can be explained by a number limited and predetermined losses ; however, of the authors as Cases (2001) suggests that the number and the type of dimensions of the risk can be associated the context of purchase, underlining the possibility of dimensions of the risk which had not been identified yet by the literature. But especially, the 6 losses presented by Stone and Grönhaug (1993) does not capture all the variance of the total perceived risk.

In the research of dimensions of the perceived risk, the approach multiattribute thus came to be interfered by proposing the postulate which there as many exists dimensions as of determining attributes for a given product, in a situation of purchase given, for a segment of individuals given. Thus, the consumer perceives the risk on the level of each determining attribute of the product and with his associated characteristics.

For Stolen (1995), the approach multiattribute is « certainly more relevant than the approach starting from great dimensions ... no decomposition of the total risk is essential arbitrarily »^13(*). Moreover, according to Halvena and De Sarbo (1991), it is always possible to reduce the high number of attributes to some relevant axes.

However this approach is difficult to implement. Indeed, it is a question, for the consumer, of associating the number of attributes a number of factors and to perceive the risk on the level of each determining attribute. That is unrealistic in certain situations like that of the choice in limited time (Yates and Stone 1992). The representation of the risk is thus simpler when it is attached to each alternative in a total way.

The behaviors of the individuals are different vis-a-vis the risk. In order to observe more in-depth the behavior of the consumer vis-a-vis these dangers, we will propose to you measuring instruments of the risk at the individual and the tools which he uses in order to reduce these threats.

Vis-a-vis the risk, various features of personality coexist. We have the takers of risk (risk seeker) or éviteurs of risk (risk avoider). The scale of the most known measurement is the scale of « psychological tolerance with the risk » (Kogan and Wallaach 1964). Each individual evaluates the level of risk of a whole of situations. The individuals who evaluated a risked situation as not very risky are considered to be taking of risk and conversely for the individuals having evaluated like very risky a situation not very risky.

Woodside (1972) proposed a scale « taking risk ». Each individual must choose between 2 different scenarios, of which one is riskier than the other, but gets of advantage of satisfaction according to his level of risk.

« any action, on the initiative of the purchaser or the salesman, used as a strategy of resolution of risk »^14(*) (Roselius 1971). The reducer of risk is a consequence of the existence of the risk (Stem, Lamb, Mc Crokle 1990).

The reduction of the risk by the consumer is explained by the principle of Homeostasis. There is an optimal level of risk which the individuals are ready to tolerate to continue or not the purchase (Shanteau 1992), beyond this level of acceptable risk (Dowling and Staelin 1994), specific to each individual, balance homeostasic is broken, a cognitive imbalance is created, this imbalance must be restricted by the reduction of the risk.

Methods exist to reduce this risk ; they are justified by Cases (2003) « because a risk is perceived in more the share of the decisions of purchase, the consumers must then manage it and have methods to make their decisions in all confidence »^15(*). These methods, Olshavsky (1985) divided into 3 categories : the cognitive process interns with the search for information and comparisons, the process based on others with the councils, and finally the emotional process with the positive experiment which binds the consumer to the mark. These 3 processes then, will be more or less studied according to their availability (quality, quantity), the incurred risk and uncertainty.

According to Roselius (1971), the consumer can reduce the pre-decisional risk according to 4 strategies : to decrease the probability of failure of the purchase, to reduce the severity of the loss in the event of failure, to modify the type of loss by another for which it has more raised tolerance and to give up the purchase or to choose to buy and thus absorb the unsolved risk.

These strategies result in various techniques, for the consumer and the salesman, in order to make acceptable the perceived risk and to simplify the decision-making.

Bauer was the first to identify the couple quality/price (produced at high or cheap price), the influence of the group and fidelity with the mark like reducing methods of the risk. Then of many methods appeared :

A higher perception of the risk generates a search for more important information (Sheth and Venkatesan, 1968) and more active (Urbany, Dickinson and Wilkie, 1989). However, of the authors as Gemunden (1985) call into question the relation risks perceived high and search for more important information, by estimating that the search for information depends rather on the complexity of the decision of purchase. Simon (2000) also refutes this relation by affirming that this one is non-existent in several cases, when the degree of perception of the risk is lower than a minimal threshold, the cost of information is too important, it exists a risk of dissonance generated by additional information and finally, when a lack of confidence in the sources of information is real.

These 2 authors Bauer and Cox characterize commercial information (publicity) like fairly powerful reducing the risk.

- The purchase of mark and interpersonal communications (viral marketing) (Cunningham 1967).

- To propose tests or samples in order to develop the experiment of the consumer with the product (Akaah and Korgaonkar, 1988).

- The choice of the points of sale or modes of purchase (Prasad, 1975 ; Dash, Schiffman and Bergenson, 1976 ; Korgaonkar and Moschis, 1989). Indeed, some are regarded as riskier than others.

A limit is to be brought at the requests of reductions of risks of the individuals. Several studies (To place and Delquie, 1997), proved that the relation between perception of risk and request for reduction of risk is not always dependant. The consumers can take into account at the time of their decision of other factors that the existence of risk, to express their request for reduction of risk. Thus, they can for example, being interested by the potential success of the programme of reduction of risk without attaching an importance to the risk itself, or conversely, detecting a strong risk in some thing, but not to believe in the possible reduction of this risk.

In this first chapter, we will thus have insisted on the difficulty in clarifying the concept of risk and more precisely of perceived risk.

We analyzed the risk perceived in a traditional environment of purchase. Let us bring closer our problems by studying this risk perceived in a context of purchase on Internet, with a presentation of its evolution and its risks.

Confidence in the electronic trade continues to progress, passing from 23% in 2001 to 41% in 2003, however more half of the Net surfers thus do not have still confidence in the purchase in line (Médiamétrie, 2003). In order to include/understand and to better realize of the problems of the electronic trade, it is important to present its inventory of fixtures in 1998-99 and 2003-2004, for thus estimating the evolution. This study of the electronic trade will enable us to study ahead, the perceived types of risk and the means of reducing them.

In the years 1998, the potential of development of the E-trade seems very important, since the experts agree to think that it should represent between 5 and 10% of the sales turnover of the sales of detail in the years to come... Finally, at the beginning of 1999, it appears obvious that a full exploitation of this sector will not be possible that if the sites succeed in reassuring the potential consumers and to improve the contacts with them (ergonomics, speed, personalization, interactivity...).

What happenhappen does today, the sites they knew to develop and answer waitings of general public? ...

In these years of first steps of the Internet, We count, semi-1998 more than 3 million Net surfers^*, 6% of the population, with, at the end of 1998, 24% of the hearths having a microcomputer (Ministry for the Economy, Finances and Industry 1999).

* utilisateurs à titre privé et/ou professionnel s'étant connectés dans une période récente : 1 à 6 mois selon les études.

On the level of the request, according to a study of IDC (1998), 400 million francs (61.5 million euros) were spent on line for B to C. the E-trade does not represent whereas 0,05% of the sales turnover of the sales of detail.

Estimate of the incomes of the residential trade on the Internet in 1997 - 1998

On the level of the offer, one counts 54.801 French sites in July 1998, with 34.000 sites « .com ». (source AFTEL). Nevertheless, hardly 1000 of these sites are sites of sales on line.

Apart from well-known sectors like the cultural products, financial data processing and electronics, services, voyages... certain new products like the toys, flowers, greetings cards... find their market. New forms of exchange, such as the auction sales, emergent in a very important way.

Following the trends in the traditional sectors and new markets, one year later, the market B to C electronic commercial has more than triplet to arrive at 1.3 billion francs (200 d'€uros million), that is to say an increase in 227% compared to semi-1998 (Benchmark Group, March 2003 ; Fevad, 2003).

The 10 more important French merchants^*(*) realize more than 1 million francs of sale per month (155 000€).

The average audience of these sites is 6.000 visitors/month, with an average sales turnover monthly of 500.000 francs (75 000€) that is to say 400 francs (62€) per average basket.

Sales in line B to C in France by branch of industry (in francs and for year 99)

The purchasers on line, just like the sales turnover of the electronic trade, also knew a considerable growth. According to a survey carried out by TMO (1998 and 1999), 210000 French had bought on line in 1998, 800.000 at the end of 1999 and 2 million in September 2000.

In 1999, 73% of the Net surfers who buy on line are men, whereas the women account for 27% of the cyberconsommateurs (benchmark group, 1999).

According to the French organization, Médiamétrie, the bar of the 23 million Net surfers was crossed in March 2004. This threshold would have been reached thanks to the increasingly interesting offers tops flows and with largest numbers hearths equipped. In 2003, 10.5 million, is more than 50% of the hearths have a computer (Nielsen/NetRatings), which corresponds to 33.7% of French (ITU, 2003).

On these 10.5 million hearths, 6.8 million has an access Internet including 50% in high flow, that is to say an increase in 50% in 1 year (ministry for finances, 2004).

Since 1998, France strongly increased its number of Net surfers since it increased it almost 700% in 6 years. Nevertheless, in proportion of the population, it did not make up for its lost time.

On the level of the evolution, the annual increase in the number of Net surfers falls ; this one was 25% in 2001, 16% in 2002 and was only 10% for the year 2003.

On the level of the request, in 2003, 36% of the Net surfers are declared purchasing against in line 29% a year before (mediametry, 2003).

The barometer of the FEVAD, represented by 46 companies on line, shows that these companies increased on average by 55,2% their sales turnover compared to 2002 (540,66 million euros).

As comparison, the ACSEL (Association for the trade and the services on line) announced in the second quarter 2003, a growth over one year of 90% of the sales turnover of fourteen large commercial sites^*(*) (253,7 million euros) and of 91% of the number of the transactions (calculation carried out on seventeen commercial sites members of the panel^**(*)) compared to 2002.

In comparison at 1999 and the 202 million euros of Sales turnover, in 2003, France carries out 2390 million euros, that is to say an increase of more than 1000%.

In comparison with identical countries, France knows the greatest rate of increase in the number of purchasers on line, in spite of a delay always some.

The principal observation to be underlined is the delay of France in 2003-2004 in each field (Net surfers, cyberconsommateurs, high flow...), compared to other countries known as « similar » (Germany, Royaune-Plain) that, in spite of an important evolution, these last years, compared to the other country.

On the level of the equipment and number of Net surfers, France has at the beginning of this technological revolution, taken delay compared to the other countries. During years, certainly the number of computers and Net surfers strongly increased, but the increase was not enough strong to make up for lost time.

Today, the rates of increase in France are higher than the other countries, but these rates are not any more « spectacular » ; France will thus not make up for its lost time immediately, but in several years.

According to Claude Binette (1999), the delay of France in this new technology of the Internet is due to the Minitel.

Indeed, the French version of the electronic trade named a long time the Minitel. The figure d'affaires generated amounted of billion dollars, while the sales turnover of the Internet was limited to just a billion. As comparison, the electronic trade on the Minitel represented in 1998, 6 billion francs against 1.6 for the E-trade (BtoB and BtoC).

On the technical level, l'Internet is more powerful than the Minitel, because this one proposes much more possibilities. Perhaps however, the Minitel contains the keys of the success of the trade on Internet. Indeed, the material (investments) and data-processing knowledge required to use the Minitel are null ; on the level of the systems of payment, the Minitel also knew, to reassure the consumers.

If the minitel is one of the explanations of the delay which France as regards electronic trade knows, the importance of the risks perceived, on Internet, is also a negative factor. Let us study which are these risks...

In the field of the behavior of the consumer, the innovativity is the tendency of one individual to the innovation (Filser, 1994) or the predisposition to be informed on the new products and to adopt them (Goldsmith and Hofacker, 1991).

The adoption of an innovation by an individual depends, inter alia, of its degree of aversion with regard to the risk. The individuals downpours with the risks must multiply the repetitions of experiment in order to decrease the negative emotions. These experiments require sometimes knowledge who can be difficult to transfer according to sectors' ; the sector electronic commercial belongs to these fields which require a knowledge. Thus, the capacity of the consumer to control new technology is a fundamental element which influences the adoption of Internet like mode of trade (Leyzour, 1996).

The perceived brakes, by the individual, with the use of Internet are of 2 types : psychocognitifs and behavioral.

The aspects psychocognitifs are the innovativity in shopping, the self-confidence in the act of purchase, the implication with respect to the data processing and of the Internet, the estimated level of knowledge and the felt level of familiarity.

Behavioral aspects are the use of the computer and Internet, and the experiment of purchase on Internet.

Certain studies brought typologies of Net surfers according to their behavior or their attitude with respect to medium Internet and its commercial applications. Thus, the SOFRES classifies the Net surfers in three different groups. Among these groups, the group of « being wary ». They are individuals who did not buy yet on Internet and who would not intend to do it in the future. Their behavior would be explained by the fact why they regard Internet as « a space distant from them and not very sure » and also because they would test « more difficulty of using the computer and Internet that the average ».

This study perhaps supplemented by Médiamétrie, November 2003, which measured, near 9.000 French hearths, reasons of nonthe Net surfers not to be interested in the Internet. 1/3 of these not-Net surfers evokes the price of the material and of the conditions of too high access and ¼ evoke the lack of technical skills.

This innovativity and misses knowledge is however drops some, since 1/3 of reticent states to intend to use Internet soon. Moreover, the refractory core which does not see any utility to use the media on line reduces with the wire of time since they were 24% to think it it y six months and are 18% today. These 18% justify their position by the fact that they do not think of finding information interesting on Internet.

We will retain that there is thus well a difference in risk perceived between the levels of durable implication of the individuals with respect to data processing and of Internet : the individuals not familiarized with these tools test a more important perceived risk. The individuals having experience in the purchase on Internet have a sensitivity of perceived risk less. Thus, it is common of reading, in the studies published on the subject, that the experiment with the medium « play a part determining in the passage to the act of purchase », the experiment with data processing and Internet would prevent certain brakes.

Ernest and Young (1999) defined 13 factors which influence the purchase on Internet : flight of the information brought back to the CB, the sale of information concerning the purchaser by the salesman, the difficulty in finding the product sought, difficulties in seeking on the site of the salesman, confidence in the mark of the site, lack of tangibility of the products, the nature of the product, the price of the product, carriage costs, the possibility of visualizing the product on the site, the presence of depending information for the purchase on the site of the salesman, the number and variety of the products on the site, the bad description of the products and their use on the site.

Basing itself on these thirteen factors, Paraschiv and Zaharia (2000) the risks classified which influence the behavior of the consumer on Internet in seven classes : risk safety, risks intimacy, risk of waste of time, risks of mark of the site, risks related to the product, risks price and risks related to the contents of the site.

- the presence of depending information for the purchase on the site of the salesman

The study of Paraschiv and Zaharia (2000) shows that the first risk perceived by the consumers is the financial risk. It is about one of principal dimensions of the risk on Internet, with a strongly negative effect on the decision of purchase. It represents the principal fear of the Net surfers, but also of the tradesmen : interception of the data, transmission of wrong numbers of blue Charts...

In spite of fear that certain Net surfers always nourish with regard to the electronic payment by bank card, this one represents today the majority of the transactions on line with 65 to 85% of volumes.

However, the payment by check persists, two thirds of the twenty-eight principal merchants of the French Web propose this type of payment in order to reassure the Net surfers. In spite of the increasing familiarity of French with Internet or the efforts provided by the merchants to make safe the payment by bank card, the payment by check remains relatively stable with approximately 15% of flows.

This relative stability varies according to the type of sites and is thus explained by various factors.

For Télémarket, supermarket on Internet, it acts primarily of a question usually, « the customers paying by check remain faithful to this mode of payment in spite of a protected site » (Olivier Gargean, D.G.).

For sites of good bargains, this mode of payment is directly related to fear to pay on line by bank card and is the fact of people, in general little tested. Once familiarized with Internet, the Net surfers forsake the check, for the more flexible and less constraining bank card.

For the merchants of products high-tech when it is about raised amount, the use of the check book is a question of safety or limit of weekly expenditure fixed by the banks about the bank cards (Journaldunet, 2003).

Today, 39,5% of the Net surfers are declared trustful in the purchase in line (June 2003) ; to that, is added the encouraging prospects, since 1,2 million Net surfers not having never carried out purchases on line state to intend to cross the step in the six next months (vnunnet.fr). Thus let us study the actions which allowed or will allow, more still, the reduction of these perceived risks.

In order to reduce « numerical fracture » and to allure the reticent last, the government decided to set up a campaign aiming at promoting Internet (12/03). This countryside entitled « Internet declared of utility very public » has as an aim of « to promote the interest and the uses of microprocessing and the Internet, near the French little or not equipped, mainly, seniors, women, households with the modest incomes and least graduate ».

This countryside will surely have as a consequence a greater use of Internet in navigation, but not in the purchase. Indeed, contrary to the remarks of Nantel and Tellier (1992), which affirmed that the data-processing experiment decreased the level of risk towards the electronic purchase which the individual perceives, Salkin (1999) indicates that familiarity with Internet does not decrease the fear of the consumer, because tested are more conscious, that the beginners, of the risks and the limits of this type of trade.

Several research (To place and Delquie, 1997) proved that the relation between perception of risk and request for reduction of risk is not always in conformity. Paraschiv-Zaharia (2000) showed that this postulate does not apply for Internet. For this medium, there is well a relation between the request for reduction of risk and the perception of risk ; the greater the perceived risk is, the more the request of reduction of the risk is high. A strong correlation exists between the perception of risk on Internet and the request for reduction of this risk.

The study of Jacques Nantel (2001) presents methods of reduction of risk on Internet (Seybold and Marshak 1998, Forrester Research 1998).

First of all, in order to decrease the risk of waste of time and psychological, the site must develop 3 characteristics : identification of the site (corporate name, addresses...), identification of the customers concerned (consuming, undertaken,...) and the distinction enters the users (customers, employees, shareholders...). It is thus a question of facilitating the search for information characterizing the site and the salesman.

Thereafter, the search for information and the evaluation of the various choices (search engine, catalogs electronic, interactions between customers...) allows to reduce the risk of performance and waste of time.

At the time of the transaction, the risk of safety can be tiny room thanks to a procedure of single counter (regrouping of the purchases on a transaction), simple and supplements, offering to the customer all the confidentiality necessary.

Lastly, the site must propose an adequate customer service (fair with the questions, repertory of assistance...) to help the customer and to answer the behavior post-purchase.

Ancillary services must be proposed to the customers, like the possibility of configuring certain aspects of the site, in order to facilitate their decision-making process, to modify their personal information for purchases later, and finally, to propose the concept of community allowing the customers D `to exchange on the products or services. These assets can increase the implication of the consumer and thus his fidelity.

Van der Poel and Leunis (1999) also showed that the strategies of reduction of risk of the type satisfied or refunded and the reputation of the mark have more impact on the intentions of purchase of the consumers than the strategy of reduced price.

Lastly, the e-customization allows « to target the communication to help the consumer in his decisions of purchase by reducing the overload of information and making information highly suitable for the customer » (Ansari and Mela, 2003).

Since a few months, diffusion of the high flow (ADSL, cable...) allowed the reduction in the risk of waste of time and the increase in the comfort of use. According to Point-Topic, the number of subscribers high flow in France knew a growth of 50% in 2003.

Lastly, the certification of the sites makes it possible to reinforce the sedentary aspect of Internet, while offering the guarantee of a certain quality. The label attests conformity of a site, services or products which it proposes with preset requirements corresponding to waitings of the Net surfers. The aim in view is to give a better visibility to a site and the practices that this last applies in its relations with its customers. So the certification represents a commercial argument (Bernat, 2001). For a few months, one has observed a multitude of seals, logos, signs distinctive and other codes of good control.

The certification can be done by a French regulation or an independent third, setting up relatively relevant means to ensure itself of the reliability of the labellized sites. The European Commission currently thinks of a European label.

In order to limit the principal perceived risk, the financial risk, the companies worked on the guarantee of the transaction, the guarantee of safety with respect to the credit card (applications against the fraud) and the alternative option of payment.

The tools of safety are mainly « the lock » in the bar of state of the navigator and the address of the page with https:// instead of http://. These indicators reliable, since are posted by the navigator, unlike the assertions on the pages of the Web site.

Other sites, always in a concern of solving these risks, propose eight different types of payment : bank card, check, taking away and of the more sophisticated tools (account purchase, virtual credit,...).

The State, as for him, recently opened the first « Station control room of the electronic trade », decentralized unit of the Directorate-General of Competition, the Consumption and the Repression of Frauds (DGCCRF). This unit is charged to supervise the legality of the commercial transactions by Internet, to manage the mail sent by the consumers and the companies and to receive the complaints.

With the use of these various tools and practical, the Data-processing National Commission Report and Freedoms (07/03) congratulate the sites on electronic trade for the safety of the transactions and the personal data protection. The companies and the State thus became aware of the need well for investing in sedentary solutions in order to increase the field of the purchase on line. However, contrary to countries where the public discussion made rage, in France, these tools of safety set up, exist, but pass relatively unperceived in the spirits or on the site itself... (Bernat, 2001).

This 1ère left enabled us to clear up the theoretical aspect concerning the risk

Thus, we will have to underline the complexity of the risk with the perceived risk, the objective risk and the attitude towards the risk, its importance in the behavior of purchase. ; its various determinants according to the consumer, the product and the situation ; consequences on the level of its principal dimensions : temporal, financial, performance, physics, psychological and social ; and attitudes towards the risk according to individuals', with measurements and the tools for reduction.

On the level of the electronic trade, in spite of a rate of increase important and the development in the sites, high flow..., France marks a delay, in proportion of the population, compared to identical countries since 1998. The minitel plays an important part in this delay. Perceived risks, with mainly the financial risk, and the persistence of the not-Net surfers are also brakes to which the Web sites must face and propose solutions in order to develop the experiment of the Net surfer.

With an aim of analyzing the relevance of the various risks perceived in 2004, we will be delayed, in this 2nd part, with the presentation of the empirical study.

In this second part, we will present research ground carried out around our problems of the evolution of the behavior of purchase on Internet. We go first of all approached the methodological aspect with the choice of the installation of a questionnaire, then we will expose the object of this research, with the presentation of the assumptions.

A second phase, with this first part, will treat questionnaire with the aspect distribution, its contents and the measuring instruments used in order to check the assumptions.

Which are the large axes of our study ? What seeks you one to check? Such are some basic questions but necessary in order to clarify our research and to design a relevant tool of study. Thus, we will explain our bearing choice for a quantitative study, then we will detail our wills of research by presenting our assumptions.

The object of this study is to constitute an inventory of fixtures of the electronic trade with respect to the behavior of the Net surfer, in 2004, with their behavior and felt ; that by measuring the importance of the perceived risks which we will compare with 1999.

Theories exist to speak about the theory of the perceived risk, but no theory takes again an evolution of the risk perceived in the Internet. Indeed, considering the innovation of the field approached, the retreat is necessary over several years and appears only now.

To conclude this research and to gather many information, a study ground, of great scale is necessary.

My study will be thus inductive and quantitative with a determination of the model starting from many observations. My objective is to detect the essential properties of a ground of study, which is the risk perceived on Internet, and to observe the evolution of this felt. These observations grounds will enable me to establish a theory on the perceived risks and their evolutions.

To observe these evolutions, I will thus base myself on a study filed (questionnaire and analyzes 1999) and a questionnaire on the ground (2004).

This study of 1999 was carried out on 120 students, users of Internet. To 38% daily, 52% regularly and 10% seldom. On these questioned, 9% already bought on line.

The perception of the risks is classified in four classes, according to their rate of importance. The first class is consisted of the risk of safety, the second class of the risk of price, the third class of the risk of mark of the site and of the risk then the fourth class produces gathers the risk of contents of the site, intimacy and waste of time.

The request for reduction of these risks takes again the same priorities : with more the perceived risk is strong, with more the request of reduction of this risk will be strong.

On the level of the real cases of purchases, we observe various characteristics. The price has a great influence with respect to safety. Indeed, the low price is more important than the risk of safety. Questioned prefer to buy in store to pay less expensive but while spending an important search time on Internet. The people who prefer to buy on Internet perceive a risk of waste of time less important than the others. Questioned prefer to buy in the physical store of one « click and mortar » rather than to buy at this « click and mortar » on Internet and duty to pay forwarding costs.

To the level of the attitude towards the risk, more to the person downpour is said to the risk, with more it will perceive risks in the purchase on line.

For the sexes, the women test a perception of risk much more important than the men for all the risks, except the risk of safety where there is no difference between the sexes and for the risk of mark of the site where the men test a greater risk.

The women are also more sensitive to the price changes between traditional trade and trade on line to provide their number of CB and to buy on an unknown site.

This empirical study is divided into 2 stages, according to the studied population.

The first stage of the study is to observe the behavior of a population Net surfer with respect to the electronic trade in 2004.

The second stage is to compare the evolution of the behaviors of the Net surfers compared to 1999.

Which are the perceived risks most important, which makes that the Net surfer is refractory with the E-trade ?

1.1 -The spread of the risk is identical to 1999, with a distribution in 4 classes :

the risk of safety, the risk of price, the risk of mark of the site and the produced risk and finally the risk of contents of the site, intimacy and waste of time. The risk of service added in this study must be in the third class. (SHARE C)

It seems to to me interesting to draw the attention to one 8th perceived risk. Indeed, at the time of the purchase of a product, individuals will concentrate on the risks which even touch the purchase in him (information, comparisons, confidence, mark, choice...), then the additional transaction and finally costs (connection, computer...). At these perceived risks, the post-purchase risk seems to me also very present and forgotten in the preceding analyzes. This one takes again all the possible risks after the transaction : delivery (errors, time...) and customer services (assistance, councils, guarantees, refunding...). It is different from the risk of the site, mark or product, because it is about a risk which one does not check directly at the time of the decision of purchase at the moment T but at the moment t+1 (experiment with the site or the product). (SHARE F)

1.2 - Each risk is represented by several criteria. For each one of these risks, the most important criterion, in the importance of the risk, which is that compares more the electronic trade and physical trade.

This series of assumptions is based on the physical characteristics of the Net surfers. Indeed, it is interesting for the companies of the Net to be interested in which characterizes their refractories on the level of the sex, age and activity.

risk much more important than the men for all the risks, except the risk of safety where there is no difference between the sexes and for the risk of mark of the site where the men test a greater risk.

After having studied the perception of the risk on Internet according to physical characteristics' of the individual, it is relevant to correlate each factor with the mental characteristics of the Net surfer : reaction vis-a-vis the risk, experimentation and request for investments.

perception of the risk on Internet (nap of the factors). The individuals who say downpours to the risks have factors of measurement of the risk on Internet less important than the taking individuals of risks. (SHARE B-C)

3.2 - A strong relation exists between the perception of the risks on Internet and the experiment of use and purchase on Internet (category of individual). (SHARE A-C)

By gathering the factors which measure the experiment of the individual with Internet and by comparing them with the experiment of purchase on line, we obtain several categories of individuals. These categories are compared at each risk in order to determine the influence of the experiment of Internet or the purchase in the determination of the importance of the risks.

3.3 - The relation between the risk perceived by an individual and the request to reduce this perceived risk is not logical. (SHARE CD)

Before 2000, when the electronic trade yet were not very developed, the coefficient of correlation between the risk perceived by an individual and the request to reduce this perceived risk was strong (Zaharia and Paraschiv, 2000). This coefficient of correlation shows that more the perceived risk is great plus the request for reduction of risk is high. That is not always in conformity, indeed, they can believe in the impossibility of reduction of the risk and thus prefers to see investing the companies in a weaker risk but whose result will be convincing or to express their own risks and to take into account the general risks to define the investments necessary.

This relation risks perceive-request of reduction is different in 2004. Indeed, the companies invested much and the vision of the E-trade, by the individuals, changed.

After having studied the types of risks felt by the Net surfer, it is relevant to see whether these perceived risks are inherent in the product or not, bought on Internet.

4.1- The resentment of perceived risk is strongly different according to the type of product bought on Internet.

The questionnaire makes it possible to measure in a simple and effective way the preferences of the individuals.

We thus will present the method of distribution and the contents of the questionnaire. Then, we will present the statistical tools necessary to the analysis and the sum of the results as obtained thanks to a treatment on statistical software SPSS.

In a first phase, the questionnaire was tested near ten people. This pretest enabled us to modify the formulation of some questions or assertions. Thereafter, 2 modes of distributions were carried out.

This questionnaire was first of all distributed to 100 students, then distributed to 100 customers of a cinema, with a preference for the users of Internet and adult customers. This second mode of distribution enabled us to collect data with a very broad sample (level of education, age, occupation...).

This choice of 2 quite distinct groups is justified by the fact that these 2 groups profit from a different experiment with Internet. The first, throughout their studies, were formed with Internet which for them is not a new thing. The second population had to be formed with data processing and with Internet apart from the education system, it is indeed something again for them.

The design of the questionnaire is based on the questionnaire carried out by Paraschiv and Zaharia in 1999. The same structure was installation in order to facilitate the comparisons of evolutions. Questions/assertions were added or removed, in order to correspond with the Internet of 2004.

Parts A-B-G are parts with multiple choice gathering of information on questioned. Part A identifies its experiment with Internet : first contacts, frequency and purchases or not on the Web. The part B makes it possible to know its level of risk; with the risk (to gain 50€) or taker of risk is downpour (to gain 0 or 100€). The part G includes/understands general information on the individual: sex, age, dwelling, activity and income.

By gathering the factors which measure the experiment of the individual with Internet (questions 1 and 2), we obtain 2 categories of individuals :

Situation A. One offers a lottery ticket to you where you have 50% of chance to gain

For the treatment, the activity will be gathered in 2 classes : the workers (employee, liberal profession and different) and nonthe workers (student, applicant for work and inactive).

Parts C-D-E-F are parts with multiple choices on a semantic scale from 1 to 7 : of « unimportant » with « determinant ».

The part C questions on the factors which influence more the decision of purchase on Internet. The factors presented are based on the existing literature presented partly theoretical, except for factors 14 and 15. Indeed, 2 factors were added with 13 what exists, it acts of factors post-purchases which measure the risk of the customer compared to events post-payment, like the time, the customer service, guaranteed...

	Factor of influence of unspecified product-site purchase	Concerning my decisions of purchase on Internet
1	The risk of later use by others of your number of bank card	Unimportant Determining ? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
2	Confidence in the Web site of the salesman	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
3	The risk that, in the future, the salesman could sell information concerning you with other sites salesmen	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
4	The nature of the product (the product is perishable, fragile and/or requires a perilous transport)	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
5	The waste of time before finding the product sought on Internet	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
6	The absence of tangibility with the products before making the decision of purchase (possibility of touching...)	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
7	The good description of the products and their use on the site salesman	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
8	The possibility of visualizing the product on the site (photo, animation)	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
9	Facility to find information sought on the site (good organization of the site)	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
10	The presence of relevant information for the purchase (in terms of comparative prices) on the site salesman	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
11	The number and variety of the products available on the site salesman	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
12	The price of the product	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
13	The amount of the expenses of delivery	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
14	To propose a fast parcel delivery	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7
15	To propose a customer service (councils, guarantee, refunding...)	? ? ? ? ? ? ? ? ? ? ? ? ? 1 2 3 4 5 6 7

The treatment of this part is carried out while associating, according to the classification of Paraschiv-Zaharia (2000) the assertions with the risk to which they are referred.

The part D takes again a list of strategies of company which takes again the 8 perceived risks. It acts for questioned to note the rate of efforts which the companies should set up to support the sales on Internet.

The part E compares the reaction of questioned vis-a-vis the risks perceived between the purchase of a book and a portable computer.

Please specify, for each factor, the degree of risk which you feel, at the time of the purchase of each one of these 2 products.

Uncertainties concerning the stages after the sale (deliveries, contacts, assistance,...) :

The part F gathers 35 assertions concerning the purchase on Internet. Just like the part C, these assertions are gathered in 8 perceived risks.

	The PURCHASE ON Internet... WHEN I BUY ON Internet...	1 = Completely in dissension 7 = Completely of agreement
1	When one buys on Internet, one exposes oneself to more nasty surprises than of good surprises.	1	2	3	4	5	6	7
2	When one buys on Internet, there is the risk to be made steal and divert its number of bank card.	1	2	3	4	5	6	7
3	When one buys on Internet, one takes more risks than in store.	1	2	3	4	5	6	7
4	I believe that it is even surer to give its number of bank card per telephone or fax that on Internet.	1	2	3	4	5	6	7
5	With all that one hears around the electronic trade, I rely on this sector.	1	2	3	4	5	6	7
6	The brittleness of certain products makes their delivery difficult and consequently their very risky purchase on Internet.	1	2	3	4	5	6	7
7	The perishable character of the products is a major barrier to their purchase on Internet.	1	2	3	4	5	6	7
8	Impossibility of touching, of palpating, is a major barrier to the purchase of certain products on Internet.	1	2	3	4	5	6	7
9	When one buys on Internet, it is necessary to buy known marks if not one exposes oneself to nasty surprises.	1	2	3	4	5	6	7
10	When one buys on Internet, it is important to have already good experience of purchase of the product which interests us.	1	2	3	4	5	6	7
11	On Internet, there are many fraudulent sites (cashings without delivery, any postal address,...)	1	2	3	4	5	6	7
12	«Protected» Internet sites do not inspire really confidence to me.	1	2	3	4	5	6	7
13	On Internet, it is preferable to buy on a site known at a price higher than to buy on an unknown site at a weaker price.	1	2	3	4	5	6	7
14	When one buys on Internet, one exposes oneself so that the product that one will receive does not correspond to our waitings.	1	2	3	4	5	6	7
15	When one buys on Internet, the problems of delivery are important (times, costs, errors,...)	1	2	3	4	5	6	7
16	It is preferable to buy on an Internet site proposing an after-sales service, at a price more raised, than to buy on an Internet site without customer service, at a weaker price.	1	2	3	4	5	6	7
17	When one buys on Internet, the guarantees offered for the bought products are comparable with the guarantees of the same products bought in store.	1	2	3	4	5	6	7
18	When one buys on Internet, one does not profit from any customer service after his purchase.	1	2	3	4	5	6	7
19	When I buy on Internet, I have the fear not to know with which to inform me in the event of problem.	1	2	3	4	5	6	7
20	The transactions carried out on Internet can easily be observed by a third.	1	2	3	4	5	6	7
21	It is very pleasant to make its purchases on a personalized Internet site which takes account of its preferences.	1	2	3	4	5	6	7
22	I do not like the idea that somebody can perhaps record all the purchases which I make via Internet and thus learns how to know me without I knowing it.	1	2	3	4	5	6	7
23	After a purchase on Internet, one exposes oneself to receive many nondesired advertizing advertisements.	1	2	3	4	5	6	7
24	One saves time by making his purchases on Internet.	1	2	3	4	5	6	7
25	For the search for information, the contact with a salesman is more effective than to look on Internet.	1	2	3	4	5	6	7
26	On Internet, one often loses much time before finding the product required.	1	2	3	4	5	6	7
27	I lose much time to sail on the sites of sale.	1	2	3	4	5	6	7
28	I often find information relevant (in term of comparative prices) on Internet site of the salesman.	1	2	3	4	5	6	7
29	In general, on Internet, descriptions of products to be sold leave something to be desired.	1	2	3	4	5	6	7
30	In general, on Internet, the products are clearly visible by images and/or vidéos.	1	2	3	4	5	6	7
31	There are more choice of products on Internet that in the stores.	1	2	3	4	5	6	7
32	The price of the products on Internet (except additional costs : delivery...), are lower than in the traditional trade.	1	2	3	4	5	6	7
33	The price differences between Internet sites are larger than for the traditional trade.	1	2	3	4	5	6	7
34	Very included/understood, when I buy on Internet, the prices are, in general, higher than in the traditional trade.	1	2	3	4	5	6	7
35	Before buying on Internet, I would compare much the prices with the traditional trade.	1	2	3	4	5	6	7

Questions 5, 7, 21, 24, 28, 30, 31, 32 were re-coded in order to always obtain the 7= criteria apprehensive of Internet, 1= trustful in Internet.

1.1 -The spread of the risk is identical to 1999, with a distribution in 4 classes :

To check that, after having classified the risks by order of average, it is a question of a bivariée analysis, since a relation is measured between 2 quantitative variables. The test of Wilcoxon is carried out for these matched samples. A comparison with the results of the study of 1999 will be carried out.

1.2 - Each risk is represented by several criteria. For each one of these risks, the criterion which compares more electronic trade and physical trade prevails. (SHARE F)

Following the regrouping of the factors by risk of the part F of the questionnaire, the tool of regression enables us to obtain a classification of the assertions which determine more the given risk.

To check that, it is a question of an analysis Bi-varied between a qualitative variable with each time an average of quantitative variables. The samples being independent, the tests of Wilcoxon and Mann-Whitney are carried out. A comparison with the results of the study of 1999 will be carried out.

For the age, it is a question each time of analyzing a quantitative variable with an average of quantitative variables. It is thus about an analysis of linear Correlation of Pearson.

It is a question of gathering the activities in 2 classes (activity and non-activity), then of an analysis Bi-varied between a qualitative variable with each time an average of quantitative variables. The samples being independent, the tests of Wilcoxon and Mann-Whitney are carried out.

3.1 - A strong relation exists between the attitude vis-a-vis the risk (downpours or researchers) and

perception of the risk on Internet (nap of the factors). The individuals who say downpour to the risks have factors of measurement of the risk on Internet less important than the taking individuals of risks. (SHARE B-C)

It is about an analysis Bi-varied between a qualitative variable with each time an average of quantitative variables. The samples being independent, the tests of Wilcoxon and Mann-Whitney are carried out.

3.2 - A strong relation exists between the perception of the risks on Internet and the experiment of use and purchase on Internet (category of individual). (SHARE A-C)

After having gathered the individuals in 2 groups according to their experimentation of the medium, it is about an analysis Bi-varied between an average of quantitative variables and an average of qualitative variables the samples being independent, the tests of Wilcoxon and Mann-Whitney are carried out.

3.3 - The relation between the risk perceived by an individual and the request to reduce this perceived risk is not logical. (SHARE CD)

We have to make with 2 averages quantitative variables here. This assumption will thus be checked with a linear Correlation of Pearson.

4.1- The resentment of perceived risk is strongly different according to the type of product bought on Internet.

It is about a Bi-varied analysis of 2 quantitative variables, to analyze with the test of Wilcoxon for these matched samples.

By comparing the risks one by one, we obtain a perception of the risks which groups in six classes.

The first class is consisted of the risk of price and the risk of safety. Indeed, there is no significant difference between these 2 risks (Test of Wilcoxon ; Z= - 0.9 ; P=0.928).

The second class is consisted of the risk of mark of the site. Indeed, there is a significant difference between the risk of safety and the risk of mark of the site (Test of Wilcoxon ; Z= - 4.285; P=0).

The third class is consisted of the risk after purchase. Indeed, there is a significant difference between the risk of mark of the site and the risk after purchase (Test of Wilcoxon ; Z= - 1.704; P=0,088).

The fourth class is consisted of the risk of contents of the site. Indeed, there is a significant difference between the risk after purchase and the risk of contents of the site (Test of Wilcoxon ; Z= - 1.892; P=0,058).

The fifth class is consisted of the risks of products and intimacy. Indeed, there is a significant difference between the risk of contents of the site and the risk of product (Test of Wilcoxon ; Z= - 2.733 ; P=0.006) ; but there is no significant difference between the risks of product and intimacy (Test of Wilcoxon ; Z= - 0.362 ; P=0.717).

The sixth class is consisted of the risk of waste of time. Indeed, there is a significant difference between the risk of intimacy and the risk of waste of time (Test of Wilcoxon ; Z= - 2.733; P=0.006).

Assumption 1.2 was tested thanks to a regrouping of the factors, by risk, of the part F of the questionnaire and with the tool of regression. We thus could see, by risk, which characterized assertions more, the given risk.

The solution with assumption 2.1 is brought while observing that the women perceive only the risk of safety on a level significantly higher than the men (independent and paired samples).

Assumption 2.2 is tested by observing the Bi-varied correlation which is significantly relevant with the fact that the more the age increases, the more the individual perceives the risk of :

This correlation is also tested for assumption 3.2 by observing that the experiment with Internet does not seem to have of influence on the perception of the risks.

However, the experiment of purchase on Internet has an influence significantly lower for the risks of safety (Test of Mann-Whitney and Wilcoxon ; Z= - 2.994 ; P=0.003), of mark of the site (Test of Mann-Whitney and Wilcoxon ; Z= - 3.751 ; P=0.000) and of contents of the site (Test of Mann-Whitney and Wilcoxon ; Z= - 1.672 ; P=0.095).

Assumption 2.3 is solved by realizing that the workers perceive these risks in manner significantly higher than nonthe workers (independent and paired sample):

Assumption 3.1 is not confirmed, since no significant difference in perceived risks appears between the individuals who are taking of risk and the others.

The results of assumption 3.3 shows that, by analyzing the averages, the perception of the risks and the request for reduction of these risks are significantly correlated ; with, Correlation of Pearson= 0.87 ; P= 0.005.

The requests for improvements respect the classification of the following averages :

The results of assumption 4 shows, significantly, that the level of risk, for each risk, is always higher for the purchase of the computer than for the purchase of the book.

It is interesting to buy on Internet ? Without any doubt, yes. One finds in the virtual shops of the products difficult to unearth elsewhere, of the prices often less low... However, the fear of the risk remains a barrier to the purchase on line. According to a recent survey for the site lastminute.com, 1/3 of French having travelled used Internet for their voyage, with 70% which prepared their voyage on line and 30% which bought on line. There is thus well an unquestionable figure of users of Internet which any more but does not remain to be converted into on-line consumers ; is well there the principal problem...

Our theoretical part defined 3 determinants of the perceived risk : the consumer (sex, activity...), the product (price, mark...) and the situation (place of purchase, serviceable time...).

We of these determinants inspire to classify and analyze our observations. Thus, we go initially, to study the general perception of the perceived risks. Thereafter, let us study these determinants, more particularly, according to the consumer with its physical and mental characteristics, and according to the situation (with the activity synonymous with availability). Lastly, let us observe the perception of the risks according to characteristics' of the product.

Assumption 1.1 enabled us to be delayed on the classification of the perceived risks, as distinguished by questioned. We supposed that the classification was unchanged at 1999, in spite of the sum of the investments carried out in favor of the development of Internet and the electronic trade. That proves roughly exact. Indeed, the classification is almost identical to 1999. Only the inversion of the risks of site and product and the number of classes, which thus means the significant difference between the risks, change. Nevertheless, these changes are not relevant since they can be explained by the fact that the questioned population is widened compared to 1999, with employees and students.

We will thus retain, in a general way, the identical classification between 1999 and 2004.

Let us note the importance of the new founded risk, the post-purchase risk, meaning the importance with whole share of the delivery periods, errors, after-sales services, service customers, guaranteed...

In general, the risks are thus always identical in their setting ahead by the Net surfers. That can appear astonishing, taking into consideration investment and progression of the Internet. Thus, Internet sites propose gravitational prices today (sites discount), with the protected means of payment or the certificates justifying of their reliability. However, these 3 ideas gather the first 3 quoted risks.

Before putting the question to know if the investments of the companies were useless at the operational level, let us analyze the relevance of these risks, see whether they quite real or are just perceived...

In the theoretical part, we quoted Dandouau (2001) to define the objective risk while insisting on its difference with subjectivity of the perceived risk ; let us analyze if this difference is found in our study. Indeed, the 8 risks perceived on Internet are a true problem so much fears which they inspire are one of the factors inhibiting with the development of the electronic trade. Thus, many are still those which hesitate to make their purchases over Internet because of perceived risks... perceived since often evoked ill-advisedly...

In June 2002, EOS Gallup Europe made a study on problems in the purchase on line. In France, the results show that, 41% of the purchasers on line never encountered problem at the time of their transactions on the Internet.

Among the Net surfers having encountered problems, navigation on the commercial site arrives in first position with 24% of the causes of problems. The second problem, 22%, relate to use the not - authorized personal data. Being the delivery, 16% of the purchasers on line received with delay produces it ordered. The problems of communication with the after-sales service relate to 13% of probed. Publicity (or offers it) untrue comes in fifth position and relates to only 12% of the purchasers on line. The topic close to the lack of clearness in the price problematic for 6% to is probed. Only 4% were never delivered. On the other hand, contrary to the important apprehensions on this topic, the difficulties of payment appear rather minor (2% of probed), finally, only 1% of the purchasers could not turn over or not desired defective products.

The quantified table of Eurobaromètre (2002) shows the same results more or less :

Thus, one observes an enormous difference between the perceived risks and the real risks. The classifications are almost reversed...

The first reason called upon by the Net surfers is thus the fear of the payment in line (risk of safety). However, like presents it the figures and report/ratio 2003 of Yolin, the fact of diverting information on Internet is an existing, but completely negligible risk because of cryptology.

It is infinitely easier to use other easy ways to divert bank accounts : card readers of credit which make it possible to recover the data of a chart in the physical stores, purchases of lists of numbers of chart concealed with their validity date on Internet, generation of numbers of charts thanks to a software, hacking of files of tradesmen and access to the numbers of charts seized by fax, station, telephone or at the time of purchase in the physical shop...

It often happens that holders of CB see themselves output of an unknown sum, but, the majority of these people never bought on line. The true risk is thus to have a chart of payment, and not to use it on Internet. Even if the holder of a CB sees himself outputting of an unknown sum, on Internet, by flight of his number of CB, the banks must refund the output amount.

Lastly, we speak much about Blue the Carte risk, but the setting in number one of the financial risk can also be reduced by the payment by check, as presented in the determinants of the perceived risk of the first part.

It is thus obvious that this perceived risk is unjustified ; that, just like other risks, quite as unjustified...

Thus, the risk of mark of site was a reality to be supervised a few years ago with certain not very scrupulous sites. However, these small sites did not survive today vis-a-vis controls and the other more structured sites. For example, the station control room of the electronic trade, which emanates from the DGCCRF (Directorate-General of competition, the consumption and the repression of the frauds, ministry for the Economy), raised in 2001, out of 1000 sites « at the risk », only 30% of light anomalies and very few swindles. This a priori, of importance of the mark of the site, is due to sites like www.perenoel.fr which provided the media a long time (whimsical times to see null reception, errors of delivery...), leaving a negative image to the individuals who never tested this means of purchase by Internet.

Other risks like the not-reliability of information (risk of contents of the site), the produced risk or risks it intimacy are regulated by the law.

Thus, a commercial site which would provide inaccurate, obsolete or desynchronized information (produced, conditions of sale...) could undergo continuations and the consumer would profit from the problem of information (price, conditions...) always in its favor. In the same way, just like a remote sale, the law gives 7 days to the Net surfer at the time of the reception of its purchase to retract and return the product. Lastly, the risk of intimacy is protected by the Data-processing National Commission and Freedoms which regulates the information held by the company (right of glance, of correction...).

The commercial sites must work to improve the image of the provisions of services on the level safety, price, mark, post purchase... and thus to make rock these psychological problems. A continuous investment on behalf of the tradesmen on line, not at the operational level, but on the level of the psychology of the consumer is thus necessary by informing it and by educating it. That is confirmed with the study of the European Commission, 2003, on the level of the index of confidence with respect to the trade on line, France is classified only in 9th European position.

That can be explained by the fact of the tiresome delay taken as regards confidence of the Net surfers ; indeed, it is infinitely easier to sow the disorder in the spirits than to create a climate of confidence. The Net surfers thus more easily kept in mind the vexations of Internet, errors or frauds of sites, the years of premise of the trade on line. The stock exchange correlation of 2001, caused an important wave of not very solid and not very serious closings of these start-up. However, these sites left this image of insecurity on Internet. Some even wonder, if the on-mediation of these problems is not to put in the active those which wished to slow down the development of Internet to defend the cow with milk of the minitel. This on-mediation can also be allotted to the creators of the systems of payment, which succeeded in making take with Internet the responsibility for their own deficiencies as regards safety (Yolin, 2003).

Although 77% of the French sites offer clear information as regards safety, 47% do not propose this information in a sufficiently accessible way. France is nevertheless a leader in Europe by these figures.

If France is 9th as regards confidence and leader as regards sedentary information, it is although in spite of this information, the delay of France, on the level of the perception of the risks, is enormous. It is thus a question well of re-examining the ergonomics of the sites to put this safety initially ahead and not the products, that much more than the other countries. Only this solution can make it possible France to make up for its lost time as regards confidence in the purchase on line.

To continue in this analysis of risk perceived, it is interesting to analyze, in the list of the factors which brings to quantify a risk, one saw it, often in inaccurate manner, the most important factor to determine the coefficient of determination of the risk. Our observations can be divided into 2, with first of all a change of behavior of navigation on the fabric, gathering 4 factors and the investments to be realized for the 4 other factors.

First of all, at the level of the risk of price, this one is proposed with important price differences between Internet sites. The risk of price between Internet and the traditional trade are less. Thus, the multitude of the sites of trade, in particular the innovation of the sites discount, which are little in the traditional trade, allowed the Net surfers to gain in confidence price, in the purchase on Internet. Henceforth, the principal criterion of this risk is the doubt to buy on the site proposing the best report/ratio price quality. This doubt is due to this multitude of sites. Thus, the risk of important price in our classification of the risks, is more by the fact that the Net surfers compare much on Internet, rather than they estimate that the prices are more important than in the traditional trade.

The risk of contents of the site is mainly proposed by the positive factors of choice and visibility of the products. The lack of descriptions and information comes afterwards. Thus, the Net surfers appreciate the sum of the products present on Internet and their soft copy. These positive factors are quoted than the lack of textual presentation of the products. Indeed, the lines tops flows allow visual possibilities more and more (photographs, vidéos...).

On the level of the risk of product, the individuals judge in the 1st risk related to the specificity of the product, then judge the risk of choice. Initially the risk produced according to its characteristic comes (perishable, important to touch...) then the risk comes from negative experiment because of a bad choice (different from our waitings, to buy a product of mark...). Thus, the Net surfers test more risks to order a specific good like food (perishable) or an armchair (to be tested) that a good for which they are likely to be disappointed in the use (microwaves not powerful enough). This description, is explained by the spending pattern « physics », but also by the fact that the Net surfers inform themselves more on their purchases because of control of tool Internet and thus have less doubts about their purchases.

It is finally relevant to observe that the risk of waste of time lived by the Net surfers, is more in research of a particular product between several sites, that in navigation on a specific site. Thus, henceforth, the Net surfer enormously compares the products between several sites and thus spends more time in this type of research than on only one site.

The first part, of these internal analyzes of the classes, is explained by the fact why, in a general way, the Net surfers feel better formed that before with Internet (investigation journalDuNet, 2003). This greater knowledge of the Internet, enables them to have more confidence in the fact that few fraudulent sites exist (risk of mark) and to optimize their purchases (comparison of the characteristics produced, of the prices, takes care of the promotional operations) with gravitational sites. That explains the risk of prices perceived between the sites. This greater ease of navigation leads them to compare a sum of products, photographs... (risk of contents of the site), which enables them to be surer of their purchase (produced risk) and which brings them to more sailing between the sites and thus to spend more time.

For the risk of safety, the individuals underline more, the difference in safety with a physical store which they test. Thus, on the level of safety, they prefer to buy in store and to give their number of CB per telephone or fax that on Internet. The experiments and testimonys which they could hear affect less their determination for this risk. That explains the persistent importance of the risk of safety. The individuals remain persuaded of their fears for this risk, in spite of the communication carried out to reduce this risk and the positive experiments of their entourage.

The risk of mark is quantified especially because of the lack of psychology concerning the protected sites. Indeed, the fact that protected sites « do not inspire really confidence » to the users is important. On the other hand the fact that which there exists still much of fraudulent sites is less. Thus, a 1ère stage in « education » of the consumer was realized by the sites and the media by showing that the fraudulent sites were negligible and by giving the means of differing from these sites (labels, certificates...). However, a doubt exists, now, on total confidence to bring to these sites known as protected. A 2nde stage is thus to bring on the confidence of the Net surfers while making clearer these tools of safety (visibility, explanation...).

The after-sales risk is proposed with the fact that after the sale, no help is brought for the use of the product. The problems involved in the guarantee and the delivery are secondary. An investment of the companies is thus well to realize on this connection with the customer : customer service...

In the risk of intimacy, the Net surfers consider more the aspect negative that somebody knows them without they knowing it, and summons it nondesired advertizing advertisements, rather than the positive aspect which is to see itself recording its preferences and thus to profit from a personalized site. As we will see it later, in the analysis of the risks according to the age, it is well a problem of nonknowledge of the person who knows their preferences, since the consumer always likes that the physical tradesman knows his preferences.

The last 4 risks represented show the importance of the convictions of the Net surfers vis-a-vis testimonys and communication, their nonconfidence in the sites made safe, the lack of connection with the customer after the purchase and the aspect of intimacy lived like a monitoring and not a facility of navigation. These factors are thus many courses which the E-trade can cross while investing in a better presentation of the tools of security and a better relation (presentation of the identity of the company, co-ordinates...).

Assumption 2.1 was based on the results of the study of 1999 and affirmed that for 5 of the 7 risks presented, the women tested a risk more important than the men. Only the risks of mark, superiors for the women, and of safety, identical to the men, were distinguished from this observation.

In 2004, a stabilization of the risks between men and women really exist. Indeed, 7 of the 8 risks do not present significant distinctions between men and women. Only the high risk of safety is specific to the women.

Two points are thus distinguished from this analysis, the fact that we attend a fall of the inequalities of the risks perceived between the sexes and the fact that the variation of risk of safety increased for the women.

In 1999, the inequality of the risks perceived on the level of the sexes, profited from an explanation of use : 30% of the users of Internet were women (Novatris Source). In 2004, this figure east passed to 45% (Source FEVAD).

This reduction of inequality of men use/women thus has directly an influence on a balancing of the perceived risks.

This variation of use was explained by the complexity of Internet and the weak presence of Internet in the hearths. First of all, at the behavioral level, the men profit from a spontaneous aspect for new technologies, but Internet was not presented in gravitational a enough way for the women but well in a too technical way. Moreover, the formation with this medium was carried out in the companies, but this access to the training of Internet was less for the women taking into consideration level of occupation of the 2 sexes : 50% for the women and 63% for the men. Lastly, the more unquestionable taste of the men for the unknown and the risk, were also a factor determining in their attraction for the use of Internet (Ministry for research and new technologies, 2003).

Very quickly, as Cloutier (1999) underlines it, the companies wished to reduce this fracture and to allure these customers, for economic reasons, since principal in the traditional trade. Thus, Internet was submerged by a wave of creation of sites dedicated to the woman since 1999. The Internet in residence and the development of the use by the children of this medium also were of the accelerating factors of the use of Internet by the women, and thus of a greater confidence in this medium.

The second point of this analysis of the sexes, relates to a rate of risk perceived, on the level of safety on Internet, significantly higher at the women than at the men. This over-exposure of the risk of female safety can be explained by a female psychology of « less self-confidence ». Thus, the study of the ministry site « a feeling of self-confidence and weaker economic safety » at the women that at the men. With that is added, the ignorance of the technical aspects of safety employed for crypter and made safe the transactions, 41% of the questioned women are said completely exceeded by new technologies against 33% for the men. This percentage is significant ignorance of Internet and its safety measures used.

Assumption 2.2 affirmed that for all the risks, the age was an important factor of distinction ; thus at more the age is raised, with more the perceived risk is strong.

Our analysis shows that that is true only for 3 risks : the risk of waste of time, the risk of product and the risk of intimacy. These 3 risks thus gather the topics of the training of navigation on Internet (risk of time) and usually of consumption (risk of product and intimacy). None of the 5 other risks belongs to these topics.

Thus, the study of Kaplan (2000) explains the prevalence of control of these 2 topics at young people.

The young people always lived with Internet. It is first of all at the school that the majority of the young people discovered Internet.

Thereafter, the students unconsciously interiorized the Internet which influences the whole of their behaviors of purchases. They became active consumers, in the search of convenience and good bargains, which fully use the tools at their disposal.

The young credits adapt the Internet to their practices of preliminary purchases. In example, these young people perceive the online advertizing and off line who relate to Web sites like a publicity with whole share of consumption ; 86% of the young Net surfers already clicked on an advertizing stringcourse whereas only 43% of the adults.

This part of the population, very sensitive to new technologies and innovations, would have thus acquired experiment on line and would be smelled at ease to carry out purchases on line, which is for them natural. They thus do not test a waste of time and to buy on the fabric belongs to their spending pattern, without risk of tangibility or negative aspect not not to better know the site, which knows us perfectly to him.

For the people paid, their activity makes that they do not have as many time as the young people to sail on Internet. The sum of the sites, of the received malls... pushes this category to propose this risk. This risk is also perceived by the seniors owing to the fact that they do not control tool Internet. The old people have to learn the use of Internet by themselves or their family, but to in no case by professionals at the school or in company, that can thus explain the difficulties of navigation and the waste of time met.

The risks of intimacy and product make parties of the relation and spending patterns with the salesman of the older people ; they like to go in their merchant. These people are the first to like that their usual tradesmen know their practices, however, the fact that it do not have there a human contact with the person who knows their preference is the cause of this factor of intimacy. To go in their merchant to profit from personal information, of a human contact, to touch the product... is also of primary importance for them. It would be even about a spending pattern synonymous with pleasure which the young people do not need.

Today, the appropriation of Internet mainly remains the fact of individuals of less than 49 years (82,1% of the Net surfers) and especially of the young people of less than 35 years. According to Médiamétrie, the 11-34 years represented, at the end of 2002, nearly 57% of the population Net surfer, whereas their weight in the French population is only 36,3%. The old Net surfers of 50 and more, on the other hand, under are represented, this phenomenon accelerating as from 65 years. The latter account for only 3,2% of the Net surfers, whereas their weight in the French population is 18,9%. However, just like for the women, a few years ago, with the family pressures, the development of the accessibility and of the sites for the credits or the seniors, an increase in this population starts to be felt. Thus, an ocean carrying trade and acquisition of experiment will be acquired by the people at the age more advanced. It would not be astonishing that in a few years, the risks between young people, adults and more than 50 years are balanced.

Assumption 2.3 affirmed that for all the criteria of risk, those were more important for the people having a wage activity compared to the individuals not having wage activities, except for risk of price because of their income.

Our results show that our assumption was false with a balance for the risks, except for the risks of intimacy, and mark waste of time of the site where an important variation exists between our 2 populations.

We will analyze each risk judged like superior, by classifying them in 2 categories, the over-exposure of the risks of intimacy and waste of time because of their work, and the presence of the risk of mark, bound with nonthe presence of the risk of price, because of their income.

The fact of carrying on an activity, on the level of the monitoring and time of the employees explains the risk of intimacy and waste of time.

The risk of intimacy takes again the fact that the salesman can sell with other sites information which it recorded concerning the individual, thanks to questionnaires or more technically thanks to the informers (cookies, spyware...) installed automatically on the PC. 2 axes are thus released from this definition of risk, being observed and tracked on Internet and to receive a sum of information nondesired by thirds.

This fact of being observed and of being tracked is very present on the places of work. Like the recall the site www.droitdunet.com, it is normal that the employer has a right of monitoring on the employee, with an aim of guaranteeing the integrity of the information processing system of his company, of avoiding an illicit or faulty use of Internet and of taking care of the productivity of his teams, finally, it can get informed about the professional malls or files receipts by its employees. Thus, the employer has the right to install a data-processing monitoring in order to control the activity of paid and to regulate the use of the electronic messaging. The company must simply inform the employee of it. It is to him however interdict to control any mall or personal file.

Thus, the use of the working station at personal ends perhaps a cause of legitimate dismissal recognized by the courts. According to an investigation of KLegal-Personnel Today Magazine (2002), in the British companies, there are more sanctions related to the abusive use of Internet than of traditional sanctions (violence, dishonesty, non-observance of the safety requirements...). The British study shows thus that 40 % of the questioned companies at least once a month supervise the use of the connection Internet of their employees.

In France, Benchmark Group (2002), publishes a study which shows the extent of the personal use of connection Internet on the place of work. Close to a French employee on four, having a connection Internet to the office, passes more than one hour per day to use this access for a personal use. They are one on two to use it more than one half an hour.

Thus, the fear of observation on the place of work can explain the bond carried out between intimacy with work and intimacy on the sites. The employee, even in his residence, can fear the fact of feeling to observe, as with his work ; the non wage-earning person, does not know to him this feeling of monitoring to work and thus thinks of it less at the time of its connection. The fact that the law the labor does not regulate the conditions of use of data processing to work involves a blur on this cyber-monitoring for the employee and the employee which does not know its rights. The same blur exists for the Net surfer who does not know his rights (CNIL...).

In addition to this risk of intimacy, it is obvious that, because of their activity, the paid people have less time to devote to activities Internet than the others (student, inactive...). This observation is also measured owing to the fact that a majority of questioned are students and thus control more tool Internet.

This prevalence of the risk of mark is certainly most interesting to note. Indeed, with their high income compared to the individuals not profiting from wages, these Net surfers prefer to buy on a known site that on an ignored site but certainly less expensive, and thus to profit from a safety of the site and to reduce the perceived risk.

It is interesting to see that the risk of price was however not quoted in a higher way by the non wage-earning persons. Thus, the individuals are rational and the price is not considered to be more or less important whatever the income ; the consumer seeks to pay less expensive for 2 identical products, some is his income. On the other hand, the individuals with high incomes will be tempted to pay expensive for a better service.

On assumption 3.1, we supposed that the attitude vis-a-vis the risk and the perception of the risk on Internet were bound, with more the person said downpour to the risk, with more it perceived risks in the purchase on line. If that proves actually shown, this conclusion would bring an explanation to the differences in use of Internet and a classification of the individuals could be carried out by the companies with an aim of segmenting the population.

* ¹²Cox D.F. (1967), «Risk Handling in Consuming Behavior: Year Intensive Study off Two Boxes», Risk Handling in Consuming Behavior, Cox ED., Boston, Harvard University Near, p.34-81.

* ¹³ Stolen P. (1995), « The concept of risky perceived in psychology of the consumer: antecedents and theoretical statute » Research and Applications in Marketing, vol.10, n° 1, p.40-56.

* ¹⁴ Roselius T. (1971), «To off consume Rankings Risk Methods Reduction», Newspaper off Marketing, vol.36, N°1, April, 1, p.56-61.

* ¹⁵ Boxes HAVE (2003), « The risk perceived, a mediating variable in the process of reduction of the risk on Internet «, Publication AFM, « To perceive, identify and manage the risks in marketing «, p.137-153.

* * It acts by order of Alapage (Produced cultural), Al (Billeterie), Degriftour (Voyages), LED (Data-processing), Fnac (Produced cultural), Gateway (Data-processing), Spring (General practitioner), Redoute (General practitioner), Nouvelles borders (Voyages) and the SNCF (Voyages).

* * Watercolour, Chapitre.com, Chateauonline, Redoute, Mistergooddeal, Nomatica, Ooshop, Paniers.com, Pixmania, Priceminister, Surcouf, 3 Suisses.com, Voyages-sncf.com, Welcomeoffice

* ^** Alapage, Watercolour, Chapitre.com, Chateauonline, e-Bay, Fnac.com, Redoute, Mistergooddeal, Nomatica, Ooshop, Paniers.com, Pixmania, Priceminister, Surcouf, 3 Suisses.com, Voyages-sncf.com, Welcomeoffice

Evolution du comportement du consommateur français dans l'e-commerce B2C

* not developed previously 2.2 dimensions in the literature

If the minitel is one of the explanations of the delay which France as regards electronic trade knows, the importance of the risks perceived, on Internet, is also a negative factor. Let us study which are these risks...

Initiating label