Toward A Trustful B To B Relationship

Being trusted by the client has to be considered as an intangible asset that constitutes a considerable competitive advantage. To enjoy this advantage it becomes ineluctable to determine the trust’s antecedents. This survey tents to bring elements of answer for this problematic in an industrial context. The emphasis has been placed, in one side, on the importance of trust as a condition of successful relationship marketing, in the other side, on its conceptualization. Finally, and while distinguishing two targets of trust, the seller and his salesperson, an empiric verification close to electronic industrial clients, indicate that several variables influence trust’s development towards the seller firm and its salespersons and that it (trust) can be transferred from the latter to the former.


Introduction
Markets globalization's has motivated companies to establish long-term relationships (Anderson and Weitz, 1989;Bergadaa and al, 1998).However, all commercial exchange system cannot exist without trust that acts as strength that seals client and seller relationship and influence its long-term orientation (Bergadaa and al, 1998;Ganesan, 1994;Doney and Cannon, 1997).In this context, a better understanding of client -seller relationship should facilitate the development and the commercial efficiency between enterprises partners (Bergadaa and al, 1998).
Authors introduce the enterprise in the process of development of trust (Doney and Cannon, 1997) and they recognize that sellers have an important function to fill to facilitate trust development towards their firms (Bergadaa and al, 1998).Therefore, our problematic will be as follows: How trust to a seller firm is influenced by his characteristics as well as by his relationship with the client and how trust towards a salesperson is influenced by his characteristics as well as by his relationship with the client?While distinguishing, these two targets of trust (seller firm and salesperson), we will try in the following development to focus on the importance of trust as a condition of successful relationship marketing.Then and through our literature review based on Doney and Cannon (1997) research, we will present the conceptual model that identifies several variables that can influence trust development.Finally, the above model will be the object of empiric verification in a B to B context different to above authors' research field.

Why should we study the trust's concept?
In organization's relationship, trust is considered as an essential element that guarantee the two relationship parts success (Hawes and al, 1989).When a raised level of trust exists each part is convinced that the other is involved in their mutual success.This sensation is reinforced by the fact that their successes are often interdependent.A customer having trust is a customer who has propensity to increase his purchases (Doney and Cannon, 1997).These authors stipulate also that if customers and sellers trust each other, the degree of complexity of their negotiations would automatically decrease; what permits them to discuss some important topics for their mutual success rather than small details.
For trust in service context as well as to products brands, Sidershmukh and al (2002) and Gurviez (2002) confirmed that consumer's trust towards service provider and respectively to a product brand raise his fidelity.
The aforementioned reasoning permits us to assess trust importance in B to B, B to C and customer-products relationships, what brings us to dig more this concept that in spite of being studied in industrial context remains always a little bit ambiguous.

Defining Trust
The literature treating trust definition is abundant.Nevertheless, this concept remains often vague (Sahli-Abbes, 1999).Doney and Cannon (1997) define trust as "the perceived credibility and benevolence of a seller or his sales person": - The dimension of credibility refers to the expectation that a client can rely on seller's statements. - The dimension of benevolence refers to the consideration that a seller is interested on his client's benefits and welfare.

Towards a wider definition
Sahli-Abbes (1999) consider a wider definition of trust's concept imposes itself.Trusting a partner, means believing he is credible, benevolent, honest and reliable.Hence, the last author considers four different dimension of trust.

Seller reputation
The reputation has been defined as the public information concerning "the merit of trust" (trustworthiness) of an actor.The reputation of a society is considered like a key factor to create trust towards organizations, it has been well underlined by several authors for example by Ganesan (1994) and Doney and Cannon (1997).
In marketing literature, firm's reputation has been considered as the assessment aggregated of all firm's aspects, or the perception of the capacity of a firm to satisfy different parties' waiting.The reputation is as the central element of trust foundation when partners do not know themselves and they want to achieve a transaction.Thanks to its transfer through firms, a favourable reputation puts in value the seller's credibility (Ganesan, 1994) and encourages them to establish relationships.
While examining all these points of view, we notice that the pomposity is put on the nature of the reputation as an asset that is shared by different relationship parts, because they construct it and can destroy it as well.The aforementioned analysis permits us to formally state our first hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H1: Client trust influences positively seller's reputation.

Seller size
The seller firm size can be defined while underlying many indicators like the annual turnover, the number of employees, as well as seller's investments (Doney andCannon, 1997, Silem, 1994;Matri Ben Jemaa and al, 2006).
A seller's position on a marketplace is one of the reasons for partners' choice as well as for trustful relationship development (Doney and Cannon, 1997).In the same vein, Silem (1994) considered that firm's size as well as the capacity to make investments is positively correlated.Therefore, more sellers make investments less they will have opportunistic behaviour in order to keep their market position.
To confirm previous affirmations, Doney and Cannon (1997) proved empirically that more the size of the firm is big more the industrial purchasers consider that it deserves trust.Thus, we can consider seller size as a trust determinant (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H2: Client trust is positively associated to seller's size.

Seller and client relationship
Seller eagerness to make specific investments An investment is considered as particular or specific if it looses its value when the relationship is broken.These investments can be equipments, machines, facilities and even employee's particular expertise (Doney and Cannon, 1997).Granovetter (1995) consider that idiosyncratic investments convert an economic and transactional relationship to a social one.Thus, a certain dependence among exchange parties, corresponding to losses produced by the relationship eventual termination (Guibert, 1999).In the same context, Buvik and John (2000) consider that specific investments between clients and seller reduces transaction costs and endeavour trust development.Whereas, Anderson and Weitz (1992) suggest that it is implausible that specific investment existence cause opportunistic behaviour as this type of behaviour menaces relationship continuation.Thus, our third hypothesis stipulating the eagerness to make specific investments determine the client trust in a B to B context (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H3: The eagerness of the seller to make specific investments is positively associated to the client trust.
The sharing of confidential information The sharing of confidential information refers to sharing of private information that causes troubles to the supplier if revealed (Doney and Cannon, 1997).
Clients who are convinced that their seller do not deserve trust have tendency to behave on a non deserving trust manner, as for example to pass this information in the industrial surroundings.Sharing of confidential information generates also costs economy as it creates a certain synergy between exchange parties for the firm and therefore an increase of the outcome.This sharing confirms also that a supplier eager to share confidential information accepts to be in a vulnerable situation as the client can reveal it.Thus, it deserves trust.However, Pallas and Richard (1999) don't seem to be quite approving advantages of the confidential information sharing on the long run.In fact these authors stipulates that it is necessary consider that, on the long run, it exists an involuntary disclosure risk (or voluntary) of some confidential information by the client that could generate strong losses for the firm.Thus our fourth hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H4: Client trust is positively related to confidential information sharing by the seller.

Duration of relationship
The duration refers to the period of time seller and client have been in contact without considering the interaction frequencies.Dwyer and al (1987) affirm that through time, the experience with the seller increases and the couple client-seller converges on leaving behind deep disagreements.Such periods provide to the two parts a better understanding of each other what encourages the development of trust between them.In the same context, Ganesan (1994) demonstrated that the seller-client experience is positively related to the client perception of benevolence as well as seller's credibility (the two trust dimensions).This point of view is confirmed once more by the survey of Anderson and Weitz (1989) who found that trust of network members in manufacturers' increases with the duration of the relation.
In short, Doney and Cannon (1997) stipulate that the duration of time represent the investment that the two parts made in the relationship.These same authors affirm that most researchers are unanimous on the fact that trust is developed trough time.It brings us to give out our fifth hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H5: Client trust is positively related to the duration of time they have been in contact with the seller.

The Salesperson
The salesperson presents an important role for trust development.That is, his characteristics are very important.In the following, we will concentrate our development on the most common characteristics like expertise and capacity as well as the perceived likeability and perceived similarity by the client as done by Doney and Cannon (1997).Certainly, some promotional tools could also explain the trust to a certain extent but we will not deal with them in this study because we will concentrate only on seller and salesperson characteristics the most important factors that could explain trust development in B to B context (Doney and Cannon, 1997).

Expertise
Expertise can be defined as a perception of the other knowledge and competence (Ghoshes and al, 2001).
The role that plays salesperson expertise has received consideration from marketing literature.Indeed, several authors found that clients see the salesperson with more expertise to worth more trust (Doney and Cannon, 1997).
In the education area, Ghosh and al ( 2001) demonstrated that more the expertise of a college or training centre is better discerned more it would deserve trust of its potential customers (students).This point of view is supported also by Moorman and al (1993) that proved that expertise of researchers is a founding element of trust.Therefore, experts generate more trust and a positive attitude towards the ideas that they recommend (Matri Ben Jemaa and al, 2006).
In sale's relations, Crosby and al (1990) affirm that consumers count on their assessments of expertise discerned as a significant indicator of trust.In the industrial area, Michael and al (1987) affirm that to conquer his customer's trust it is imperative that salesperson becomes "customer expert" not only "product expert".In the same context, Doney and Cannon (1997) demonstrated that salesperson expertise determines clients trust in a B to B context.Thus the sixth hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H6: Client trust is determined by the expertise of the salesperson.

Capacity
Salesperson capacity is defined as the client belief that the salesperson is capable to fulfil his obligation as well as engagements consistently to previous agreements (Doney and Cannon, 1997).
The salesperson's capacity to fill his liabilities depends on the type of the relation he maintains with other actors, and more precisely of systems or networks in which he is inserted, it can be notably very important in the case of research agreements (Zaltman and Moorman, 1988).
Empirically, Moorman and Zaltzman (1992) found that when a researcher's capacity in an organization is raised, researcher's trust would be more important.Doney and Cannon (1997) sustain this idea and state that the capacity of a salesperson encourages trust development toward him.For example, a salesperson who promises to send an order could be doubted if the client supposes that this salesperson does not control the situation permitting his promise achievement (Doney and Cannon, 1997).Thus the seven's hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H7: Client trust is positively linked to seller's salesperson capacity.

Client and seller relationship
Likeability "Salesperson likeability refers to the client's evaluation that the salesperson of the seller firm is welcoming, good and agreeable" (Doney and Cannon, 1997), it is also the seller's capacity to adjust his behaviour so that a favourable interpersonal interaction can occur with a big variety of client's personalities.
Researchers in B to B areas, show that the trust sensation as well as maintain towards salespersons is determined by his likeability (Doney and Cannon, 1997).This point of view is also supported by Weitz and al (1986) who insisted on the notion of likeability while developing the notion of "adaptive selling" that designates the change of the salesperson behaviour during his interaction with the customer according to the situation of sale.This adaptation permits the development of trust towards these personals.
Finally, Rich and Smith (2000) demonstrated that more the salesperson is reactive more it will be easy for him to identify the social profile and features of his interlocutor.These authors go beyond, while demonstrating that the feminine salespersons is more capable to identify profiles or social styles better than the masculine personal and so they have a more elevated adaptation capacity what makes that they are discerned as being nicer.This observation could be a big contribution and could bring us to wonder if it is thus more discriminating to replace all salespersons by feminine personal.Thus our hypothesis number eight (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): The likeability of a salesperson determines clients trust.

Similarity
The similarity can be defined as the resemblance supposed between the source and the receptor of a message.The similarity is esteemed by the client's conviction that he shares the same interests as well as values with the salesperson.
Value sharing corresponds to the extent to which partners have some similar beliefs on the importance and the relevance of some behaviour, certain goals, and some procedures (Guibert, 1999).Johnson and Johnson (1972) stipulate that the similarity can be the motor that allows a part of a relationship to foresee that the other will facilitate the realization of his objectives and will encourage the institution of a relation based on trust.Dwyer and al (1987) confirmed that the shared values permit trust development; the same case is for Morgan and Hunt (1994).Hence, these last two authors stipulate that when exchange partners shares values, they will have more trust.This point of view is also supported by Donald and Brock (1997) who proved empirically that less there is similarity between two parts less will be the assessment of trustworthiness.Thus our ninth hypothesis (Donald and Brock, 1997;Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H9: Client trust is determined by the similarity of the salesperson.

Contact frequency with the salesperson
By contact, we designate interaction between seller and client.Therefore a frequent contact designates several interactions between these two parts.When the salesperson has frequent contacts with the client for a business or even other causes, trust can be constructed because the client sees the seller in different situations.What permits to know him more through different reactions this salesperson has (Doney and Cannon, 1997).Therefore, these authors stipulate that the frequent interactions encourage trust while allowing the client to understand better the salesperson to the extent to predict this personals future behaviour.
Empirically, and in insurances area Crosby and al (1990) found that a positive link exists among the contacts occurrence and the trust towards salespersons.
To the sense of Doney and Cannon (1997), the social interactions provide a casual environment that reinforces information's flue and help to construct inter-individual relations and encourage an enhanced mutual understanding and thus the establishment of trust.These authors stipulate that, the social and professional interactions can generate trust, because clients attribute benevolent intentions to the salesperson with which they share professional and social ties.Zaltman and Moorman (1988) reinforce more the above reasoning while insisting on the importance of interactions that permits the communal understanding of the two exchange parts (in their case of survey it was directors and researchers) through experiences.Therefore our tenth and eleventh hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H10: Salesperson frequency of business contact determines clients trust.
H11: Salesperson frequency of business contact determines clients trust.

Duration of relationship
By duration of relationship, we designate the number of years during which the client knew the salesperson.The construction of a long-standing relation becomes a strategically vital objective (Cravens,1995), and this thanks to advantages generated by this type of relation (long-standing relationship) as the development of mutual trust as well as added value (Wood,1995).
In the scientific researches area, Zaltzman and Moorman (1988) demonstrated that the absence of lasting relations between researchers and directors reduces the opportunity to establish an elevated level of trust although a research of good quality can be provided.These same authors affirm that one of the factors that improves the understanding between researchers and directors are the duration of the relation that is reflected by the number of common year of experience.What permits trust development while helping every part to understand constraints and difficulties undergone by the other.Besides, we notice that several empiric studies proposed to study the relation between the workforce stability (rate of salesperson's rotation) as well as trustworthiness (Donald and Brock, 1997) while concluding that it exists a strong opposing inter-relationship between the rate of rotation and trust towards this personal.
In the industry area, Doney and Cannon (1997) stipulate that the investment of salesperson in his relation with the client increases through time and since this investment will be put in danger following to a non-deserving trust behaviour, this personal would then try to be trust worthy.Accordingly, the duration of client-seller relationship should reinforce trust because it gives the insurance that the future conduct of the salesperson will be the projection of the current one (Doney and Cannon, 1997).This reasoning permits us to advance our twelfth hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H12: Client trust is determined by the duration of time he has been in contact with a salesperson.

Trust toward the seller and trust toward the salesperson
In B to B milieus clients suppose that the salesperson conduct replicate values as well as the strategies of the seller they are working for (Doney and Cannon, 1997).Sirdermukh and al (2002) share the same opinion, indeed they affirm that salespersons proceed according to the management board orders as well as indications.Therefore, as soon as a client does not know the seller, trust towards this latter could be developed on the assessment of salesperson trust deserving behaviour.Therefore, client trust towards salesperson is transferred to the seller firm (Doney and Cannon, 1997).
In spite of the fact that Doney and Cannon (1997) arrived to prove empirically that trust of salespersons can be conveyed to the seller and that seller's trust can be transmitted to the salesperson.For this research, we will focus our analysis on the first sense.First, because as we have the constraint to focalize on only one sense due to the statistical analysis and second because this sense seem to be more interesting than the second one (Pallas and Richard, 1999).
While doing reference to the here over development, we propose to verify empirically that trust towards salespersons is transferred to the seller.It will be the object of our thirteenth hypothesis (Doney and Cannon, 1997;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006): H13: Client trust is positively related to client trust in the salesperson.

Conceptual model presentation
This model has been adapted from Doney and Cannon (1997) research.Variables of the model are regrouped in four categories as follows: Insert Figure 1 here

Variables measure
Used scales have been seized from Doney and Cannon (1997) research.These two authors have generated multi-item scales based on measures used in previous research as well on interviews with the purchase personals from several firms.All variables have been measured while using a Likert on a seven points scale.

Sample description
The sample frame comprised 65 international high tech firms belonging to the electronic industry (implanted in Europe as well as in America).The entire sample was e-mailed a questionnaire and a letter inviting them to participate in this survey.A total of 47 questionnaires were returned dully filled (a 72% response rate).Respondents were basically male (78.7%), with an age of 30 to 40 years for 44.7% and a percentage of 36.2% for the interval 40 to 50 years.The level of instruction is a superior level for 78.7%, with an experience more than 5 years in purchases for 59.6% and seniority more than 5 years for 48.9%.
Firms to which belongs these respondents are generally public companies and make part essentially from automotive (38.3%), followed by Telecom (27.7%) and then electronics industrial sector (17%).This distribution puts the accent on the importance of trust due to the dynamism of these sectors (technological changes, demand fluctuation..).(Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Measure scale reliability
The alpha of Cronbach has been calculated for every scale, the gotten values provide a good reliability (between 0.64 and 0.94).

Hypotheses Validation
Hypotheses developed in this survey imply relations of reason to effect.To verify them we will use the analysis of regression.For this, we used following indicators: - Results are summarized as follows: Insert Table 1 here The regression analysis of the variable "reputation " and the variable " size " (independent variables) as well as the variable "Trust towards the seller " (dependent variable) appears as follows: The value of VIF provides a reasonable fit for the two variables.What proves that there is not an interrelationship between these independent variables.The value of F =12.828 was statistically significant with Sig =0.000 what provide a reasonable fit.This value permits us to affirm that it exists at least an explanatory variable in the model.Values of T are superior to 1.96 in the two cases with a lower significance to 0.05, this permit to advance that the two variables reputation and size of the seller are explanatory ones.That is, the two first hypotheses are confirmed.So reputation and seller's size are two antecedents of trust towards the last (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Characteristics of the seller firm relationship: (H3, H4 and H5)
Insert Table 2 here The analysis of regression between the abovementioned variables and the variable "Trust towards the seller " brings us to conclude that "Willingness to customize" as well as "Confidential information sharing" are two variables determining trust whereas the duration of the relation is not (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Characteristics of the salesperson: (H6 andH7)
Insert Table 3 here The analysis of regression shows that the hypothesis N° 6 is accepted and N°7 is rejected.

Relationship between trust towards seller and trust towards salesperson: (H13)
Insert Table 5 here The last hypothesis is confirmed.

Synthesis
The first result is the importance of reputation as well as the seller's size as determinants of trust.This result goes in the same sense as previous researches.
Regarding seller firm relationship characteristic's, the variables "Willingness to customize" and "Confidential information sharing" interferes like elements permitting trust development, whereas the duration of the relation with the seller has not been kept as trust determinant.Concerning salesperson it has been proved that his expertise, likeability as well as his frequent business contacts are some important elements to conquer purchaser's trust.Whereas his capacity, his similarity, the frequency of his social contacts as well as the duration of his relation with the purchaser don't have an effect on trust towards him.Finally, and concerning the transfer of trust generated by the salesperson to the seller, the result confirms the existence of a positive relationship between the first and the second variable confirming thus previous research results (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Consequently, the final model will be as follows: Insert Figure 2 here

Discussion
The results confirm the importance of the seller's characteristics (size, reputation).These variables appear to be important in different domains as many authors are unanimous on their effect on client's trust (Ganesan, 1994;Dasguptas, 1988;Anderson and Weitz, 1989..).It has also been demonstrated that the "Willingness to customize" is also an antecedent of trust.Indeed this kind of behaviour shows that the business seller is sincere what makes that he deserves trust (Ganesan, 1994;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
In accordance with the previous researches, private information sharing is an important variable in the determination of trust since it gives sign of the good faith and seller's benevolence (Doney and Cannon, 1997).We should also note that this variable can also acquire a high level of value for our sample since it is about " Hi-Tech Firms " where the dimension benevolence presents a crucial importance for trust development (Bell and al, 2002;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Contrary to certain researches but in accordance with the research of Doney and Cannon (1997), the duration of the salesperson and seller relationship appears not to have an effect on the development of trust towards the seller as well as towards his salesperson in the B to B context.This could be explained by the fact that the majority of studies confirming the importance of relation's duration have been elaborated for distribution channel area (Anderson and Weitz, 1989;Ganesan, 1994;Morgan and Hunt, 1994).Indeed and in comparison with distribution channel, the industrial markets have tendency not to accept elevated levels of dependence (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
For the salesperson and in accordance with previous research results, it appears that their expertise, their discerned likeability as well as the frequency of their business contacts constitutes keys variables to trust development.Indeed these three variables have been strongly solicited in several areas but the importance of the variable expertise could be in particular underlined in our case of survey since it is about high technology product.Concerning the variable likeability, it shows big importance and found a big theoretical support (Rotter, 1980;Swans and al, 1988;Hawes and al, 1989;Doney and Cannon, 1997).Finally, encouraging a mutual understanding, the business contact frequency between purchaser and salesperson permits the development of trust (Zaltman and Moorman, 1988;Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Besides, it has been demonstrated that other variables relative to salesperson as his capacity, his similarity, and frequency of his social contacts don't have an effect on trust development.Indeed researches in B to B didn't insist too much on these variables contrary to those in service areas (Crosby and al, 1990;Zaltman and Moorman, 1988;Moorman and al, 1992) that see that these salespersons play a key role.This explanation could be sustained also by the setting of our research where the industrial purchasers of a high technology product keep always as first objective problem resolution and honesty (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
In short, the last statement is the transfer of trust generated by salesperson to the firm.This report is compliant to the survey of Doney and Cannon (that are among the rare researchers to work on two targets of trust), from where the importance to differentiate between these two targets (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Theoretical implications
The above results go in the sense of a confirmation of the importance to determine trust antecedents in a client seller relationship.Some theoretical implications emerge (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
A First contribution is expressed by the confirmation of the variation of variables importance determining trust according to the context of survey.Indeed it was proved that the pomposity is put on salesperson's features as variable antecedent trust in the service context, whereas and due to high level of dependence in distribution channel the accent is rather put on the duration of the relation.In short, in B to B context it is rather the firm characteristics (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
A second contribution to the theory permit to confirm the adapted model of Doney and Cannon (1997) in a different research context and that trust has several determinants (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Managerial implications
First, results show that the industrial client give a lot of importance to the seller's characteristics (reputation and size), since while doing substantial investments, it is strongly unlikely that these sellers behave in an opportunistic manner.It appears thus important that this firm puts in relief these two aspects when meeting its customers (fair or visit), as well as on its advertising supports (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Second, it appears interesting to share confidential information, but with vigilance because it could be a double-edged weapon.In fact, as Pallas and Richard (1999) stipulates, it is important to take in account that with the duration it exists an involuntary (or voluntary) disclosure risk of some confidential information, what could generate strong losses for the firm (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Conclusion
This survey aimed to confirm trust antecedents for client seller relationships in a B to B context, for that the pomposity has been first put on this relationship characteristic, then we put in relief the importance of trust for these relations.
Finally and while having reference to our literature, a conceptual model (adapted from Doney and Cannon) has been presented on differentiating between two targets of trust: the seller and its salesperson (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Results showed that seller's characteristics (size and reputation), confidential information sharing, as well as his willingness to make some specific investments, play an important role for trust development towards him.It suggests that to develop and maintain a relation of trust with his customer, the seller must concentrate his efforts this way (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Regarding salesperson expertise, likeability, as well as formal contact frequency revealed a great importance, what emphasizes the importance of salesperson training either on behavioural level or on product characteristics (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Results also showed that trust generated by salespersons is transferable to the firm and it (trust) is developed differently towards the two targets.This result is very important because it shows that besides his assets (tangible or intangible), the seller can develop trust through his salesperson (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Therefore we conclude that beneficing of client's trust is to be considered as an intangible asset that constitutes a considerable competitive advantage (Bergadaa and al, 1998).To enjoy this advantage, the issue will not be only developing trust but also knowing how to maintain it.Hence, firms already beneficing of their clients trust should research continuously how to assure trust maintenance (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).

Limitations and further research
This work does not constitute a finished study.On the contrary, it only tries to suggest tracks and axes of improvement for enterprises.
First, the retained variables are not the only ones to contribute to trust development towards the industrial seller.Indeed, in spite of the integration of several variables determining trust, it is possible to identify other ones, as promotional tools used by the seller.Further research could perhaps study this aspect (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
Furthermore, the heterogeneity of our sample (enterprises implanted in Europe as well as in America) doesn't allow to generalize our results in an affirmed manner due to the cultural divergence.It would be thus interesting to underline this aspect while leading a comparative survey to study the impact of culture on clients trust development (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
In short, we must note the static character of our approach of trust.The choice of a transverse perspective, in spite of its advantages, doesn't permit to give idea about the dynamism of trust development process through time.The overtaking of this one-sided approach could be operated while doing longitudinal studies (Matri Ben Jemaa, 2003;Matri Ben Jemaa and al, 2006).
VIF : Factor of inflation of the variance -F : The test of Fisher -T : The test of Student 4.4.1 Characteristics of the seller firm: (H1 and H2)

Table 4 .
Results of the regression characteristics of the salesperson relationship and trust towards him

Table 5 .
Results of the regression trust towards the salesperson and trust towards seller