Measuring the acceptance of electronic marketplaces: A study based on a used-car trading site

Authors

  • Tobias Kollmann

    Corresponding author
    1. Studied economics at the universities of Bonn and Trier (Germany), specializing in marketing. He received a Ph.D. in marketing from the University of Trier after completing his thesis on “The Acceptance of Telecommunication and Multimedia Systems”. His research interests include electronic marketplaces, customer behavior, one-to-one marketing and virtual transactions. He is a lecturer at the International University College Bad Honnef, Bonn (Germany) on the subject of electronic marketplaces.
    Search for more papers by this author

Address: Agrippinenstr. 3, 53115 Bonn, Germany

Abstract

Electronic marketplaces act as intermediaries between supply and demand. Such forums are often organized by a central operator and are a common form of electronic commerce. The operators are independent entrepreneurs who are competing against one another for business. Their objective is to win over users for their own marketplace in order to generate fees by matching buyers and sellers. This is only possible when the marketplace design is ‘acceptable’ to the users. Using data from an empirical study, the present article analyzes criteria which marketplace users considered important in determining their acceptance of an electronic marketplace. The data were collected from the users (buyers and sellers) of a used-car trading site.

Introduction

The role of electronic marketplaces is to act as an intermediary between supply and demand (Sarkar, Butler & Steinfield, 1996; Schmid, 1993). Using a simplified classification, it is possible to distinguish between two general types of marketplaces on the World Wide Web (Bailey/Bakos, 1997): markets with and markets without a central operator who brokers the transactions between buyer and seller. Markets without a central operator are merely electronic trading forums. They feature a thematic listing (links) of potential trading partners but do not offer the buyer any concrete buying opportunities (Bakos, 1991). The result is a so-called electronic shopping mall or market community, which provides consumers with a catalog of suppliers in a specific product field. In marketplaces with a central operator, the organizer or broker intervenes directly in the market process. The operator gathers product offerings and buyer wishes and matches this information within its database according to defined co-ordination criteria (Chircu & Kauffman, 1999). This proactive intermediation service between supplier and customer is offered as an entrepreneurial product. The result is an online auction or electronic marketplace that facilitates transactions for certain items (Whinston, Stahl & Cho, 1997). The aim of an auction is specifically to offer an object to as many bidders as possible in order to raise the price. In the case of electronic marketplaces, in contrast, there are mostly many sellers and many buyers simultaneously facing one another.

Operators of electronic marketplaces offer a co-ordination service for economic transactions on the WWW as economically independent entrepreneurs (Schmitz, 2000). The existing resources in computer technology really do make it possible for them to assume the role of proactive market managers (Malone, Yates & Benjamin, 1987). Not only do they determine the rules of trade under which the business transactions take place (Schmid, 1997), they also collect requests for transactions from both market actors and attempts to match these using criteria of efficiency and effectiveness. The added value, which ultimately forms the basis for the fee the operator collects from both market participants, is made up of the increased market transparency for the buyers and the greater efficiency of the intermediation for the sellers (Kollmann, 1999a).

Acceptance as a Competitive Criterion for Marketplace Operators

The marketplaces of the central operators are furthermore in direct competition with one another. Their goal is to win over suppliers and customers for their own intermediation platform and to match buyers and sellers for a fee (Kollmann, 1999b). This is only possible when the marketplace design is ‘acceptable’ to the users. Acceptance is seen here as a construct which reflects the marketplace user's belief in the ability of an electronic marketplace to match the needs of supply and demand successfully. The present article analyzes criteria which were important for the acceptance of an electronic marketplace from the standpoint of the marketplace users (buyers and sellers). These criteria can be used as the basis for a marketing concept designed to attract new participants to a marketplace. For that reason, the distinctive characteristics of electronic intermediation will be discussed in the following section. Subsequently, they will be incorporated into a model for the measurement of marketplace acceptance. The ensuing acceptance model will then be verified using data from a survey of users of an electronic marketplace specializing in used cars.

Basis of a Measure of Acceptance of Electronic Marketplaces

When examining market acceptance for an enterprise organized as an electronic marketplace, there are two distinctive characteristics which are of import: a tripolar participation structure and a derivative service aspect. The tripolar participation structure refers to the triangular transaction configuration (Figure 1). In contrast to a traditional exchange of goods between two parties (Alderson & Martin, 1965), a third partner intervenes here in the bilateral exchange relationship between buyer and seller. In this configuration all parties (seller/buyer/marketplace operator) are in direct contact with each other and the goods or services can only be exchanged when all three parties participate in the transaction (Figure 1). The third partner in the exchange, the marketplace operator, does not transform the exchanged goods or services. The operator simply provides an information channel which is requisite to the conclusion of the transaction and which reduces the costs of this transaction (Lee & Clark, 1996). Acceptance is thus determined by three market players with opposing optimization criteria (needs structures). Whereas supply and demand are attempting to optimize the design of the individual transaction, with the supplier acting to increase revenue and the demander personal benefit, the marketplace operator is seeking to optimize his profits as a function of the fees charged for matching the transaction partners. The acceptance of an ‘electronic marketplace’ enterprise, therefore, depends on two client groups, the demand side and the supply side, with disparate objectives (bilateral acceptance aspect).

Figure 1.

The basic structure and players in virtual marketplaces

The derivative service aspect describes a situation in which a good (in this case the intermediation service provided by the marketplace operator) does not result in direct, but rather an indirect utility, which first arises after use has been made of an interactive relationship within a communication system (Farrell & Saloner, 1985; Katz & Shapiro, 1985; Wiese, 1990). The derived benefit to the person using the good increases with the quantity and intensity of use of the other participants (Weiber, 1992). Typical examples of goods with direct network externalities are all types of telecommunications and information systems in which the benefit to each subscriber increases when another subscriber enters the system, since the existing network is expanded. Another consequence of this externality is that the benefit derived from a marketplace not only depends on the readiness to provide a service and the performance of the operator but also on the readiness to use the service and the performance of the operator's clients, i.e. on the buyers and sellers themselves. The marketplace operator cannot act as a mediator unless requests or offers are made for a product on the operator's site (Figure 1). Furthermore, the operator's success will increase with the number of requests and offers that exist on the site. The ‘electronic marketplace’ is thus dependent on the readiness of others to participate in it (acceptance) and, regardless of that readiness, does not provide a primary service (slavish acceptance aspect). An example should clarify these relationships:

The operator of AutoScout24 has been running this used-car site since 1997. Sellers enter their cars into the database and the operator mediates between buyers and sellers on the basis of the search data entered by the buyers. On average there are approximately 180,000 cars in the database, making AutoScout24 one of the largest marketplaces in this field in Europe. Around 2,500,000 users access the site every month to look for a car or to put a car up for sale. The marketplace only exists online and only deals with the information on the cars in the database (e.g. technical data, photographs, test reports). It is not possible to inspect the vehicles before intermediation takes place since there is no central car park or the like where the cars in the database can be put on display. It is a matter of structuring the object related information, not the real physical objects. AutoScout24 itself thus does not obtain an ownership interest in the objects.

The functional method of the electronic marketplace is thus as simple as may be imagined: A seller has the possibility of specifying a used car via an interactive entry module and to place this directly in the marketplace database. The marketplace operator is then commissioned to perform mediation (slavish acceptance aspect). Demanders have the choice of either searching in the database themselves or have the marketplace operator search by entering their search wish (activity of the market channel). Objects are then offered by this channel that fit the search criteria, but alternatives to the wishes initially given are also offered. On the other side the operator also informs sellers about the search wishes of demanders. The electronic marketplace AutoScout24 thus supports marketing on the seller side and on the buyer side ensures a certain level of market transparency. An actual transaction is only made, however, when both customer groups are satisfied with the results of the mediation (bilateral acceptance aspect). AutoScout24 is thus in competition with other marketplaces specifically on the German-speaking WWW that also offer a point of contact for the used-car trade (e.g. faircar, autocity, autoweb, mobile etc.), but also with real mediation institutions (e.g. newspapers). It is clear that in this kind of object mediation, other mediation mechanisms besides that of price apply. Other limiting conditions apply here than those in pure auctions (Lee, 1997).

Problems Related to the Electronic Mediation of Transactions

The tripolar participation structure (seller, buyer and marketplace operator) and the derivative service aspect complicate the acceptance issue. This can be illustrated with the following real-life situation. This practical example is based on an electronic real-estate marketplace with numerous houses/apartments for rent in its database, a good number of daily users looking for an apartment and thus the real opportunity of a transaction's being mediated. There are three typical acceptance problems which can arise in this constellation:

  • Problem 1: The site lists 250 apartments for rent in neighborhood X but does not have 250 people looking for an apartment in neighborhood X since the users prefer to live in neighborhood A. Consequence: The database volume is insufficient for a match and primarily the suppliers are disappointed with the marketplace (focus on supply).
  • Problem 2: The site lists 250 people looking for an apartment in neighborhood Y but does not list 250 places for rent in neighborhood Y. Rather, it only lists apartments in neighborhood C. Consequence: The database volume is insufficient for a match and primarily the demanders are disappointed with the marketplace (focus on demand).
  • Problem 3: The site lists 250 apartments for rent in neighborhood X and has 250 people looking for an apartment in neighborhood X but none of these match because all apartment seekers want a balcony and none of the apartments has one. Consequence: The quality of the database is too poor for a match and both the suppliers and seekers are disappointed with the marketplace (focus on intermediation).

The result is a two-sided acceptance problem for the marketplace operators which (cf. supra) which can be structured into four problem zones of quantitative and qualitative nature (Kollmann, 1998a). The main quantitative aspects are:

  • 1Chicken and Egg Problem
    The mutual interactive relationships in electronic marketplaces are the cause of the ‘chicken and egg’ problem (Durand, 1983; Earston, 1980), otherwise referred to as the ‘vicious circle’. The ‘chicken and egg’ problem can be summed up as follows: If there are not enough suppliers, or if the acceptance of the offered items is too low, no buyers will use the marketplace. If the acceptance of the buyers is too low, suppliers will not use the marketplace. This dilemma is one barrier to the acceptance of the ‘electronic marketplace’ institution.
  • 2Critical Mass Problem
    The installed basis, i.e. the number of users already present in a marketplace, determines the utility of the marketplace for new users since a greater number of users also increases the number of potential transactions (Farrell & Saloner, 1986). The larger the installed basis, the larger the derivative utility (cf. supra) for the marketplace participants. Once a certain number of users has been exceeded and the derivative utility has accordingly passed a certain level, it can be expected that the users will also accept the marketplace in the future and the number of new participants entering the marketplace will grow faster. The minimum number of users required before a marketplace “…can develop a sufficient usage for a long-term application for a user circle is referred to as critical mass” (Weiber, 1992).From the standpoint of the marketplace operator, this creates the situation of a double critical mass. What is more, both masses exist in a mutually dependent relationship. For the supplier, a certain number of demanders of a certain quality must be present to make the marketplace attractive to him. At the same time, a certain number of suppliers with certain characteristics must exist before buyers will enter the marketplace. In addition, there are the qualitative aspects of the problem:
  • 3‘Quality of Intermediation’ Problem
    The number of market participants alone, whether on the supply or demand side, does not yet allow any conclusions about the quality of the matched transaction partners or how well their needs can be satisfied with regard to the transaction items. It still remains to be seen, therefore, how the quality expectations of both parties are met by the marketplace. That is why a close match between what people are looking for and what is being offered plays an important role in this context.
  • 4‘Quality of Business Deal’ Problem
    Owing to the virtual nature of an electronic marketplace, it is not possible to inspect the items up for sale. After the marketplace operator has fulfilled the role of middleman between buyer and seller, i.e. after the preselection process for the items which were of interest to them is complete, the buyer still has to visit the seller to view the actual item. The ‘reality gap’ which can arise should the item not live up to the expectations created by the description on the electronic marketplace will have a negative impact on the electronic marketplace.

Based on the marketing problems described above, as seen from the standpoint of the marketplace operator, an initial concept for measuring user acceptance (by both suppliers and demanders) of electronic marketplaces has been developed.

Methods

The issue of user acceptance invites the question: Which indicators are of what significance for the acceptance of an electronic marketplace? The answer to this question can be used to develop a marketing concept which should identify the relevant aspects of a communication strategy for marketplace operators. This could help improve the acceptance of one's own marketplace by increasing the number and loyalty of buyers and sellers at a site. The question remains, though, how to measure the acceptance of a marketplace on which a central agent is merely responsible for information intermediation without any knowledge of the items in reality? A fitting instrument for this context has yet to be discussed in the economics journals. Up until now, the discussion has dealt mainly with the definition of the phenomenon per se and limited itself to the suggestion that the acceptance of an electronic marketplace should be determined from the standpoint of the marketplace user. One reason for this lack of attention is surely the fact that marketing experts have considered electronic marketplaces, as they do bricks-and-mortar marketplaces, merely to be another sales channel. Before an appropriate yardstick for acceptance can be found, therefore, it is necessary to develop a new causal model to define the acceptance of an electronic marketplace from the standpoint of the user. These findings can then form the basis of a general marketing concept for the operators.

An analysis of potential indicators which could describe the acceptance of electronic marketplaces, which can be described in general terms as intermediation institutions, is not only of theoretical interest, but owing to the growing number of such sites on the WWW, also of great practical interest (Wigand & Benjamin, 1996). In light of the theoretical problems of electronic marketplace acceptance (cf. supra), two central objectives were pursued in this study:

  • 1Identification of central indicators for electronic marketplace acceptance from the standpoint of the marketplace participants.
  • 2Verification of the ensuing multivariate models for measuring the electronic marketplace acceptance as relevant for the supply and demand sides.

The first step in the procedure chosen to bring us closer to the complex idea of ‘marketplace acceptance’ was to create a basis of understanding on which the first indicators were developed (Homburg & Giering, 1996). Based on an examination of the relevant literature, discussions with the management of various electronic marketplace sites on the Internet were carried out. The results were incorporated in a pre-test done with operators of electronic marketplaces on the German-language WWW in March/April 1997. The return rate was aceptable, with 37 of the 127 questionnaires being sent back (Kollmann, 1998b). The results of the survey (e.g. problem specifications of the definition of success construct and indicators) and the further theoretical considerations (Kollmann 1998a, 1999a, 1999b) resulted in an acceptance model for electronic marketplaces which will be described in greater detail in this article.

Structure of the Acceptance Model

The general tenor of the answers given by the marketplace operators in the pre-test was that an electronic marketplace is successful in practice when it brings together supply and demand at the information level in a cost-efficient and user-friendly atmosphere which guarantees a better chance of a real transaction being concluded than is the case on a competing intermediation institution. If the marketplace is successful in doing this, it is also acceptable to the users. It was in this context that the influencing factors were derived and the theoretical model constructed. It should provide an explanation for how the decision to participate in an electronic marketplace, which is based on subjective criteria, is made. The central construct being discussed here, namely ‘marketplace acceptance’, can consequently be seen as a synonym for the decision to participate in a market, which thus ensures the marketplace's success. This allows the formulation of the following definition (result of the survey of marketplace operators, see above):

  • Marketplace acceptance, from the viewpoint of the marketplace participant, reflects the subjectively perceived ability of the ‘electronic marketplace’ or its operator to adequately match supply and demand.

Examples of positive acceptance would be the first-time use of a site's services or the repeated use thereof. Negative acceptance would lead to the decision not to use a site's services or cause a participant to leave the marketplace.

Owing to the tripolar participation structure of electronic marketplaces, the influencing factors must be determined from the standpoint of two customer groups (supply and demand side). In accordance with the above description of the acceptance issue, the demand side will be looking at the number and quality of the suppliers and their items. Of further importance to them will be the conditions for accessing the database and their ability to make their search criteria clear to the marketplace operator. By contrast, the supply side will be more interested in the number and quality of the purchasers in the marketplace and the conditions under which they can enter items into the database. Getting good value for their money will be of relevance to both the supply and demand sides. They will be comparing the costs for using the service with the results offered by the marketplace operator. Furthermore, the distinctive nature of electronic intermediation and the inherent concentration on the information level will condition the question of the real consequence of intermediation; in other words, is the information on the marketplace in a position to generate a real sale?

The next step was to use the influencing factors identified above to develop theoretical constructs. Each construct reflected a group of influencing factors. Subsequently, certain hypotheses were tested in an empirical study to determine their link to ‘marketplace acceptance’. The ‘readiness to use’ construct includes search mask appearance and the ease of data input when searching for a product (demand side; H1d) or when posting an item for sale (supply side; H1s). The ‘database quality’ construct, from the supplier's point of view (H2s), stands for the number and quality of the information requests by the demand side. A general estimate only could be made here, because the determining of a ‘critical mass’ was made on a purely subjective basis. Further to this the search is oriented on individual cases (the right demander) and not on the total selection. For the product seekers (H2d), it stands for the type, number and quality of the items in the database. Here too it is not the total available ‘critical mass’ that is in focus, but the specific individual case (the right car). The ‘intermediation service’ construct (H3d/s) describes the manner in which the transaction partners are matched and the closeness of the match to their initial search parameters. The ‘actual transformation’ construct (H4a/n) refers to the transformation of the mediated information into a real transaction, the ultimate measure of the intermediation institution. Finally, the ‘intermediation costs’ construct (H5d/s) reflects the expenses incurred by the market participants for using the services provided by the marketplace operator. It includes the costs of both the accessing the marketplace (e.g. communication costs) and the site's user fees. These costs represent only a portion of the total transaction costs, however, (Giaglis, Klein, & O'Keefe, 1999; Malone & Rockart, 1992). The reason is that in the real transaction of business (e.g. real inspection of or provision of the car) additional costs will be incurred. Since these cannot be charged to the marketplace only those costs are examined that result directly from electronic mediation. The term ‘intermediation costs’ is thus used instead of ‘transaction costs’. The role played by these constructs in determining the central marketplace value was examined in this study. The correlations, which can be positive or negative, can be expressed in the following hypotheses:

  • Hypothesis Group 1 (positive correlation):
    The greater the database quality (H1s/d) and the readiness to use (H2s/d), and the better the intermediation service (H3s/d) and the actual transformation rate (H4s/d), the greater the acceptance of the electronic marketplace.
  • Hypothesis Group 2 (negative correlation):
    The higher the intermediation costs (H5s/d), the lower the acceptance of an electronic marketplace.

These correlations in themselves are in no way surprising. Of greater importance is the weighting or significance of the individual constructs in determining the overall acceptance.

Background of the Data Collection

The empirical basis for testing the hypotheses was provided by two online surveys done with the users of AutoScout24 in May/June 1999 and March/April 2000. The automotive industry in particular and its system of dealers (real intermediary) is influenced by the appearance of electronic marketplaces (virtual intermediary) (Klein & Selz, 2000). Each survey was carried out over a period of four weeks. A button with a link to the survey was installed on the AutoScout homepage. The users were asked to share their general opinion of electronic marketplaces. A filter question was used to split up the buyer and seller groups. After all unusable responses had been eliminated, the 1999 survey produced data records for 821 buyers and 115 sellers. The 2000 survey, which was filled in by 1,322 buyers and 473 sellers, used exactly the same questionnaire and was performed to check and confirm the acceptance model. In spite of the relatively large number of data records, it is still necessary to question the representativeness of the results since all respondents came from the ‘online user’ group and the survey only pertained to one specific topic (cars). Nonetheless, the results can be seen as a first indication of the validity of the acceptance model.

The hypothetical constructs were operationalized within the survey using a multiple item concept which employed a six-point forced rating scale (no evasive option was provided). The items were selected to reflect the hypothetical construct and described the concepts behind the constructs. The questions were formulated as personal statements to reduce the risk of misunderstanding and to help the respondents identify with the issue. The answers given were an expression of agreement or disagreement with the statements on the questionnaire (e.g. “A clearly designed homepage is important to me”). As a result, the respondents were incited to call up their ‘self-images’ and to describe these using the items on the questionnaire (Fisseni, 1990). The agreement or disagreement was expressed by clicking on the appropriate point of the graphical rating scale on the online questionnaire. This method placed a greater value on the subjective perception of the subjects. In this context, the overall hypothetical construct ‘marketplace acceptance’ was also recorded using global personal statements as a combination of all evaluation values (e.g. “I believe that I have an overall positive opinion of electronic marketplaces”).

The description of the model makes it clear that, with regard to the presumed correlations, it not a question of the connections between the measured variables but between the hypothetical constructs which cannot be directly observed themselves. As a result, a causal analysis is required to verify empirically the underlying hypotheses in their entirety. Hence, the survey data were analyzed using the LISREL approach of structural equation or causal analysis (Jöreskog & Sörbom 1988). Before the procedure was applied, an analysis was performed (normal distribution, univariate check, reliability and validity) to verify whether the question chosen to operationalize the hypothetical variables represented the theoretical constructs in an appropriate manner. This is the case, for example, when the measuring item showed a correlation of > 0.7 (good) or > 0.6 (satisfactory) with the underlying construct (Backhaus et al., 1998). The items which fulfilled this criterion and which were used in the LISREL analysis are shown in Figure 2. After the correlation matrix was created, the ULS method was chosen as the iterative estimate procedure since not all measured items displayed a normal distribution.

Figure 2.

Operationalization and degree of proximity to theoretical constructs

* The closer this figure is to “1”, the better the representation of the construct by the measuring item.
(d) demand side (s) supply side

RESULTS

Following the bilateral approach of supply and demand side, the fully standardized results of the LISREL model for the ‘supply side’ showed that the acceptance of an electronic marketplace can be very nicely illustrated with the generated constructs. In a first step, the directions of effect of the hypotheses (sign of the path coefficients) were confirmed in all cases. Looking at the intensity of the effects, it becomes clear that the 1999 ‘intermediation service’ construct with a path coefficient of 0.65 (2000: 0.71) had the strongest impact on the assessment of electronic marketplace acceptance. Consequently, the supplier's key criterion for market acceptance was the market's ability to provide buyers with an authentic purchase interest (qualitative assignment). In addition, ‘actual transformation’ (1999: 0.27; 2000: 0.31) and ‘readiness to use’ (1999: 0.22; 2000: 0.18) had a relatively strong influence on the acceptance of electronic marketplaces. By contrast, ‘database quality’ only had minimal influence. This can probably be explained by the fact that most suppliers do not really care how many buyers are active on the market (quantitative characteristic) as long as there is one who is willing to buy their product.

In general, the global quality criteria of the LISREL model for the ‘supply side’ can be characterized as very good (Backhaus et al., 1998). The quality-of-fit index (source variance) was 97% (2000: 96%), adjusted quality-of-fit INDEX (source variance with consideration given to degrees of freedom) was 96% (2000: 95%) and the root-mean-square index (measure of the variances not explained on average) was only 6% (2000: 7%), which are all very good values, so it is safe to assume that the explanations of the initial situation provided by the data are reliable. The quality criterion for the explanation of the total structural equations of the marketing model was also very good at 92% (2000: 91%) (total coefficient for structural equations). The data also provide a good explanation of the central ‘marketplace acceptance’ construct from the supply side's point of view with values of 93% (2000: 91%) (THETA-Value = 0.07 and 0.09). The illustration in Figure 3 provides a summary of the results.

Figure 3.

Structure model results from the LISREL calculations

The LISREL model for the ‘demand side’ was also able to portray the ‘marketplace acceptance’ and the constructs generated to explain it adequately. The directions of effect in the presumed hypotheses (sign of the path coefficients) were also confirmed in all cases. The critical remark must be made here, though, that the negative correlations between costs and acceptance were next to naught. One explanation for this could be that having to pay for an Internet service is still perceived as an absolute ‘disqualifying’ criterion by participants on the demand side. With regard to the intensity of the other effects, though, it becomes clear, though, that the 1999 ‘intermediation service’ construct with a path coefficient of 0.54 (2000: 0.59) had the strongest influence on the assessment of electronic marketplace acceptance for the buyers as well. As opposed to real marketplaces on which the buyer must find the seller alone, the emphasis here is on a supplier's being identified by the electronic marketplace. In contrast to the supply side, however, the buyers rated the number of sellers or offers on the marketplace as an important criterion (1999: 0.31; 2000: 0.42). Most likely this reflects the desire to have more choices, with the logic being that a higher number of potential sellers means a better chance of finding the item one is looking for. As well, ‘actual transformation’ with a value of 0.25 (2000: 0.32) still had a relatively strong influence on the acceptance of electronic marketplaces. In particular, the actual availability played an important role here.

The global quality criteria of the LISREL model for the ‘demand side’ can also be characterized as good. The quality-of-fit index was 98% (2000: 97%), adjusted quality-of-fit was 97% (2000: 96%) and the root-mean-square index was only 5% (2000: 6%), all of which are very good. The quality criterion for the explanation of the overall structural equations of the marketing model can still be labeled good at 80% (2000: 82%) (total coefficient for structural equations). Finally, the data made it possible to explain the central ‘market acceptance’ construct to an extent of 80% (2000: 84%) all the same (THETA-Value = 0.20 and. 0.16). The illustration in Figure 3 also provides a visual summary of these results.

Overall, the empirical data gathered in the 2000 survey largely confirmed the results from the 1999 survey, making it safe to assume a high reliability and accuracy of the measurements (Peter, 1979). This assumption would appear justified considering the consistency of the results after the repeated measurements.

Discussion

The results of the empirical study have shown that electronic marketplace acceptance, at least in the case of autoscout24, can possibly be adequately measured with the model proposed in this article. The critical comment must be made, however, that only visitors to autoscout24 were questioned. A parallel questionnaire of visitors to other marketplaces for used goods or other objects (e.g. real property) was not made. It can therefore not be ruled out that the answers might be influenced by the experiences of other marketplaces. The results might also be different from marketplace to marketplace, because despite the relatively large number of those questioned, demographically different users are to be found there. Further, only online users were questioned, not those possible users of the marketplace who do not have Internet access. It must therefore be established that control groups that could verify the results are lacking in both online as well as the offline areas. In summary the need remains for further research for establishing the acceptance of electronic marketplaces. A survey of different marketplaces with different trade goods could be made in a first phase. In addition offline surveys of potential users who use traditional mediation organs for example should be conducted, (e.g. newspaper advertisements). An objective of the current survey was thus to lay the first foundation for future activities. In particular the first indicators for practice, i.e. for the marketplace operators, may be offered as a result of the double confirmation of the statements made.

Hence, this could provide the operators of other electronic marketplaces with an instrument to verify the acceptance of their marketplaces from their clients' point of view. Should they be able to confirm a similar weighting of the individual factors with regard to the overall acceptance, the operators should pay special attention to the following points when putting together their individual marketing concept:

  • 1Even if the ‘readiness to use’ is of lesser importance (e.g. clear page layout and good navigatability), the data show that ease of data input is more important for the sellers than the buyers. As a result, the marketplace operator should consider creating online interfaces which automate the posting of the ads by the supply side.
  • 2The marketplace operator is best served by a two-pronged strategy when it comes to database quality. For the supply side, the operator must document and advertise the quality of the searches performed (e.g. their authenticity). Of particular relevance here is the acceptance problem of the intermediation quality. On the other hand, the operator should emphasize the number of items available in the database to attract more buyers, which raises the question of the chicken and egg problem and critical mass on the supply side.
  • 3As expected, the intermediation service, i.e. the matching of supply and demand, is of key importance for both sides of the market. This service is the main determinant of the acceptance of an electronic marketplace and its operator. Consequently, offering this service should be the guiding principle for the marketplace operator in the role of entrepreneur. The electronic intermediation is the factor which makes the operator distinctive when compared to a real marketplace, and the matching success rate can be seen as a competitive criterion vis-à-vis other electronic intermediation platforms. The fact that both buyers and sellers use the intermediation service provided by the marketplace operator underscores the tripolar structure illustrated here. The marketplace operator must ensure that the product criteria are clear, the product descriptions are complete and the product requests are authentic. It should also be possible for the buyer and seller to contact each other directly.
  • 4The actual transformation, i.e. the transformation of the electronic intermediation into a real exchange of goods, also plays an important role for electronic marketplaces. Discrepancies between the electronic information and the items in reality are attributed in retrospect to the electronic marketplace, which has a negative impact on its image. This risk particularly depends on the quality of the information posted to the operator's database by both buyer and seller (externality or derivative service aspect). As a result, the marketplace operator must ensure that items being offered on the site fulfill certain objective quality criteria (e.g. MOT inspection) so that the seller's descriptions are not too far off the buyer's expectations. If expectations are not met, no transaction will occur, which impacts negatively on the real transaction rate, another communications instrument used by the marketplace operator to promote his site.
  • 5The results obtained with regard to intermediation costs show clearly that buyers are currently not willing to pay for this type of service on the Internet. For this reason, the only financing option open to the marketplace operator are fees charged to the supply side (e.g. for each ad placed in the database).

Ancillary