Key factors in forming an e-marketplace: An empirical analysis
Lee Tzong-Ru
a,*, Li Jan-Mou
baDepartment of Marketing, National Chung-Hsing University, Taiwan, ROC
bDepartment of Transportation Technology and Management, National Chiao-Tung University, Taiwan, ROC Received 1 November 2004; received in revised form 12 April 2005; accepted 10 October 2005
Available online 2 November 2005
Abstract
Currently, the major marketing channel for flower suppliers and retailers in Taiwan is the flower wholesale market. However, when the retailers make purchases in the wholesale market, the dominant suppliers offer poor service, and the retailers find it inconvenient to collect information on the price of flowers. Our study shows that the E-Commerce mechanism of the e-marketplace can improve trading efficiency and lower the cost of collecting information as well as the purchase price. According to our analysis, the e-marketplace can use ‘‘a com-bination of pictures, literal description, and regulated classification’’ to introduce the quality of flower products. By Fuzzy Delphi, the key factors which affect the operation modes between the retailer and the e-marketplace are ‘‘cooperation on urgent orders’’, ‘‘accuracy of order processing’’, and ‘‘order processing efficiency’’. Then, based on the three key factors, we use Fuzzy Multiple Criteria Decision Mak-ing to find what operation modes the e-marketplace should take to cooperate with the retailer. Retailers find the three operation modes ‘‘actively placing orders’’, ‘‘jointly negotiating prices’’, and ‘‘free bidding’’ equally compatible, so we suggest that the e-marketplace should provide these modes at the same time for retailer use and later the retailers can adjust the modes according to their business performance. 2005 Elsevier B.V. All rights reserved.
Keywords: E-commerce; E-marketplace; Floral industry; Fuzzy Delphi; Fuzzy Multiple Criteria Decision Making; Kano analysis
1. Introduction
Internet has quietly linked global markets, but it is unli-kely to alter the trading mode and preference in each mar-ket in the short term; thus, to operate e-commerce in a certain market, what we first need to know is the trading mode and preference in that market. E-marketplace is a form of e-commerce, supplier and the retailer deal through the e-commerce mechanism. However, they do not neces-sarily deal through the trading mechanism provided by the e-marketplace. This paper argues that the e-market-place for the floral industry in Taiwan aims to meet the demand of both supplier and retailer. The technology for providing an e-marketplace is readily present. This study unfolds the criteria for joining an e-marketplace and pro-poses a conceptual framework to study the key factors in
forming an e-marketplace. The study subject is the floral industry in Taiwan.
Currently, the major marketing channel for flower sup-pliers and retailers is the wholesale market, where the wholesaler bids for the supplierÕs flowers and then sell them downstream to the retailer (Fig. 1). Although there are hundreds of flower retail websites on Internet and in recent years e-commerce has been used extensively, most retailers still make purchases in the traditional wholesale market. However, when the retailers make purchases in the whole-sale market, the wholewhole-salers offer poor service; besides, it is inconvenient for buyers to collect information on the prices of flowers. An e-marketplace is able to solve the problem by providing retailers a new marketing channel and offer-ing timely information on website about the production-marketing of flowers so as to reduce the cost of collecting information as well as the purchase price, and to improve trading efficiency.
The paper suggests key factors in forming an e-market-place for the floral industry in Taiwan, investigates by
1567-4223/$ - see front matter 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.elerap.2005.10.004
*
Corresponding author.
E-mail address:trlee@dragon.nchu.edu.tw(T.-R. Lee).
questionnaire the attitudes and interests of both supplier and retailer in selling or purchasing flowers through e-com-merce, and analyzes the possible cooperation modes among supplier, retailer, and e-marketplace. The scope of the paper is to explore the possibility of an e-marketplace for the floral industry in Taiwan. The objective is to find the key factors in forming an e-marketplace for the floral industry in Taiwan. The main contribution of this paper is to highlight the key factors in forming an e-marketplace for the floral industry in Taiwan.
2. Literature review
Ratchford et al. [12] discussed a model of demand for Internet and other information sources on the premise that Internet is the most efficient source of providing functions and prices; however, his empirical subject was the automo-bile, while agricultural products like flowers do not have clear functions, which thus require other substitutes like quality check or standard classification.
OÕKeefe and Loebbecke[11] suggested that researchers should not blindly suppose that the virtual world is of interest in its own right because most of the time people live in the physical world and have already developed a mature approach to integrate the experiences they have had in both real and virtual worlds. Thus, when we analyze the business mode in the virtual world like the electronic marketplace, we should take into account the business mode in the real world.
Steinfield et al.[15]explained how organizations in the Netherlands combine the operation of the real and the vir-tual by proposing four points: cost savings, improved dif-ferentiation, enhanced trust, and market extension, OÕKeefe and Loebbecke[11]thought that these four points could apply to other related studies on the combination of e-commerce and activities in the real world. Our study dis-cusses the key factors which influence the formation of the e-marketplaces in four aspects: logistics flow (cost savings), business flow (market extension), cash flow, and informa-tion flow.
Using Extended Web Assessment Method (EWAM), Schubert [13] discovered that, from the consumerÕs point of view on consumer goods and on e-commerce service of Internet banks, most websites fail to satisfy consumer expectations; on the other hand, those websites with high-quality service are not necessarily successful in their business. Schubert thought that the soundness of the Busi-ness Model and the assessment of the Website should be discussed separately. Our study analyzes the flower e-mar-ketplace based on the soundness of the Business Model and
find the key factors in forming the flower e-marketplace based on the expectations of both supplier and retailer.
Luo and Seyedian [10]quoted the viewpoint of Kenny and Marshall [8] that contextual marketing refers to the extent to which e-businesses use the ubiquitous Internet to provide customers with relevant information in the right context and in real-time, and that contextual marketing is important because users are already overloaded with infor-mation. What they need most is relevant information pro-vided in real time at the point of need. Our study confirms this viewpoint, that is, an e-marketplace has to provide the proper business modes for both supplier and retailer.
Chen [2]in her essay ‘‘Study on the Evaluation Proce-dure of Selecting Airport Location’’ used Fuzzy Delphi and Fuzzy Layer Analysis as evaluation tools. By question-naire, and taking the professionals familiar with transpor-tation and delivery and the staff in Taipei Agricultural Marketing Company as research subjects, Cheng[4,5]used Fuzzy Multiple Criteria Decision Making to evaluate the delivery modes between the company and its 18 affiliated supermarkets; the result show that ‘‘commission delivery’’ is the most feasible delivery mode.
Kano analysis is proposed by the Japanese Kano [6], which was initially used to systematically deal with the demand from customer and then transfer the demand onto the improvement of products so as to improve the compet-itiveness of enterprises. Kano considered that the major customer demands are: must have, linear satisfier, and delighted.
In respect of flower consumption behavior, Lee and Cheng[3]in their study ‘‘The Business Environment Analy-sis and the Sales Channel Strategy Making of the Flower Stores (Part 2)’’ analyzed the current general environment confronted by flower stores, their roles in the marketing channels, and the right strategies they should take; they made strategic analysis and suggested that the flower stores should put more emphasis on their business management strategy and well employ their unique advantages and reduce threats so as to fight against external market impacts. Lee[9]in his study of ‘‘New Sales Channel for Flowers – Application of Internet’’, by questionnaire and the applica-tion of independent checking and LogitÕs mode, analyzed the background of the interviewees, discussed the relation-ship between the intervieweesÕ background and their pur-chase behavior, as well as finding out the feasible separated markets for bouquet promotion. The result shows that the feasible separated markets is composed of consumers ‘‘who have computers with Internet access’’ and ‘‘whose education level is above college or graduate school (inclusive)’’.
Comsumer Wholesale
Market Wholesaler
Suppliers Retailer
3. Methodology
Besides using descriptive statistical method, this study employs Fuzzy Delphi to extract the key factors which affect the willingness of flower supplier and retailer to join the e-marketplace, and further uses Fuzzy Multiple Criteria Decision Making[14]to choose the more feasible coopera-tion modes from all the possible cooperacoopera-tion modes. Our study also uses Kano analysis to analyze the supplierÕs atti-tudes towards the operation mode of the e-marketplace. The following is an introduction of the three methods. 3.1. Fuzzy Delphi
Based on Kauffman and Gupta [7], Fuzzy Delphi is a variation of the Delphi method using triangular fuzzy num-bers in which communication with experts is the same, but the estimation procedure is different. Some important remarks about this method can be found in Kauffman and Gupta[7]. Here, the function of this method is to list several factors in order according to their importance, which is shown in the form of a geometric mean. The equa-tion of the geometric mean is shown as:
Xi¼ mffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffia1i a2i a3i ami
p
; ð1Þ
where Xi is the score of question i, m is the number of
respondents, ajiis the score of respondent j to question i.
3.2. Fuzzy Multiple Criteria Decision Making
This method evaluates the feasibility of each item; the higher the calculated figure, the more feasible is the item. This paper adopts the method to evaluate all possible coop-eration modes for the flower supplier who would like to par-ticipate in the flower e-marketplace, then to find out the more feasible cooperation modes. The steps are as follows: Step 1: Decide Factor Set
Factor Set U = {ui, i = 1, 2, . . ., n}, uiis the key
fac-tor i
Step 2: Find the weight of each factor in Factor Set W = {wi, i = 1, 2, . . ., n}, wiis the relative weight of
the key factor (ui)The weight of each factor is
shown by the triangle fuzzy figure, as Eq.(2)shows: wi¼ fwiL; wiA; wiHg; i¼ 1; 2; . . . ; n; ð2Þ
where wiLis the lowest of all the scores in question (i),
wiAis the geometric mean of all the scores in question
(i), wiHis the highest of all the scores in question (i).
Step 3: Find the value of each item to each factor, which is shown by rjk, representing that the performance
score of the factor j in the item k; the higher the score, the better the factor j performs in the item k. rjkis also
shown by the triangle fuzzy figure, as Eq.(3)shows: rjk¼ frjkL; rjkA; rjkHg; j¼ 1; 2; . . . ; v; k ¼ 1; 2; . . . ; p;
ð3Þ
where rjkis the performance score, rjkLis the lowest
of the geometric means of the score of the factor j in the item k, kjkA is the geometric mean of the
score of the factor j in the item k, k.rjkHis the
high-est of the geometric means of the score of the factor j in the item k.
The fuzzy evaluation matrix Rkof all factors in the
item k, as Eq.(4)shows:
Rk¼ r1k r2k .. . rvk 8 > > > > < > > > > : 9 > > > > = > > > > ; ¼ r1kL r1kA r1kH r2kL r2kA r2kH .. . .. . .. . rvkL rvkA rvkH 8 > > > > < > > > > : 9 > > > > = > > > > ; . ð4Þ
After calculation, we get the triangle fuzzy figure of the item k, which has to be positively naturalized before comparison. The process of positive natural-ization is in Step 4:
Step 4: Fuzzy Compound Calculation
Bk¼ W Rk ¼ fw1; w2; w3; . . . ; wng r1kL r1kA r1kH r2kL r2kA r2kH .. . .. . .. . 0 B B @ 1 C C A ¼ ½ðw1L; w1A; w1HÞ; ðw2L; w2A; w2HÞ; . . . ; ðw3L; w3A; w3HÞ r1kL r1kA r1kH r2kL r2kA r2kH .. . .. . .. . 0 B B @ 1 C C A ¼ ½bkL; bkA; bkH ð5Þ where Bk is the compatibility of the item k, also
shown in the fuzzy triangle figure [bkL, bkA, bkH]
Step 5: De-fuzz and Sort
After fuzzy calculation, we get a set of figures rep-resenting the triangle fuzzy function, and we use the simple triangle centering method to calculate the compatibility of each factor.
3.3. Kano analysis
Zhang and von Dran[16]used KanoÕs model of quality to analyze the customerÕs quality expectation of the web-sites of special form (CNN.com) and discovered that KanoÕs model of quality could be extensively applied to many areas or many types of websites; through long-term study they discovered that customer expectation on quality changes with time, so it fails to use the same single quality checking table to evaluate quality expectation. KanoÕs model can be employed to identify quality expectations and time transition of quality factors; customers in the same web areas do not consider all quality factors equally important; the quality factors in different web areas have different importance, but certain factors exhibit great importance in all web areas.
This method classifies customer demand threefold: must have, linear satisfier, and delighted. ‘‘Must have’’ means
the property or function customers consider the product must certainly have or they cannot accept the product. Kano analysis inquires customer demand by questionnaire with two questions as a set, including both a positive and a negative question. For example:
Positive: What is your opinion if Service A is included? Negative: What is your opinion if Service A is not included?
With respect to the questions, the interviewee has five choices: (1) I like it very much, (2) It must be included, (3) I remain neutral, (4) ItÕs so-so, and (5) I dislike it. The results are transformed and shown in Table 1; the transformed results include ‘‘must have’’, ‘‘the more, the better’’, ‘‘delighted’’, ‘‘there is some problem’’, ‘‘the con-trary’’, and ‘‘no difference’’.
Suppose the interviewee thought Service A must be included (Choice 2) and disliked no Service A (Choice 5). After the transformation inTable 1, we know that Service A belongs to the function ‘‘must have’’.
InTable 1, ‘‘no difference’’ means that customer remains neutral and is not affected by the property of the product. ‘‘The contrary’’ means that the interviewee dislikes or does not need the property of the product. ‘‘There is some prob-lem’’ means that the interviewee chooses both ‘‘I like it very much’’ and ‘‘I dislike it’’ in answering both the positive and the negative question. ‘‘The contrary and ‘‘There is some problem’’ indicate that the customer shows contradiction in answering the questions, but these situations more or less appear in common questionnaire investigations. After we have interviewed a certain number of customers, we tally the number of times in each grid ofTable 1and relate it in terms of percentage, to show the viewpoint of the customers.
4. Empirical results 4.1. Suppliers
In respect of flower production, according to the Profile of Agricultural Suppliers in Taiwan[1], in June 1996, there were 486 groups of flower suppliers in Taiwan, mostly from the central countries of Nantou, Changhua, and Taichung. Currently the major sales channels for suppliers are the four wholesale markets in Taipei, Taichung, Changhua, and Tainan. There are two questionnaires with two parts
therein both for the supplier and the retailer, respectively. The first part of the questionnaire deals with the supplierÕs related sales background, including gender, age, level of education, and his/her major sales channel; the second part is about the related B2B (Business to Business) trading information of the supplier, including his/her willingness in selling flowers in the e-marketplace, and his/her pre-ferred cooperation mode with the e-marketplace. The Del-phi processes were conducted from June 28, 2002 to July 7, 2002, and the 150 interviewees (all of them were experts in this industry) are chosen by random sampling from suppli-ers in the central countries listed in the Profile of Agricul-tural Suppliers in Taiwan. 135 questionnaires were retrieved, with a high effective retrieve rate of 90%. There were five iterations in the whole Delphi processes to get the final results. On the last iteration, our researchers con-duct personal interviews and the issue is very attractive to interviewees; besides, the questions has little to do with the business confidentiality of the interviewees, so most of them feel free to answer the questions.
Table 2shows most interviewees are male (88.9%), over 40 years old (37.8%), with senior high level of education (40.7%), with an average of 9.29 years on flower produc-tion and their planting areas are mostly in Nantou (43%). Flower suppliers with the above characteristics are the principal research subjects of our study.
Table 3 shows the current sales modes of the intervie-wees. The wholesale market is the principal sales channel (54.3%), and cut flowers (75.8%) are a major production.
Table 1
KanoÕs transformation of customer demand Positive description Negative description
I like it very much Must have I remain neutral ItÕs so-so I dislike it
I like it very much There is some problem Delighted Delighted Delighted The more, the better
Must have The contrary No difference No difference No difference Must have
I remain neutral The contrary No difference No difference No difference Must have
ItÕs so-so The contrary No difference No difference No difference Must have
I dislike it The contrary The contrary The contrary The contrary There is some problem
Table 2
Background of interviewees: suppliers Characteristics of interviewees Statistics Gender Male (88.9%), female (11.1%) Seniority on flower production 9.29 years on average
Age 50 and over (20%) 35–39 (14.8%)
40–44 (17.8%) 45–49 (14.8%)
25–29 (14.8%) 20–24 (3%)
30–34 (14.8%)
Level of education Senior high (40.7%) Vocational school (11.1%) Junior high (25.2%) College (3.7%) Elementary school (19.3%)
Planting area Nantou (43%),
94.1% of the interviewees do not deal with contract produc-tion and the usual goods-delivering period is within 7 days (50%). Most suppliers who do not deal with contract pro-duction show great willingness to sell flowers through Internet (83%). Suppliers who do not deal with contraction product but are willing to sell flowers through Internet tend to take the wholesale price of the day as reference (40.2%) before they decide on the price of their flowers. The inter-viewees who do not deal with contract production and are not willing to sell flowers through Internet to flower stores think that ‘‘ on-line demand is deficient’’ (5.39 points), and ‘‘the sales procedures online are complicated’’ (5 points). We suggest that the e-marketplace should simplify the online sales procedure as much as possible, but we have not recommended any specific standard online-sales proce-dure to respondents.
While suppliers cooperate with the e-marketplace, the activities are categorized into four parts: logistics flow, cash flow, business flow, and information flow. There are 23 fac-tors altogether, and the interviewees are asked to rate the importance of each of the factors based on a 10-point scale, 10 being the most important and 1 the least important. We use Fuzzy Delphi to calculate the geometric mean of each factor, and the results are shown inTable 4.
Fig. 2shows the scores of all the factors in linear illustra-tion, Based on the experience principle of ‘‘choosing no more than 7 successful key factors in a study’’, we classify the fac-tors into 5 groups. The geometric mean of the first group is the highest, indicating that it is most representative of the intervieweesÕ opinions, hence the most feasible website key factors. From this we extract, in descending order, the four
key factors which affect the cooperation modes between the supplier and the e-marketplace, namely ‘‘accuracy of order processing ’’, ‘‘trading credit investigation’’, ‘‘quality check’’, and ‘‘production project’’. ‘‘Order processing cor-rectness’’ is under Business Flow, ‘‘trading credit investiga-tion’’ under Cash Flow, and ‘‘quality check’’ and ‘‘production project’’ under Logistics Flow. There are no factors attributed to Information Flow, possibly because the suppliers seldom use related production information to make production planning, and they are not quite familiar with the application of production-marketing information.
According to the analysis, we suggest that the e-market-place should strengthen flower quality check to ensure flower quality stability, and take into account produc-tion-marketing information while planning production projects, cooperating with the suppliers on planned pro-duction. Besides, the e-marketplace should check the trad-ing credit of both sides to prevent bad debts, and improve the accuracy and efficiency of order processing so as to upgrade customer service.
Presently, the major production-marketing modes for suppliers are ‘‘spot selling’’ and ‘‘contract production’’. We also include ‘‘free bidding’’ which is now popular in Internet. The contents of the three modes are as follows: 1. Spot selling: As soon as the retailers join the
e-market-place and e-market-place orders online, the suppliers sell their flowers based on the orders.
2. Contract production: After confirming the total volume of orders and the transaction day, the e-marketplace arranges the delivery date with the suppliers by contract.
Table 3
Current sales modes of the interviewees
Item Statistics
Major sales channels and ratio of sales volume to total sales volume
54.3% to wholesale markets 7.9% to retailers
22.6% to collecting centers 3.4% to flower stores
10.2% to exporters 1.6% to other channels
Major types of flowers they supply Cut flowers 75.8%, Potted flowers 15.8%, Others 8.4%
Dealing with contract production No (94.1%), Yes (5.9%) Usual goods-delivering period Within 7 days (50%)
2 months and over (25%) 7 days to 1 month (12.5%) 1 to 2 months (12.5%) Attitude of the flower suppliers who do
not deal with contract production towards selling flowers through Internet to flower stores
Yes (83%), No (17%)
Price-deciding reference of flower suppliers who do not deal with contract production
Wholesale price of the day (40.2%) Wholesale price of three days ago (26.8%) Wholesale price of the day before (16.1%) Wholesale price of one week ago (11.6%) Others (5.4%)
Why flower suppliers who do not deal with contract production are not willing to sell products through Internet to flower stores
On-line demand is deficient (5.39 points) Classification-packaging standard is complicated (3.39 points)
Sales procedures online are complicated (5 points) Production scale is small (3.26 points) Price-deciding methods are complicated (4.04 points) They show little interest (2.83 points) Note. 7-point scale – 7 the highest, 1 the lowest.
3. Free bidding: Suppliers provide information about their products including literal description or pictures, which are then put on the e-marketplace for retailers to bed. We employ Fuzzy Multiple Criteria Decision Making to evaluate the compatibility of the three cooperation modes. Table 5 lists the results. The evaluation shows that ‘‘con-tract production’’ is the most acceptable mode to suppliers, ‘‘free bidding’’ the second, and ‘‘spot selling’’ the last. The compatibility of ‘‘contract production’’ is very close to that of ‘‘free bidding’’, both higher than that of ‘‘spot selling’’. Next, we analyze the suppliersÕ attitude towards 13 operation modes with Kano analysis. Table 6 shows the results.
In Kano analysis, the questions are addressed positively and negatively to each operation mode, giving a total of 26 questions. The following is an illustration:
Positive question: What is your opinion when the oper-ation mode is ‘‘integrating all orders and then purchas-ing flowers from suppliers’’?
Negative question: What is your opinion when the oper-ation mode is ‘‘not integrating all orders and then pur-chasing flowers from suppliers’’?
There are five options: (1) I like it very much, (2) Must have, (3) I remain neutral, (4) ItÕs so-so, and (5) I dislike it. We arrange the opinions of the interviewees (Table 7). With reference toTable 1, the results show that 32 intervie-wees are for ‘‘ must have’’, 31 for ‘‘the more, the better’’, 22 for ‘‘delighted’’, 44 for ‘‘ no difference’’, 4 for ‘‘there is some problem’’, and 2 for ‘‘the contrary’’. Thus, for this operation mode, the principal attitude of the interviewees is ‘‘no difference’’, followed by ‘‘must have’’, ‘‘the more, the better’’, ‘‘delighted’’, ‘‘there is some problem’’, and ‘‘the contrary’’, respectively.
Table 6 shows the attitude of the interviewees towards the 13 operation modes. As ‘‘there is some problem’’ and ‘‘the contrary’’ appear sparingly, and both are negative, we put them under one heading. We tally the number of times of all the options and compute the percentage. For example, in the operation mode ‘‘integrating all orders
Table 5
Compatibility of the operation modes of the e-marketplaces
Operation mode Compatibility
Contract production 1.21058
Free bidding 1.20870
Spot selling 1.18955
Table 4
Factors affecting the operation modes between the supplier and the e-marketplace
Category Serial number Factor Average
Logistics flow 1 Quality check 8.81
2 Production project 8.52
3 Flower-packing and staff training 7.58
4 Efficiency of car dispatch 7.15
5 Freight charge 7.04
6 Form and amount of compensation 7.04
7 Urgent order processing 6.46
8 Quantity and capacity of cars 6.36
Cash flow 9 Trading credit investigation 8.81
10 Whether capital is sufficient or not 7.65
11 Trading price and trading volume 7.31
12 Processing fee 6.19
Information flow 13 Provision and compilation of production-marketing information 8.25
14 Volume and stability of demand 8.14
15 Efficiency of deciding on trading price 7.53
16 Establishing trading mechanism 7.42
17 Price prediction 6.68
Business flow 18 Accuracy of order processing 8.91
19 Order processing efficiency 8.19
20 Dealing with damaged flowers 8.08
21 Processing of customer complaint 7.58
22 Goods tracking 6.99
23 Frequency and degree of contact with buyers 6.6
(1) (2) (3) (4) ( ) (6) (7) (8) (9) (10) (11) (12) (17) (16)(15) (14)(13) (18) (19) (20) (21) (22) (23) 7 8 6 9 10 5
and then purchasing flowers from suppliers’’, 33% of the interviewees find ‘‘no difference’’, 24% ‘‘must have’’, and 23% ‘‘the more, the better’’. Therefore, this operation mode favors ‘‘no difference’’, ‘‘must have’’ and ‘‘ the more, the better’’.
The ‘‘Conclusion’’ column shows that the four principal attitudes of the interviewees towards each operation mode
are ‘‘no difference’’, ‘‘must have’’, ‘‘the more, the better’’, and ‘‘there is some problem’’. ‘‘No difference’’ appearing most frequently probably because the interviewees are not quite familiar with the current operation modes of the e-marketplace. ‘‘There is some problem’’ and ‘‘the con-trary’’ appears only in the operation mode ‘‘flowers are grown locally and/or imported’’. To get the real picture, we compile the statistics of the two options separately, and find that 49 interviewees (36%) have chosen ‘‘the con-trary’’ on this item and 6 (5%) for ‘‘there is some problem’’. In other words, 36% of the interviewees consider the oper-ation mode unnecessary; thus, the suppliers favor flowers grown locally.
Interviewees favor ‘‘must have’’ towards four operation modes, viz, ‘‘integrating all orders and then purchasing flowers from suppliers’’, ‘‘setting up an arbitration mecha-nism to deal with quality check, damage goods, and com-pensation’’, ‘‘suppliers choose orders they prefer online’’, and ‘‘a professional computer company is in charge of the website and maintenance’’.
Table 6
Attitude of interviewees towards the operation modes of the e-marketplaces
Operation mode Description
Must have The more, the better
Delighted No difference There is some problem/ The contrary
Conclusion
Integrating all orders and then purchasing flowers from suppliers
32 (24%) 31 (23%) 22 (16%) 44 (33%) 6 (4%) No difference, Must have, The more, the better Suppliers get better pricing based on
their sales volume online
9 (7%) 15 (11%) 17 (13%) 78 (58%) 16 (12%) No difference
Setting up an arbitration mechanism to deal with quality check, damaged goods, and compensation
33 (24%) 24 (18%) 11 (8%) 59 (44%) 8 (6%) No difference, Must have
Suppliers become members, provide only registered members of suppliers can sell their products online
24 (18%) 11 (8%) 13 (10%) 75 (56%) 12 (9%) No difference
Suppliers choose orders they prefer online
46 (34%) 18 (13%) 16 (12%) 42 (31%) 13 (10%) No difference, Must have
Suppliers have to be shareholders of the e-marketplace or pay a fee to become members.
12 (9%) 8 (6%) 6 (4%) 88 (65%) 21 (16%) No difference
Flowers are grown locally and/or imported
5 (4%) 5 (4%) 4 (3%) 66 (49%) 55 (41%) No difference, There is some
problem, The contrary Professional transportation
companies are in charge of transportation and delivery
25 (19%) 22 (16%) 10 (7%) 67 (50%) 11 (8%) No difference
A professional computer company is in charge of the website and maintenance.
31 (23%) 32 (24%) 10 (7%) 57 (42%) 5 (4%) No difference, Must have, The more, the better
Ranking of trading volume and sales volume by suppliers and buyers are posted on the e-marketplace
15 (11%) 10 (7%) 20 (15%) 68 (50%) 22 (16%) No difference
On-line trading is only for flowers of large trading volume
15 (11%) 4 (3%) 11 (8%) 77 (57%) 28 (21%) No difference
On-line trading is open to all breeds of flowers
20 (15%) 22 (16%) 17 (13%) 67 (50%) 9 (7%) No difference
Providing members with compiled production-marketing information (price in all wholesale markets, yield in place of production, and related activities)
28 (21%) 41 (30%) 17 (13%) 46 (34%) 3 (2%) No difference, The more, the better
Table 7
Attitude of interviewees towards the operation mode ‘‘integrating all orders and then purchasing flowers from suppliers’’
Positive Negative I like it very much Must have I remain neutral
ItÕs so-so I dislike it
I like it very much 2 0 15 7 31
Must have 0 0 7 2 13
I remain neutral 1 0 12 5 7
ItÕs so-so 0 0 13 5 12
I dislike it 0 0 1 0 2
Interviewees favor ‘‘the more, the better’’ towards three operation modes, viz, ‘‘integrating all orders and then pur-chasing flowers from suppliers’’, ‘‘a professional computer company is in charge of the website and maintenance’’, and ‘‘providing members with compiled production-marketing information (price in all wholesale markets, yield in place of production, and related activities)’’.
Interviewees favor both ‘‘must have’’ and ‘‘the more, the better’’ towards two operation modes, viz, ‘‘integrating all orders and then purchasing flowers from suppliers’’, and ‘‘a professional computer company is in charge of the website and maintenance’’.
From the analysis we notice that, the e-marketplace has to employ Internet marketing to increase and integrate orders and then purchase flowers from suppliers. A profes-sional computer company should be in charge of the website and maintenance. The e-marketplace has to set up a clear and fair arbitration mechanism to deal with quality check, damage goods, and compensation. To attract suppliers to browse the website, the e-marketplace should allow suppli-ers to choose ordsuppli-ers they prefer online and offer as much production-marketing information as possible.
4.2. Retailers
The first part of the questionnaire deals with the retai-lerÕs background, including gender, age, level of education, frequency of Internet usage and service area (location of flower stores). The interview was conducted from June 28, 2002 to July 7, 2002, and the 200 interviewees were cho-sen by random sampling from retailers (flower stores) in areas north of Hsinchu county listed in Flower Retailers in Taiwan in 1996. 199 questionnaires were retrieved, with a high effective retrieve rate of 99.5%. Our researchers con-ducted personal interviews and the issue was very attractive to interviewees; besides, the questions had little to do with the business confidentiality of the interviewees, so most of them felt free to answer the questions.
Table 8 shows most interviewees are female (67.3%),
under 35 years old (66.3%), with senior high and above
in education level (90.5%), do not often use Internet to look up information (40%), and have their flower stores mainly in Taipei county/city (73.4%). Flower retailers with the above characteristics are the principal research subjects of our study.
Table 9shows the purchase modes and the major busi-ness items of the flower stores being interviewed. The major purchase channels include the wholesale market, brokers, importers, flower suppliers, and others; and the wholesale market is the most popular (73%).
The standard for purchase includes, ‘‘stable flower qual-ity’’, ‘‘goodwill of the seller’’, ‘‘prices of flowers’’, ‘‘suffi-cient breeds of flowers’’, ‘‘ purchase volume’’, and ‘‘transportation ’’. The retailers put much emphasis on quality (6.46 points), so flower brokers and wholesalers should actively and continuously provide flowers of high quality to establish goodwill and to attract flower stores to buy their products.
Table 8
Background of interviewees: retailers
Characteristics of interviewees Statistics
Gender Female (67.3%), male (32.7%)
Age 25–29 (27.6%) 35–39 (10.6%)
30–34 (27.1%) 45–49 (5.5%)
40–44 (15.6%) 50 and over (2%)
20–24 (11.6%)
Level of education Senior high (42.7%) Junior high (7%)
Vocational school (32.7%) Graduate school (2.5%)
College (12.6%) Elementary school (2.5%)
Frequency of internet usage Not often (40%) Almost every day (5%)
Never (26.3%) Once a month (5%)
2–3 times a week (12.5%) Once half a month (0.6%)
Once a week (10.6%)
Service area (Location of flower stores) Taipei city (54.3%), Taipei county (19.1%), Taoyuan county (11.1%), Hsinchu city (9.5%), Taoyuan city (6%)
Table 9
Purchase modes and major business items of the flower stores
Item Statistics
Purchase channels & ratio of purchase volume to total purchase volume
73% from wholesale markets 13% from brokers
7% from importers 5% from flower suppliers 2% from other purchase channels Standard for purchase Stable flower quality (6.46 points) Goodwill of the seller (6.01 points) Price of flowers (5.99 points)
Sufficient breeds of flowers (5.77 points) Purchase volume(5.58 points)
Transportation (5.37 points) Note. 7-point scale – 7 the highest, 1 the lowest
Business items of the flower store
Selling flower (97%) Designing bouquets (90%) Decorating meeting venues (72%) Landscaping (57%)
Giving classes on flower arrangement (23%) Selling on-line (13%)
As for business items, ‘‘selling flowers’’ (97%), ‘‘design-ing bouquets’’ (90%), and ‘‘decorat‘‘design-ing meet‘‘design-ing venues’’ (72.7%) rate high.
Table 10shows that the major breeds sold by the retail-ers are daisies, roses, and orchids.Table 11shows the sales and the delivery modes of the flower stores. The average number of self-owned trucks is 1.38, and flower stores with trucks are able to purchase and deliver goods using their own trucks. When the delivery volume is very large, they entrust the transportation companies to do the service. 55.95% of the flower stores have delivery service, 81.56% of which are delivered by self-owned trucks and 14.46% are entrusted to other transportation companies. The sales are mainly cut flowers (59.85%), and the delivery period is mostly on the same day (43.7%) and/or 2–3 days (43.7%). 19.6% of the flowers are sold through e-commerce, 50% of which are interested in using Internet to sell their prod-ucts, and 7.5% greatly interested. The interviewees not using e-commerce to sell their flowers claim that ‘‘on-line promotion doesnÕt work well’’ (5.29 points) and ‘‘no need to promote online’’ (4.75 points).
Table 12 shows the feasibility and reasons for selling flowers online. All the factors affecting the feasibility of selling flowers successfully online have high scores, viz. ‘‘quality stability’’ (6.32 points), ‘‘quality recognition’’ (6.14), ‘‘freshness’’ (6.09), ‘‘classification-packing stan-dard’’ (5.94), ‘‘transporting and delivering’’ (5.79), ‘‘unit price’’ (5.68), and ‘‘demand’’ (5.46), showing that they are all important for successful business performance. Cut flowers scoring over 5 points include bouquet (5.72), lily (5.50), rose (5.42), carnation (5.16), and million stars (5.05). Pot flowers scoring over 5 points include compound pot flowers (5.72), lucky bamboo (5.68), pachira macro-carpa (5.53) and phalaenopsis (5.42).
The questionnaire lists four possible introductions of flower products in the e-marketplace, viz. ‘‘literal descrip-tion’’, ‘‘picture illustradescrip-tion’’, ‘‘regulated classificadescrip-tion’’, and ‘‘a combination of pictures, literal description, and reg-ulated classification’’. ‘‘Spot check’’ is also listed in order to compare the intervieweesÕ attitude between direct spot check and online literal and pictorial description. Table 13 shows that ‘‘a combination of pictures, literal descrip-tion, and regulated classification’’ (5.97), and ‘‘spot check’’ (5.81), score highest, showing that the retailers will accept pictures, literal description, and regulated classification even though they do not see the real objects.
With regard to the expectations towards the contents of the e-marketplaces, items scoring over 5 points include ‘‘price in all flower markets’’ (5.79), ‘‘yield in all places of production’’ (5.3), ‘‘collecting trading details of flower
stores’’ (5.11), and ‘‘information about flower suppliers’’ (5.09).
To realize the intervieweesÕ ideal delivery modes, we investigate their ideal delivery time and their evaluation of the transportation companies, and the results are shown
in Table 14. The ideal delivery times are 08:00–11:59
(46.7%), and 04:00–7:59 (37.7%), showing that the retailers prefer to receive the flowers they order before the business hours so they have enough time to arrange or pack the flowers. In choosing transportation companies, they pay great attention to service factors, including the quality of service (6.66), cooperation on urgent orders (6.62), punc-tual delivery (6.58), and efficient delivery (6.52). In fact, all 8 factors score over 5 points, and they should all be taken into consideration.
There are 20 factors which possibly affect the coopera-tion modes between the e-marketplace and the flower store, and they are categorized into four parts: logistics flow, business flow, cash flow, and information flow. The
inter-Table 10
Major breeds of flowers purchased and their purchase rate
Daisy Rose Orchid Lily Tulip Other breeds
Average purchase rate (%) 38.6 36.9 33.2 32.2 28.6 29.3
Note. Average purchase rate = purchase volume of a particular breed‚ total purchase volume. Table 11
Sales and delivery modes of flower stores
Item Statistics
Number of self-owned truck 1.38 on average
Delivery ratio Ratio of flowers delivered to total sales volume, 55.95%
Delivery mode Ratio of flowers delivered by self-owned trucks to total volume delivered, 81.56% Ratio of flowers delivered by entrusted vehicles to total volume delivered, 14.46% Sales Ratio of cut flower sales to total sales,
59.85%
Ratio of pot flower sales to total sales, 25.74%
Ratio of accessory sales to total sales, 10.09%
Ratio of other sales to total sales, 4.32% Delivery period On the same day (43.7%)
2–3 days (43.7%) 4–6 days (10.1%) 7 days–4 weeks (2%) Over one month (0.5%) Trading through e-commerce Yes (19.6%), No (80.4%) Interest in trading through
e-commerce
Yes (50%), No (42.5%), Yes, greatly interested (7.5%)
Reasons for not using e-commerce
On-line promotion does not work well (5.29 points)
No need to promote online (5.29 points) Business scale is not large (4.75 points) Unfamiliar with computer (4.66 points) Note. 7-point scale – 7 the highest, 1 the lowest.
viewees are asked to rate on a 10-point scale, 10 being the most important and 1 the least important. We use Fuzzy Delphi to calculate the geometric mean of each factor and the results are shown inTable 15.
Fig. 3 shows the scores of all the factors in linear illus-tration and we classify them into 5 groups. The three key factors which affect the cooperation modes between the retailer and the e-marketplace are ‘‘cooperation on urgent order’’, ‘‘accuracy of order processing’’, and ‘‘order pro-cessing efficiency’’. We suggest that the e-marketplace should put more emphasis on how to correctly and speedily process orders from the retailers and cooperate on the urgent orders by immediately dispatching cars for delivery. To analyze the feasible operation modes that the e-mar-ketplace should use to cooperate with the flower stores, we list three operation modes for evaluation: 1. Actively plac-ing orders: After becomplac-ing members of the e-marketplace, the retailers can search all production-marketing informa-tion on the e-marketplace, and then relate to the supplierÕs type, class, volume, trading time and place, and the ideal price of the flower products. The supplier who is interested in this order can directly contact the retailer directly and negotiate on the price. 2. Jointly negotiating prices: The e-marketplace first relates the class, trading time, and vol-ume of the flower products to all members so they are able to register the class, trading time, volume and place for col-lection; the bigger the order, the more the discount. 3. Free bidding: Suppliers offer retailers on the e-marketplace the class and the lowest bidding price of their products. Within a stipulated period of time, members place beds on the products.
We employ Fuzzy Multiple Criteria Decision Making to evaluate the compatibility of the three operation modes. Table 16shows the results: the scores of are very close, indicating that the flower retailers find them all equally compatible, probably because currently there is still no e-marketplace formed based on the expectations of the retailers, and the retailers are not familiar with the modes. We suggest that the e-marketplace can employ the three modes at the same time to cooperate with the retailers and evaluate the feasibility of each of the modes in the process.
Table 12
Feasibility and reasons for selling flowers online
Item Statistics
Factors affecting the feasibility of selling flowers online Quality stability (6.32) Transporting and delivering (5.79) Quality recognition (6.14) Unit price (5.68)
Freshness (6.09) Demand (5.46)
Classification-packing standard (5.94)
Feasibility of selling cut flowers online Bouquet (5.72) African Daisy (4.70)
Lily (5.50) Gladiolus (3.90)
Rose (5.42) Chrysanthemum (3.50)
Carnation (5.16) Other breeds (0.47)
Million Stars (5.05)
Feasibility of selling pot flowers online Compound pot flowers (5.72) Begon· Semperflorens-cultorom (4.74) Lucky Bamboo (5.68) Impatiens balsamina (4.61)
Pachira macrocarpa (5.53) Duranta repens (4.58)
Phalaenopsis (5.42) Others (0.42)
Codiaeum (4.93) Note. 7-point scale – 7 the highest, 1 the lowest.
Table 13
Expectations of Interviewees towards the E-marketplace
Item Statistics
Ideal introductions of flower products
A combination of pictures, literal description, and regulated classification (5.97)
Spot check (5.81) Pictures (5.67)
Literal description (5.48) Regulated classification (5.47) Expectations towards
the contents of the e-marketplace
Price in all flower markets (5.79)
Yield in all places of production places (5.30) Collecting trading details of flower stores (5.11)
Information about flower suppliers (5.09) Related flower activities (4.75)
Related measures from the agricultural authority (4.38)
Note. 7-point scale – 7 the highest, 1 the lowest.
Table 14
Ideal delivery modes for interviewees
Item Statistics Receiving period 08:00–11:59 (46.7%) 0:00–3:59 (4%) 04:00–7:59 (37.7%) 20:00–23:59 (1.5%) 12:00–15:59 (9.5%) 16:00–19:59 (0.5%) Factors in choosing transportation companies
Quality of service (6.66) Attitude towards customer complaint (6.24)
Cooperation on urgent orders (6.62)
Freight charge (6.22)
Punctual delivery (6.58) Packing (5.8) Efficient delivery (6.52) Number of self-owned
trucks (5.48) Note. 7-point scale – 7 the highest, 1 the lowest.
5. Conclusions
The scope of the paper is to explore the possibility of an e-marketplace for the floral industry in Taiwan, and we suggest key factors in forming an e-marketplace for the floral indus-try in Taiwan by analyzing the possible cooperation modes among the supplier, the retailer and the e-marketplace. This study proposes a conceptual framework to analyze the key factors affecting the formation of an e-marketplace, and also employs Fuzzy Delphi, Fuzzy Multiple Criteria Decision Making, and Kano Analysis to conduct an empirical research on the floral industry in Taiwan. The results also show this conceptual framework can be used to analyze the key factors in forming an e-marketplace with products that are difficult to standardize.
Currently, the major marketing channel for flower sup-pliers and retailers in Taiwan is the flower wholesale
mar-ket. However, when the retailers make purchases in the wholesale market, the dominant suppliers offer poor ser-vice, and the retailers find it inconvenient to collect infor-mation on the price of flowers. Our study shows that the e-commerce mechanism of the e-marketplace can improve trading efficiency and lower the cost of collecting informa-tion as well as the purchase price. According to our analy-sis, the e-marketplace can use ‘‘a combination of pictures, literal description, and regulated classification’’ to intro-duce the quality of flower products. By Fuzzy Delphi, the key factors which affect the operation modes between the retailer and the e-marketplace are ‘‘cooperation on urgent orders’’, ‘‘accuracy of order processing’’, and ‘‘order pro-cessing efficiency’’. Then, based on the three key factors we use Fuzzy Multiple Criteria Decision Making to find what operation modes the e-marketplace should take to cooperate with the retailer. Retailers find the three opera-tion modes ‘‘actively placing orders’’, ‘‘ jointly negotiating prices’’, and ‘‘free bidding’’ equally compatible, so we sug-gest that the e-marketplace should provide these modes at the same time for retailer use and later the retailers can adjust the modes according to their business performance. Besides, by Fuzzy Delphi, the key factors affecting the cooperation modes between the supplier and the e-market-place are ‘‘quality check’’, ‘‘ production project’’, ‘‘ trading
Table 15
Factors affecting the cooperation modes between the e-marketplace and the flower store
Category Serial number Factor Average
Logistics flow 1 Cooperation on urgent orders 9.2
2 Sufficient supply 8.84
3 Good stock management 8.77
4 Diverse breeds 8.72
5 Quality check 8.69
6 Efficiency of car dispatch 8.48
7 Form and amount of compensation 8.33
8 Freight charge 8.3
9 Quantity and capacity of cars 7.45
10 Flower-packing and staff training 6.15
Cash flow 11 Trading credit investigation 8.76
12 Whether capital is sufficient 8.37
13 Processing fee 8.07
Information flow 14 Provision and compilation of establishing trading mechanism 8.23
15 Production-marketing information 8.12
16 Trading price prediction 8.12
17 Efficiency of deciding on trading price 8.03
Business flow 18 Accuracy of order processing 9.46
19 Order processing efficiency 9.06
20 Goods tracking 8.63
6
8
9
(5) (6) (8) (1) (2) (3) (11) (7)7
(4) (9) (10) (12) (13) (14) (15) (16) (17) (20) (19) (18)10
Fig. 3. Linear illustration of scores of all factors.
Table 16
Compatibility of the operation modes of the e-marketplace
Operation mode Compatibility
Actively placing orders 0.92089
Jointly negotiating prices 0.92086
credit investigation’’, and ‘‘accuracy of order processing’’. We then employ Fuzzy Multiple Criteria Decision Making to evaluate the operation modes that the e-marketplace should take to cooperate with the supplier. ‘‘ Contract pro-duction’’ and ‘‘ free bidding’’ are the preferred operation modes. Using Kano analysis, we realize the supplierÕs atti-tude towards all the preferred operation modes. To attract more suppliers to join and trade on their websites, the e-marketplace should upgrade on-line promotion. In addi-tion, the e-marketplace should integrate all orders and then purchase flowers from suppliers, and set up a clear and fair arbitration mechanism to deal with quality check, damag-ing goods, and compensation; and, in order to attract sup-pliers to surf their websites, the e-marketplace should allow them to place orders online and offer as much production-marketing information as possible.
Acknowledgments
The authors appreciate the treasured opinions from two anonymous reviewers. They really help to make the paper perfect.
References
[1] COA, Executive Yuan, Profile of Agricultural Production-Marketing Classes in Taiwan, printed by the Agency of Agriculture and Forest of the Taiwan Province, September, 1996.
[2] Hsiao-Lin Chen, Study on the Evaluation Procedure of the Airport Location Selecting, Essay of Graduate School of Department of Transportation & Communication Management Science, National Cheng Kung University, June, 1994.
[3] Tzong-Ru Lee, Hui-Fang Cheng, Analysis of the Business Environ-ment of Flower Stores and Making the Channel Strategy, Part 2, 155 Issue of Taiwanese Flower Horticulture, July, 2000, pp. 30–31.
[4] Hui-Fang Cheng, The Application and Study of Decision Theory in Supply Chain Management, Essay of Graduate School of Depart-ment of Agriculture Marketing, National Chung Hsing University, 2000.
[5] Hui-Fang Cheng, The Application and Study of Decision Theory in Supply Chain Management-A Example of Taipei Agricultural Prod-ucts Marketing Corporation, Department of Agricultural Marketing National Chung-Hsing University, 2000.
[6] D.P. Clausing, Total Quality Development: A Step-by-Step Guide to World-Class Concurrent Engineering, New York, 1994.
[7] A. Kaufmann, M.M. Gupta, Fuzzy Mathematical Models in Engi-neering and Management Science, Elsevier Science Pub. Co., New York, 1988, Sole distributors for the U.S.A. and Canada.
[8] D. Kenny, J. Marshall, Contextual marketing: The real business in the Internet, Harvard Business Review 78 (6) (2000) 119–125. [9] Tzong-Ru Lee, A new channel for selling flowers – an application of
Internet, Journal of the Agricultural Association of China I (2) (2000) 183–196.
[10] X. Luo, M. Seyedian, Contextual marketing and customer-orienta-tion strategy for e-commerce: an empirical analysis, Internacustomer-orienta-tional Journal of Electronic Commerce 8 (2) (2003) 95–118.
[11] B. OÕKeefe, C. Loebbecke, Introduction to the special section: strategies for furthering electronic commerce, International Journal of Electronic Commerce 7 (1) (2002) 91.
[12] B.T. Ratchford, D. Talukdar, M.S. Lee, A model of consumer choice of the internet as an information source, International Journal of Electronic Commerce 5 (3) (2001) 7–21.
[13] P. Schubert, Extended web assessment method (EWAM): evalua-tion of electronic commerce applicaevalua-tions from the customerÕs viewpoint, International Journal of Electronic Commerce 7 (2) (2002) 51–80.
[14] T.L. Satty, Exploring the interface between hierarchies, multiple objective and fuzzy set, Fuzzy Sets and System 1 (1978) 57–68. [15] C. Steinfield, H. Bouwman, T. Adelaar, The dynamics of
click-and-mortar electronic commerce: opportunities and management strate-gies, International Journal of Electronic Commerce 7 (1) (2002) 93– 119.
[16] P. Zhang, G.M. von Dran, User expectations and rankings of quality factors in different web site domains, International Journal of Electronic Commerce 6 (2) (2001) 9–33.
