Conjoint Analysis: A Study of Canned Coffee in Taiwan

Transcription

1 238 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August 2008 Conoint Analysis: A Study of Canned Coffee in aiwan Meng-Long Si 1, Cen-Yin Liu 2, Biing-Wen Huang 3, Souua Lin 4, Ke-Cung Peng. 5 1 Dept. of Social Studies Education, National aitung University, aiwan 2 Graduate Scool of Management, I-Sou University, aiwan 2 Department of Information Management, zu-hui Institute of ecnology, aiwan 3 Dept. of Applied Economic, Cung-Hsing University, Corresponding Autor, aiwan 4 Applied Foreign Language Department, National Kaosiung University of Applied Sciences, aiwan 5 Department of Agribusiness Management for National Pingtung Science and ecnology University, aiwan Summary is paper intends to explore consumer preferences for canned coffee attributes, to determine te optimal combination for consumers, and to provide manufacturers a reference for teir marketing strategies. In tis study, consumers in aiwan were divided into several demograpic regions (Nortern, Central, and Soutern areas) and individual preferences for instant coffee attributes were compared. By adopting conoint analysis we found price was te maor concern for all consumers. In addition, consumers in Nortern and Central areas cared more about flavor, wile Souterners more about te brand name. Consumers in aiwan sowed significantly demograpical difference in teir preferences over te combination of canned coffee attributes. Key words: conoint analysis, canned coffee, product attributes, consumer preferences 1. Introduction Opportunity recognition and exploitation are te main concern for new product development [1] and entrepreneursip [2]. However, different customer segments in te market ave divergent perceptions regarding product attributes. Smit et al. [3] suggest te subtypes of value tat can be created from different value creating processes. For example, Starbucks creates functional value mainly via product attributes and appropriate features (suc as product taste, package, customization, ot drinks for cold days, and cold drinks for warm days). A. C. Nelson, a survey to global retail information, found drinks were categorized as te fastest-growing products in te world. e survey covered 47 countries and markets in Nort America, Europe, Mideast, Africa, Asia-Pacific Region and Latin America and found drinks industry as one of te prevailing industries. Among all drinks, liquors, teas and coffee serve as te tree maor dominant drinks. As francised cains, suc as Starbuck Coffee, Yes Coffee, Danty Coffee, and Seattle Coffee, enter aiwan, te population of coffee drinkers grow significantly. Instant coffees, ree-in-one or wo-in-one, ave sown double-digit growt rate and vigorated coffee drinks market [4]. Coffee as a drink as claimed a place in every ouseold, as various coffee products are selved in convenience stores nationwide. Evidently, new products are introduced to oin te ever-fierce market competition. In recent years, more and more companies oined te ever-eterogeneous instant coffee market competitions. Consumers are faced wit a variety of coices over brands, taste, packages, content volumes, and prices. Wat prompts te consumers in selecting a certain coffee product? Do tey demonstrate specific preference for teir purcases? Wat influences teir purcase beavior? Wat are te attributes for te coffee products tey purcase? Hence, a study of te combined attributes of te factors for purcasing teir coffee products will determine te preferred instant or canned coffee products. e purpose of tis study is to determine te caracteristics and priority of consumer preference structures; to segment te various preference structures and caracteristics of demograpic variables of consumer s groups; and to determine attributes combination of preferred instant coffees for various structural consumer groups as reference for manufacturer s marketing segmentation of te instant coffees. 2. Literature Review Relevant studies of coffee predominantly investigated consumer beaviors. For example, seng [5] surveyed neigboring consumers of convenience stores in aiwan and found tat 67.8 percent of surveyed consumers of instant coffee were between 20 and 29 in terms of age level. e preferred brand was Mr. Brown, wic enoyed te predominant market sare and te longest istory in te market. Lin [6] also pinpointed, consumers of age 15 to 30 accounted for 66.5 percent of sale in instant coffee, particularly canned coffee. Consumers preferred canned coffee over coffee in etra-pak because of better taste and adequate supply. Mr. Brown coffee enoyed te most preferred brand because of iger exposure troug commercials and advertisements. Manuscript received August 5, Manuscript revised August 20, 2008.

2 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August Lu and Hung [7] studied te preference for instant coffee and consumption caracteristic demonstrated by young consumers. ey found tat over eigty percent of te young consumers of age between 19 and 22 preferred instant coffee. eir preferred brands included Mr. Brown, La Gauce de la Seine, and Muster, available mostly at convenience stores. eir taste-oriented preferences for instant coffee were basically focused on te variety, brewery, and taste. However, some brand-oriented young consumers were easily affected by te promotion, advertisements, commercials, and packaging; wile some oter price-oriented yout groups were concerned about price and volume. In a study of college students in te Metro-aipei, Su and You [8] found te young consumers of instant coffee did not seem to care about ingredient indication; teir purcases were critically influenced by preoccupied product image and reputation, even te package patterns and colors. In anoter study, Cristovam et al. [9] investigated te gender caracteristic of preference for Italian concentrated coffee. e subects were 100 consumers wo consumed six brands of ric coffee and demonstrated te gender difference in preference for Italian coffee and consumption patterns of milk. Consumers seemed to accept blend coffee and reflected teir cange in consumer preference. Heidema and ong [10] undertook a covariance analytic study of perceptive caracteristics of eigt brands of coffee in five European countries and found consistent variations in samples, nationality, gender and age, werein gender and age sowed significant impact upon brands of coffee, and te content volume of four brands of coffee sowed mutual influence. Besides, De ong [11] conducted a study of perceptive caracteristics among European consumers of eigt brands of coffee and found te expressive variation of product structure reaced an average of 92 percent and sowed attribute consistency in most products. e significant relevance in product attribute resulted in cognitive description of consumer preferences in various nations. Amed et al. [12] examined te influence of COO (country of origin) relative to product attributes of bread and coffee in Singaporean s preferences of domestic and foreign food products. e results suggested tat in te presence of some specific product attributes suc as price and brand, te impact of COO was weak. Pelsmacker et al. [13] investigated te etical consumption beavior troug consumer s coffee-buying decision. Consumers in Belgium sow teir preferences and make trade-offs between different attributes of a coffee product by simulating real marketplace situations. It was found tat consumers in different age ad te extent difference in willing to pay for te fair-trade attribute wen buying coffee. erefore, a product or service based upon consumer-oriented attributes will become pivotal. From afore-mentioned literature, few studies ad undertaken on te overall attributes of instant coffee from consumer preference perspective. It is suggested tat preferred coffee products be developed for potential markets wen te favorable combined attributes of consumers and consideration factors toward purcase of instant coffee products can be determined. 3. Proposed Metodology In traditional teory of supply and demand, effect is a function of a product; terefore, price is a considering factor for interpreting consumer demand. However, wen consumers face a purcase decision of a certain product, tey are confronted wit various combinations of product attributes for a eterogeneous product. e uman purcase decision involves multiple considerations, so consume preference is a function of product attributes rater tan a function of te product. is study adopted conoint analysis to determine te caracteristics of instant coffees as te basis for analyzing consumer preference for instant coffees and for furter investigating te variations between subects in various regions and market segmentation. Green and Srimivasan [14] indicated tat conoint analysis is an analytic metod wit decompositional approac for evaluating te preference structure of te known testees and overall assessment. ree assumptions are made: First, tere is trade-off relation between attributes; secondly, te overall value of te testees is added; and tirdly, no interactional outcomes upon te overall effect will be considered. Hence, te overall effect is yielded by adding up te cost effect of eac attribute and represented as: U = V = V + V + + V 1 2 L = 1 U represents te overall effect of te -t Were product V 1 + V2 + L + V represents te attribute effect value of 1, 2,, in product. e applications of conoing analysis are widely found. Hair et al. [15] suggested, if te proposed attribute bencmark is not available in te market, ten attribute combination of consumer preference can be determined and te market sare for products of various attribute combination can be estimated. Consequently, researcers may determine wit conoint analysis wicever attributes are crucial for consumers and wic products wit watever attribute combination are most acceptable by consumers. Green and Rao [16] ad applied conoint analysis to te realm of marketing for determining te conoint impact of

3 240 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August 2008 two or more variables upon variable ranking. us, te conoint analytical mode can be represented as: Y = V k + k k + + '... = 1 p ' Were Y represents an overall estimate of consumer toward a product, tat is, te composition effect value of product. represents attributes of te product, = 1,2,...,. k represents te attribute bencmark of te product, k = 1,2,..., K. V k = 1 of te product. p ' k K represents te main effect of attribute bencmark ' k represents te interaction effect of two attribute bencmark of te product K represent te interaction effect of multiple attribute bencmarks. Fractional factorial design is often adopted in te conoint analysis, werein te main effect and te interactional effect of te two attributes bencmarks are retained [17]. us, equation can be simplified as: Y = V + k = 1 p ' k ' k Green and Srinivasan [14] specified te processes of a conoint analysis as: 1. Coose preference mode; 2. Data collection; 3. Construct contoured subects; 4. Describe subects; 5. Measure variables; and 6. Evaluate parameters, as sown in able 1. able 1 Processes and modes of conoint analysis Analysis Process Metod 1. Coose vector models, ideal-point preference models, part-wort utility mode models 2. Data collection 3. Construct contoured subects 4. Describe subects 5. Measure variables Bifactorial (trade-off), Full Concept Metod (concept assessment) Fractional factorial design, Balance incomplete block design, Partially balance incomplete block design textual description, narration, picture demo, obect demo Econometrics scale (ratio scale, Interval scale) Non-econometrics scale (ordinal scale, nominal scale, Pairwise Comparison scale) 6. Evaluate MONANOVA, PREFMAP, parameters LINMAP, LOGI, PROBI Source: Green and Srinivasan, 1978; Yuan, Researc Design and Pragmatic Model 4.1 Determination of Attributes of Canned Coffee is study is based upon te Yearbook 2003 of IC (Integrated Consumer endency) Marketing, surveyed 386 subects wo purcased coffee drinks in te past tree monts, including 191 Norterners, 88 Central Residents, 91 Souters, and 16 Easterners in various counties of aiwan. Results of factors of purcase are broken down as able 2. able 2 Factors of instant coffee purcase aiwan Nort. Central. Sout. East. aste/flavor 55.4% 48.7% 59.1% 69.2% 37.5% Name brand 47.9% 50.3% 43.2% 47.3% 50.0% Price 22.0% 21.0% 23.9% 14.3% 25.0% Package 20.3% 13.0% 22.8% 18.7% 6.3% Content vol. 9.6% 11.5% 10.2% 5.5% 6.3% Samples Source: IC Marketing Yearbook e top five factors to be considered for coffee purcase are taste, name brand, price, package, and content volume. erefore, te five factors contribute to crucial attributes for consumers to purcase instant or canned coffee. As te IC Yearbook [18] indicated, te top four name brands are Mr. Brown, Wincafe, UCC and 36 Francs, wile te dominant consumed packages are in can, tetra pak, and cup. e most preferred taste of instant coffee is blend, Blue Mountain, Cuppuccino, and Mandeling. Most packages contain a volume capacity between 120 cc and 500 cc. No standardized categorization was found. is study took 250 cc as a criterion for categorization: less tan 250 cc and more tan 250 cc. e price of te canned or instant coffee, taking te market price as a basis and te pricing policies, as suggested in Wang s (2003) study, for 250 cc cupped coffee (La gauce de la Seine from Unipresident Corp.) as reference, ranges from N$20 to N$30 (or USD $0.69 to 0.99). In tis study, te levels of price attributes are: N$15-20 (USD $0.49 to $0.69), N$20-25 (USD $0.69 to $0.89), and iger tan N$ 25 (USD $0.99), respectively. 4.2 Questionnaire Design e questionnaire includes two parts. e first part contains personal information as variables for market segmentation; wile te second part is a survey based upon

4 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August conoint analysis for determining te preference ranking of test properties. Data in tis study are collected wit Full Concept Metod; te combined attributes of all tested products are recorded on a card. e attributes of name brand, package, taste, content volume, and price construct a matrix of items. at means a combination of 288 Stimuli. e use of ortogonal arrays wit te SPSS (Statistical Package for te Social Science) For Windows 12.0 reduced te test items from 288 to 16 (as in able 3) and produces a card for easy testing of te testees and better reliability of te answering te survey. Item able 3 16 cards of test properties matrix produces cards of test properties matrix brand package flavor volume Price N$ USD $ 1 Brown can general < 250cc 15~ francs can general > 250cc 15~ Brown can Mandeling < 250cc >25 >.89 4 Brown etra Pak Blue mountain > 250cc 15~ UCC cup blend > 250cc >25 > francs cup Blue mountain < 250cc 15~ Wincafe can Blue mountain > 250cc >25 >.89 8 Wincafe can cappuccino > 250cc 15~ Wincafe cup Mandeling < 250cc 15~ UCC can Blue mountain < 250cc 20~ UCC can cappuccino < 250cc 15~ francs can Mandeling > 250cc 20~ francs etra Pak cappuccino < 250cc >25 > UCC etra Pak Mandeling > 250cc 15~ Brown cup cappuccino > 250cc 20~ Wincafe etra Pak general < 250cc 20~ Data Sampling e subects (instant coffee consumers) in tis study were from various demograpic regions of aiwan, namely: Nortern, Central and Soutern regions, represented by residents of aipei City and County,.aicung City and County and Kaosiung City and County, respectively. Qualified subects were tose wo ad purcased instant coffee witin te past tree monts. In a conoint analysis, te size of sampled subects is not specified. Cattin and Wittink [19] suggested a size ranging from 100 to 1,000; and in most typical studies, a size between 300 and 500 is most common. Akaa and Korgaonkar [20] even suggested a sample size of less tan 100, but insisted te representativeness of te entire target population. e sample size in tis study is 385, including 209 Norterners, 95 Central residents and 91 Souterners. Surveyors were sent for collecting data by random sampling at primary outlets of instant coffee, including convenience stores and wareouse stores, and supermarkets. 4.4 Constructing and Evaluating Pragmatic models is study added up te total effect values of test items by adopting a model of composition effect function. e effect mode can be represented as: 5 = 1 ( X ) U = V k k e overall effect of consumers to product, U could not be observed and was terefore unavailable. However, tere was a linear relationsip wit te estimated outcomes of consumers [21]. us, te overall effect is represented as: U = a + by + e Can be rewritten as: 1 Y = ( a + U e ) (6) b Were Y represents te estimated observable outcome of consumers and e is zero value of normal distribution, te difference is te deviation of te constant. Replace equation (6) wit. According to tis study, an equation of pragmatic model function can be establised by inserting te attribute benc mark of instant coffee, and be represented as: Y 5 k = α + V X + ε (7) = 1 k = 1 k k Were Y is te overall estimate value of test items by consumers, i.e., te composition value of test item, = 1,2,...,16 test items. α denotes te diagonal item; V k denotes estimate parameter; denotes te attributes of product, = 1,2,..., 5 ; k denotes te bencmark of products, k = 1,2,...,k ; X k denotes te k t benc mark of attribute is a virtual variable; and

5 242 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August 2008 ε denotes te error term. After setting te virtual variables as able 4, te following equation of estimation can be inferred as: Y = α + V + 11X11 + V12X12 + V13X13 + V21X 21 V22X 22 + V 31X 31 + V32X 32 + V33X 33 + V41X 41 + V51X 51 + V52 X 52 + ε (8) Wit ordinary least square regression (OLS), te regression coefficient α and te various attribute composition effect V k are obtained and defined as follows (able 4). able 4 Definition of attribute benc mark for instant coffee brand ( X, X X ) 11 12, package ( ) 21,X Brown Wincafe UCC 36 francs Can X etra pak cup flavor X, X, ) Content vol. ( X 41 ) ( X33 price ( X 51,X 52 ) Source: tis study blend Blue mountain cappuccino Mandeling < 250cc > 250cc N$15~20 (US$.49~.69) N$20~25 元 (US$.69~.89) > N$25 (> US$.89) (1,0,0) (0,1,0) (0,0,1) (0,0,0) (1,0) (0,1) (0,0) (1,0,0) (0,1,0) (0,0,1) (0,0,0) (0) (1,0,0) (0,1,0) (0,0,0) After obtaining te composition effect of various attribute benc mark, te relative importance of all attributes can furter be obtained. e relative importance means te value of maximal effect minus te minimal effect of attribute divided by te sum of te maximal effect minus te minimal effect of all attributes and can be represented as: RIA = A = 1 ( max( U ) min( U )) ( max( U ) min( U )) A (9) Were RIA representing te relative importance of attribute, and A represents attribute. 5. Analysis of Empirical Results 5.1 Composition Effect of Instant Coffee e composition effect of all attributes of instant coffee for consumers in all regions of aiwan is specified as (able 5): blend coffee was most preferred flavor wit effect value of Wile Mr. Brown was te most welcome brand, wit an effect value of Cup coffee, priced more tan N$25 (US$ 0.89) wit content volume more tan 250 ml was most preferred by consumers. e most unfavorable combination of attributes for instant coffee was: 36 Francs (brand), canned, cappuccino, less tan 250 ml and priced between N$15~20 (US$ 0.49~0.69). able 5 Cost-benefit of instant coffee attribute bencmark in all regions Brand Package Flavor Content Volume Price Nationwide Nortern Central Soutern Brown Wincafe UCC francs can etra pak cup bend Blue mountain cappuccino mandeling <250cc >250cc N$15~20 US$.49~ N$20~25 US$.69~ >N$25 >US$ e most preferred combination of attributes for instant coffee consumers in Nortern aiwan was: Falvor: Mandeling wit effect value of 3.25; brand: Wincafe,

6 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August ; price: over N$25 (US$ 0.89); package: etra Pak; contents: more tan 250 ml. e most unfavorable combination of attributes for instant coffee was: 36 Francs (brand), canned, cappuccino, less tan 250 ml and priced between N$15~20 (US$ 0.49~0.69). In Central region, te most preferred combination of attributes for instant coffee was: Flavor: blend wit effect value of 3.5; brand: Wincafe, 1.75; price: over N$25 (US$ 0.89); package: cup; contents: less tan 250 ml. e most unfavorable combination of attributes for instant coffee was: 36 Francs (brand), canned, Mandeling, more tan 250 ml and priced between N$15~20 (US$ 0.49~0.69). In Soutern region, te most preferred combination of attributes for instant coffee was: Brand: Mr. Brown, wit effect value of 3.5; price: over N$25 (US$ 0.89), wit effect value of ; flavor: cappuccino; package: canned; contents: more tan 250 ml. e most unfavorable combination of attributes for instant coffee was: 36 Francs (brand), tetra pak, blue mountain flavor, less tan 250 ml and priced between N$15~20 (US$ 0.49~0.69). It wort noting tat te preferred attribute combination in aiwan, Nortern, Central, and Soutern regions did not appear in te predesigned 16 cards for te enlisted product attributes. is illustrated te advantage of conoint analysis and consumers desirable products can be redesigned by recombining te preferred attributes. 5.2 Relative Importance of Instant Coffee It is known from able 6 tat te nationwide priority rankings for purcasing instant coffee are: price (28.7%), brand (24.62%), content volume (22.17%), package (14.72%), and flavor (9.71%). As sown, price was te maor concern for te purcase of instant coffee, wile packaging and flavor were not te decisive factors. erefore, manufacturers sould pay more empasis upon pricing policy and corporate image tan packaging materials and flavors. able 6 Relevancy of instant coffee in various regions in percentage (%) Brand Package Flavor Content Price Nationwide Nortern Central Soutern Note: te numbers witin parenteses means te preference ranking Consumers in te Nortern region demonstrated teir preference ranking of product attributes as: flavor (42.07%), brand (28.50%), price (13.82%), package (12.89%) and lastly, contents (2.71%). at is, te Norterners concern primarily te flavor, rater tan package and content volume in purcasing instant coffee. Consumers in te Central region demonstrated teir preference ranking of product attributes as: flavor (32.12%), brand (30.66%), price (16.79%), price (11.86%), and lastly, contents (8.76%). at is, te Central residents concern primarily te flavor, rater tan te price and content volume in purcasing instant coffee. Consumers in te Soutern region demonstrated teir preference ranking of product attributes as: brand (42.74%), price (28.34%), flavor (12.57%), package (8.80%) and lastly, contents (7.54%). at is, te Souterners concern primarily te name brand, rater tan package and content volume in purcasing instant coffee. o summarize, flavor is a maor concern for consumers in te Nortern and Central regions, wit relative importance of te Nortern region (42.07%) iger tan tat of te Central region (32.12%). at indicated Norterners paid ig attention to flavor wen purcasing instant coffee, wit Norterners preference over Mandeling and Central consumers in favor of blend flavor. Name brand serves as te secondary attribute concern in aiwan nationwide, and in Nortern and Central regions for instant coffee consumption. However, Souterners deemed name brand te primary concern, wit relative importance of 42.74% in contrast wit for Central, 28.50% for Nortern regions, and national average 24.62%. Souterners preferred Mr. Brown, wereas Central and Nortern consumers preferred Wincafe. Consumers in Nortern, Central, and Soutern regions considered te content volume te least important attribute for purcasing instant coffee wit Nortern region sowing te lowest relative importance of 2.71%. Nortern and Soutern consumers of instant coffee preferred content volume over 250 ml, wile te Central consumers less tan 250 ml. Price is te first considering factor for consumers nationwide, owever not for consumers in Nortern, Central, and Soutern regions upon purcasing instant coffee. is revealed te advantage of conoint analysis. Consumer preference for product attributes can be identified wen regions were subdivided. Consumers nation widely preferred pricey instant coffee, wit unit price of more tan N$25 (US$.89) as preferred coice. e finding can serve as a reference for instant coffee suppliers for pricing policies. 5.3 Ranking of te Stimuli e total effect of all test item cards can be obtained by adding up te compositional effect of eac attribute. e total effect te 16 product combinations are sown as in

7 244 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August 2008 able 10. On a national average, card number five revealed te igest total effect of ; wile card number 11 te least total effects of In anoter word, te test attributes on card number five were most preferred, wile tose on card number 11 te least. Assume tere are 16 instant coffee products on te market, te consumer nationwide will most prefer UCC, cupped, blend flavor wit content volume over 250 ml and N$25 (US$.89). In contrast, canned UCC cappuccino flavor less tan 250 ml and under N$25 (US$.89) is least preferred. Nortern consumers gave product combination on card number 16 te igest total effect. at is, tey preferred product number 16 most, and te number 11 least. And Central consumers preferred number 5 most, and number 12 least; Soutern consumers preferred number 3 most, and number 6 least. In general, product combination number 5 is te most enoyable instant coffee for all regions, Nortern, Central, and Soutern regions (witin te top tree in all regions). e composition effect of product attribute is a relative value and can be compared only wit te effect values of all products in a group or by a consumer, rater tan wit te preference of various groups and consumers. o investigate te segmentation difference of all regions, tis study will explore te difference between te effect value of all segmentation attributes and teir relative importance in terms of teir ranking order. 6. Conclusions is study intended to investigate te preference for instant coffee demonstrated by regional consumers of te nation. e findings were: Price of instant coffee products was consumers primary concern nationwide, followed by te brand, capacity, packing materials, and finally te taste. Consumers preferred Mr. Brown most, packed in cups, blend flavor, over 250 ml in contents and price over N$ 25 (or US$.89). In terms of regional difference, Norterners concerned most about te flavor, followed by name brand, price, package, and lastly content volume of te instant coffee products. e most preferable attributes were Wincafe, in tetra pak, Mandeling flavor, volume more tan 250 ml and price more tan N$25 (US$.89). In Central region, consumers concerned most about flavor, trailed by name brand, package, price and content volume. ey most preferred Wincafe, cupped, blend flavor, wit content volume less tan 250 ml and price over N$25 (US$.89). In Soutern region, consumers concerned most about name brand, followed by price, flavor, package, and content volume. ey most preferred Mr. Brown, canned cappuccino, wit content volume more tan 250 ml and price over N$25 (US$.89). e following conclusions can be drawn from te researc findings: Consumers in Nortern and Central regions sowed consistent preference ranking of instant coffee attributes as flavor, name brand; content volume was least considered attribute in all regions; owever, consumers preferred iger content volume package, wile consumers in Central region preferred instant coffee wit content volume under 250 ml; and consumers in all regions sowed positive effect over iger-priced instant coffee. e findings migt serve as a reference for teir product pricing and design. 7. Suggestions for marketing strategies Identifying market segmentation is an aspect of marketing, and a marketing plan sould address te demands of eac market sector. is study took consumers as start point. e findings sall be a concern for manufacturers and a reference for teir marketing strategies. is study found tat name brand and flavor were te two igly considered items by consumers in all regions for te purcase of instant coffee products. Hence, manufacturers sould view igly of te two attributes wen promoting te products to meet demands of local consumers. Corporate image sall be promoted to impress and attract consumers into purcasing various products Manufacturers sould undertake surveys and test consumers for any new flavor of products being marketed, so tey can understand te acceptance level of te new products and energize refresness into consumers life. is study found, te most acceptable price was over N$25 (US$.89), followed by N$20~25 (US$.69~.89) and N$ 15~20 (US$.49~.69). e researcers of tis study suggested te coffee product be priced over N$25. However, te pricing of te product sall remain flexible wen tere is gap between consumer perception and suggested retail price to ensure optimal acceptability. In terms of packaging, consumers preferred instant coffee products available in cup and cans and wit content volume over 250 ml. Manufacturers sall take tis finding as teir guiding post for marketing new products. In terms of locations of purcasing instant coffee, convenience store ranked as number one, followed by grocery center and supermarkets. Consequently, te pat of sale sall take convenience store as primary concern. Only eigt percent of te new products in te food industry will be marketed, and about 17 percent can take a sare in te market. Generally, only 1.4 percent of te new products will succeed and survive (Sloan, 1994). Periodical promotion of new products is sine qua non for sustaining te product line of te manufacturers. e life cycle of any food product is comparatively sort due to fierce competition. Instant coffee, like oter drinks, will not be available in all tastes and flavors. Manufacturers

8 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August

9 246 ICSNS International ournal of Computer Science and Network Security, VOL.8 No.8, August 2008 ave to identify te likes and dislikes of te consumers and ante up te most desired and preferred product to lock in teir consumer groups and flouris. 8. Researc limited Conoint analysis as a number of limitations. Firstly, te number of attributes and levels sould be limited to avoid respondents udging a large number of trade-off tasks, wic could lead to respondent fatigue and less valid response patterns. Secondly, conoint analysis does not give an absolute indication of te importance of attributes in all conceivable circumstances. e importance of attributes and te part-wort utilities tat result from te analysis are relative indications given te framework of attributes studied. References [1] Cooper, R. G.., Winning at New Products. 3d ed., New York: Perseus. [2] Sane, S. and Venkataraman, S., e Promise of Entrepreneursip as a Field of Researc. Academy of Management Review. 25: [3] Smit, B. and Colgate, M., Customer Value Creation: A Practical Framework. ournal of Marketing eory and Practice. 15: [4] Huang, Y. C rend of coffee drinks. Logistic World Magazine, 1, p. 69. [5] seng, R. H Canned coffees. Logistic World Magazine. 1: [6] Lin, P. S Competitive coffee market. Logistic World Magazine, 21, pp [7] Lu, C., and Hung, W. C Young consumers preference for canned coffee. In Proceedings of Symposium of Science. ecnology and Management [8] Su, C. Y., and You, C. W A study of purcasing beaviors of college students over brandname products in Metro-aipei a case study of Unipresident s La gauce de la Seine coffee. ournal of Printing Art. 11: [9] Cristovam, E., Russell, C., Paterson, A., and Reid, E., Gender Preference in Hedonic Rations for Espresso and Espresso-milk Coffees. Food Quality and Preference. 11(6): [10] Heidema,. and de ong, S., Consumer Preferences of Coffees in Relation to Sensory Parameters as Studied by Analysis of Covariance. Food Quality and Preference. 9: [11] De ong, S., Heidema,., and van der Knaap, H. C. M., Generalized Procrustes Analysis of Coffee Brands ested by Five European Sensory Panels. Food Quality and Preference. 9: [12] Amed, Z. U., onson,. P., Yang, X., and Cen, K. F., Does Country of Origin Matter for Low-Involvement Products? International Marketing Review. 21: 102 [13] Pelsmacker, P. D., Driesen, L., and Rayp, G.., Do Consumers Care about Etics? Willingness to Pay for Fair-rade Coffee. e ournal of Consumer Affairs. 39: [14] Green, P. E. and Srinivasan, V., Conoint Analysis in Consumer Researc: Issues and Outlook. ournal of Consumer Researc. 5: [15] Hair,. F., Anderson, R. E., atam, R. L., and Black, W. C., Multivariate data analysis wit readings. 3rd ed, New York: Macmillan. [16] Green, P. E. and Rao, V. R., Conoint Measurement for Quantifying udgmental Data. ournal of Marketing Researc. 8: [17] Green, P. E. and Devita, M.., A Complimentary Model of Consumer Utility for Item Collections. ournal of Consumer Researc. 1: [18] IC Lifestyle Researc Center Yearbook of IC Marketing. aipei: song-yan-se Co. [19] Cattin, P. and Wittink, D. R., Commercial Use of Conoint Analysis: A Survey. ournal of Marketing. 46: [20] Akaa, I. P. and Korgaonkar, P. K., A Conoint Investigation of te Relative Importance of Risk Relievers in Direct Marketing. ournal of Advertising Researc. 28: [21] Louviere,.., Analyzing Decision Making-Metric Conoint Analysis. Quantitative Applications in te Social Sciences, 67, London New Deli: Sage Publications.