IJFS#626_bozza Ital. J. Food Sci., vol 29, 2017 - 377 PAPER PROFILING WINE CONSUMERS BY PRICE SEGMENT: A CASE STUDY IN BEIJING, CHINA W. MU, H. ZHU, D. TIAN and J. FENG * College of Information and Electrical Engineering, China Agriculture University, 209# No.17 Qinghuadonglu, Haidian district, Beijing, 100083, P.R. China *Corresponding author. Tel.: +86 1062736717; fax: 13581585962 E-mail address: fjying0306@163.com ABSTRACT This study aimed to analyze and identify the purchase behavior and preferences of 976 Chinese wine consumers, particularly the differences in their demographics, purchase behaviors and recognition about wine based on a price segmentation categorized as High, Moderate, and Low Spenders. Price segmentation was defined as the price typically paid for a bottle (750 ml) of wine for daily drinking at home. Significant differences among the three segments were discovered based on income, educational background, job category, purchasing places, favorite brands, recognition of the importance of wine attributes and acceptable price for gift-oriented wine. Research on market segmentation based on price about Chinese wine market is limited. The current study provides insight into market segmentation and may serve as basis for future research. Keywords: consumer behavior, consumer preferences, wine market, market segmentation Ital. J. Food Sci., vol 29, 2017 - 378 1. INTRODUCTION The Chinese wine market has expanded in recent years, China produced 1.12 billion liters of wine and consumed 1.6 billion liters of wine in 2015, and China ranked ninth in terms of wine production and fifth in terms of wine consumption in the world (OIV, 2015), thus indicating that China has been a major player in the international market (VINEXPO, 2013). Trade reports show that the red wine sector is growing at a significant rate (EUROMONITOR INTERNATIONAL, 2013). The total volume of wine that China imported in 2015 reached 554 million liters, which is approximately 45% higher than that in 2014 (UN TRADE, 2015). The annual growth rate of the Chinese wine market is 25%- 30% (CHINA DAILY, 2012). However, the per capita wine consumption in China is only 1.17 liters a year; therefore, the wine consumption of each Chinese consumer is not more than two bottles (750 ml/bottle) each year, whereas the world average is 3.32 liters (Wine Institute, 2015). The Chinese wine consumption per capita is far lower than that in western countries, and thus opportunities for growth are vast. Undoubtedly, as one of the emerging and new wine markets in the world, it can be forecasted that Chinese wine market will attract more and more domestic and international wine producers, marketers and researchers. It is widely accepted that firms need to understand their consumers, which in turn requires analysis of all the dynamics and factors influencing consumers’ buying behavior so as to improve their competitive advantages and business performance (GUNAY and BAKER, 2011). Understanding the purchase behavior of Chinese wine consumers is crucial because preferences and behaviors of these consumers significantly affect the development direction of wine markets (CHRISTODOULIDOU and JAMES, 2011). Consumer behavior is a cornerstone in marketing strategy and it involves conceptual dimensions such as decision-making, values, motivations, self-concept and personality, expectations, attitudes, perceptions, satisfaction, trust and loyalty (COHEN et al., 2014). Consumers do not have homogeneous taste preferences (King et al., 2012). Thus, segmenting the market is useful because different people have various needs (DOLNICAR, 2002). Consumers within a segment are relatively similar, and consumers between segments are comparatively different (BRUNNER and SIEGRIST, 2011). Thus, segmentation is a relevant tool in strategic marketing, which helps researchers understand consumers (DOLNICAR, 2002). Market segmentation is defined as the process of dividing the total market into homogenous segments of consumers with similar needs and wants and the segmentation process addresses the needs of each subgroup efficiently (MARSHALL and JOHNSTON, 2010). Market segmentation is widely used in marketing products (RONDANCATALUÑA and ROSADIAZ, 2014). GUILLET and KUCUKUSTA (2016) provide a method for segmenting spa consumers according to their preferences in the selected spa attributes. Chen et al. (2013) use the market segmentation based on consumer motivation to analyze the bed and breakfast industry in Taiwan. PRAYAG et al. (2014) identify three clusters of travelers based on their service expectations, and determine that cultural differences are crucial for heterogeneity in travel behavior. CIRER COSTA (2013) investigates price formation and market segmentation in seaside accommodations. BRUWER and BULLER (2012) provide insights into the consumer behavior characteristics and consumption dynamics of Japanese wine consumers, explore the validity of existing generic consumer segmentation of the Japanese market, and consider its suitability to predict wine-purchasing behavior. VOORHEES et al. (2011) propose a novel perspective to improve segmentation in the hotel industry. Simpson and BRETHERTON (2004) utilize market segmentation based on lifestyle of customers to explore the wine tourism context. Ital. J. Food Sci., vol 29, 2017 - 379 Various studies on wine consumption have been conducted in different countries, but those on market segmentation in the Chinese wine market are scarce. OLSEN et al. (2015) divide wine consumers in the United States into three segments based on consumer variety-seeking behavior, namely, high variety-seeking, moderate variety-seeking and variety avoiders and find significant differences in high variety-seeking consumers compared with moderate variety-seeking and variety avoiders. MOLINA et al. (2015) identify the different segments of wine tourists who visit Spanish wineries. COHEN (2015) tests a segmentation of Hong Kong wine consumers based on the perception of wine labels and finds that approximately 95% of young Hong Kong Chinese wine consumers prefer “elegant contemporary” labels with red as the dominant color. GERGELY and DIETER (2014) test the segmentation of wine consumers based on the usage of sales channels and define six segments that are aggregated into two groups, namely, basic (discount, supermarket and food-retail consumers) and premium (cellar-door, wine-shop and multichannel consumers). TANG et al. (2011) use cluster analysis to divide wine consumers into six segments, namely, price-conscious wine consumers, involved consumers, knowledgeable consumers, image-oriented consumers, indifferent consumers, basic consumers, enjoyment-oriented consumers, and social consumers. BRUWER et al. (2002) identify nine different segmentation variables in wine market studies in Australia: quality, consumption, risk reduction, occasion based, cross-cultural, behavioral, involvement, geography and wine-related lifestyle. Wine knowledge is used in several segmentation studies on wine consumers, such as works of VIGARELLIS (2015), BRUWER and BULLER (2012), and JOHNSON and BASTIAN (2009). The main segmentation variables used in existing studies are mainly related to consumers (interest in wine, knowledge of wine, motivations, sensation-seeking attitude and behavior, cultural variables, values and lifestyle) (MOLINA et al., 2015). However, studies that conduct segmentation based on consumers’ perception of wine by price paid for a bottle (750 ml) of wine for daily drinking in Chinese wine market are non-existent. And according to the Wine Intelligence China Portraits (2015), Developing Drinkers who enjoy wine flavor, regard wine as a major part of their lives, and thereby often drink at home occupied an increasingly large proportion of Chinese wine consumers. Therefore, this topic is a new area for Chinese wine market research. The current study aims to provide insight into the consumption characteristics and preferences of Chinese consumers in wine. Further the study categorizes consumers into three segments, namely, high, moderate, and low Spenders based on the preferred wine price for daily drinking, explores the difference in the three groups in Beijing and provides valuable information and marketing suggestions on the Chinese wine market for wine marketers, wine business owners and other stakeholders. 2. MATERIALS AND METHODS 2.1. Conceptual framework Various studies focused on wine market segmentation (see literature review). Some studies highlighted consumer demographics, such as sex, age, income. Some studies focused on consumer consumption habits, such as consumption occasions, purchase places, channels of obtaining wine information. Other studies focused on consumer preferences about wine attributes, such as quality, brand, flavor and wine labels, and some examined consumer recognition of wine, such as knowledge in wine. The variables that can be used for segmentation can include consumption habits, cognitions, perceptions and acceptability. Therefore, a conceptual model was developed to show and examine the Ital. J. Food Sci., vol 29, 2017 - 380 potential segmentation variables that could be used for the segmentation of the wine market. The conceptual model is presented in Fig. 1. In the current study, price segmentation is defined as the price typically paid for a bottle of wine for home consumption, provides a chance to understand differences among the segments. Figure 1. The conceptual model. 2.2. Questionnaire The questionnaire on the consumption characteristics and preferences of consumers in wine was designed on the basis of the literature to collect information on consumer behavior. The questionnaire consisted of 20 questions, organized into four sections: consumer demographics (gender, age, level of wage, educational background and job category), consumers’ purchase behavior (frequency of wine consumption, occasions for wine consumption and purchasing places), consumer awareness and preference in wine (consumers’ recognition of wine attributes, cognition of the best or worst wine origins and awareness and preference in wine brands), consumers’ acceptable price for wine (acceptable price for daily drinking and gift-oriented consumption). 2.3. Survey In this research, Beijing was selected as the case because it is one of the most developed areas in China, and it is one of the cities with high wine consumption in China (WANG, 2006). Consumers’ demand for wine in Beijing may indicate the future demand trends in China. A cross-sectional survey was conducted in Beijing, from September to October 2014. Consumers’ characteristics were investigated on the basis of consumption behavior and wine preferences. Questionnaire surveys are usually conducted via telephone, online, or face-to-face interviews. The face-to-face interview and online survey were employed in this study to Ital. J. Food Sci., vol 29, 2017 - 381 ensure sufficient sampling, given that respondents’ cooperation is difficult to guarantee through telephone interviews. The process of the two research methods is as follows. Face-to-face interview: ü First of all, six representative wine retailers in Beijing were selected to participate in the survey. Three supermarkets (Merry Mart, Carrefour and Wu Mart) and three wine stores (Guijie branch, Dongzhimen branch and Donglu branch of Cheers, which is a popular wine store in Beijing) were identified. ü Secondly, five undergraduate students from China Agricultural University were chosen as investigators. The students were recruited and trained before the survey. ü Thirdly, the face-to-face survey was conducted by randomly selecting the interviewees. The surveys were conducted between 9:00 and 17:00 and a typical respondent who purchased wine was given approximately 10 minutes to complete the questionnaire. After the five-day survey, 410 questionnaires were obtained. The questionnaire was reasonable and concise, and the survey was supervised by trained investigators. The completion of the questionnaires was preliminarily ideal. All the answered questionnaires were valid. Online survey: ü Firstly, a famous professional questionnaire survey website in China called SOJUMP was used to ensure that random consumers in different areas could participate in the survey. ü Secondly, investigators released the corresponding website through different social media, such as WeChat and QQ to ensure the high respondent turnout of wine consumers in different regions in the survey. The online survey lasted one week and 604 questionnaires were obtained. After disregarding the incomplete or the logically paradoxical questionnaires, 566 valid questionnaires were acquired. The questionnaire used in the online survey was same as that in the face-to-face survey. However, in the online survey, respondents could fill out the questionnaire at a convenient time and without a time limit. And the questionnaire survey website provided storage and downloading functions, so that analysts could examine the results at any time. A total of 1,014 questionnaires were distributed, and 976 valid ones were obtained. The recovery efficiency reached a high 96.3%. 2.4. Statistical methods The collected information from the survey was transformed into statistical variables and processed by SPSS software 19.0. Descriptive statistical method and chi-square analysis testing were adopted to analyze the relation between consumers’ demographics and purchase behavior which includes the frequency and occasion of wine purchase and the demographic differences among price segments. The mean responses with the standard deviation, frequencies and percentages of responses in each category were calculated and presented in tables and graphs. Ital. J. Food Sci., vol 29, 2017 - 382 3. RESULTS AND DISCUSSIONS 3.1. Consumers’ demographics Table 1 shows the consumer demographics by gender, age, wage level, educational background and job category. The sample consisted of 455 women and 521 men. The most common age group was under 25, followed by 26-35. Per capita monthly income was nearly evenly distributed from 2,000 to 10,000. Well educated respondents (undergraduate and above) dominated the sample (74.9%). Staff members in enterprises (44.5%) represent the most common occupation in the job category, and they are followed by students, thus explaining the high educational level and young age of most of the respondents. Table 1. Statistical characteristics of the respondents. Demographics Category %,N=976 Demographics Category %,N=976 Gender Male 53.4 Educational background Junior high school and below Senior high school Junior college Undergraduate and above Government official Staff in enterprises Liberal profession Farmers Education departments Students Unemployed/retired Else 1.7 Female 46.6 Age Under 25 26-35 45.9 36.4 8.6 14.8 74.9 Per capita monthly income (CNY) 36-45 11.8 46-55 4.9 Job category 7.4 Above 55 Under 2000 2001-3000 3001-4000 4001-5000 5001-7000 7001-10000 Above10000 1.0 23.4 11.6 15.7 17.7 15.3 12.0 4.3 44.5 11.9 0.3 4.4 25.0 2.2 4.4 Notes: N, the total number of respondents. 3.2. Purchasing behavior Some studies considered wine as a female product (LI et al., 2011). However, a qualitative study (LIU and MURPHY, 2007) determined that wine buying is a male interest in future Asian markets. Age was also found to be correlated with wine involvement (BRUWER, 2013). Therefore, a chi-square test was used to examine the difference of gender and age on the frequency of wine purchase. A significant difference caused by age was identified (Table 2). Table 2 shows that approximately 18.0% of the men and 15.4% of the women often drank wine. While considering the difference of gender on wine drink, the chi-square test was insignificant, thus indicating that the frequency of wine consumption among men and women does not significantly differ. This result, which is consistent with that of FORBES (2012), suggests that further research could be conducted to explore the effect of gender on the frequency of wine purchases in China. The 26-35 age group dominated the cluster of often purchased wine. The under 25 age group mostly purchased wine occasionally, thereby indicating that young and middle-aged consumers preferred wine. Young people appeared to be the target of wine marketing in China (JENSTER and CHENG, 2008; LIU and MURPHY, 2007). Ital. J. Food Sci., vol 29, 2017 - 383 Table 2. Difference of gender and age on the frequency of wine purchase. Variables Wine purchase frequency (%) (N=976) Chi-square Test often occasionally never F Sig. Gender Male 18.0 73.3 8.7 3.531 0.171 Female 15.4 72.7 11.9 Age Under 25 7.8 75.9 16.3 26-35 21.4 72.4 6.2 36-45 31.3 67.8 0.9 81.735 0.000*** 46-55 27.1 66.7 6.2 Above 55 40.0 60.0 0.0 Notes: N, number of respondents, 3-never, *** Significant at 1% level respectively. Wine is a social product and is consumed in various occasions. However, few Chinese consumers buy wine for home entertainment (LIU et al., 2014). In China, wine consumption is mainly related to gift-giving, business banquets and special occasions such as the traditional holidays like the middle festival, and so on. Income and status-seeking are the main factors affecting wine consumption occasions (SUN et al., 2009; RICHIE, 2007). In this research, a multi topic was designed to investigate the occasions on which consumers consume wine. Occasions on which wine is consumed vary (Table 3). The result showed that 50.8% respondents drank wine during holidays, 41.3% consumers drank wine daily at home, and 30.0% purchased wine as gifts. The results are consistent with those of previous studies (SUN et al., 2009; JENSTER and CHENG, 2008; LIU and MURPHY, 2007). The results of the chi-square analysis (Table 3) indicated that the occasions on which the respondents consumed wine were significantly different among consumers with various income levels. Respondents with an income of below 2000 (23.7%) mostly comprised those who consumed wine during holidays. This result is consistent with that of a previous research, which determined that consumers with low income tended to drink wine on special occasions such as holidays (Sun et al., 2009). For business banquets, consumers whose income were CNY 7001- CNY 10000 (28.8%) seemed to have the money and opportunity to consume wine. The survey determined that consumers most frequently purchased wine is supermarkets (61.5%), followed by wine stores (20.4%). The proportion of consumers who purchased wine online was relatively high (9.8%). The result is consistent with that of LIU et al., (2014), who determined that Chinese wine consumers mostly purchased wine in supermarkets. This result is attributed to two reasons. First, supermarkets are convenient for consumers (LOCKSHIN et al., 2012). Second, price discounts in supermarkets attract consumers who want to save (FLYNN et al., 2010). The result also showed that online shopping for wine was still weak despite the prosperous electronic commerce in China. Many factors contribute to this result, such as the fragileness of wine bottles and false wine information. Ital. J. Food Sci., vol 29, 2017 - 384 Table 3. Influence of income on occasions on which wine is consumed. Variables Consumption occasiona (%) (N=976) Chi-square Test Business banquet Holidays As gifts Daily Others F Sig. Incomeb 0.000*** Below 2000 14 117 60 85 8 2001-3000 5 64 28 35 6 3001-4000 6 84 52 54 2 56.835 4001-5000 14 86 51 84 0 5001-7000 11 75 54 70 1 7001-10000 21 46 33 51 1 Above10000 2 21 14 23 3 Notes: a The question is multiple choice, b Average monthly income per capita. N, the total sample number of respondents, *** Significant at 1% level respectively. 3.3. Awareness and Preference in wine 3.3.1. Consumers’ cognition on the origins of wine The cognition of consumers on the origin of wine is presented in Fig. 2. The result showed that consumers considered Xinjiang Province and the Ningxia Hui Autonomous Region as the best origins of wine, and Beijing City, Tianjin City and Hebei Province were deemed as the worst wine origins in China. The northwest areas are known to be famous wine origins in China (MU and FENG, 2010), this is consistent with the survey results. However, 15% of the consumers deemed Hebei Province as the best wine origin. More than 10% of the respondents determined Gansu, Shandong and Jilin Provinces as the best and worst wine origins. These results indicate that consumers’ limited knowledge or lack of professional knowledge about wine origins. Figure 2. Consumers’ cognition on wine origins in China. Ital. J. Food Sci., vol 29, 2017 - 385 3.4. Consumer awareness and preference in wine brands Figure 3 shows consumers’ familiarity with wine brands. Brand effect is very important to wine. Although the cognitive value of consumers in the existing brands of wine in China does not represent the actual value of the product, cognitive value is closely related to actual value. To some degree, cognitive value reflects the actual value of products (MU and FENG, 2010). Thus, 10 wine brands are considered in the current study and the respondents were requested to choose three brands that they were most familiar with and the three brands they were willing to buy. The results showed that Greatwall (71.6%), Changyu (70.8%), and Dybasty (35.7%) were the three highly cognitive wine brands for Chinese consumers. This result is related to brand history, advertising, and other factors (MU and FENG, 2010). Figure 3. Consumers’ familiarity with wine brands. As shown in Fig. 4, in terms of buying, 39.8% of the respondents considered Changyu as their first choice, followed by Greatwall (26.3%). Forbes et al. (2014) verify that brand name, in the absence of other product information, influences consumer perception of quality and price, and purchase intentions, and that some categories of brand names perform better than others. It is notable that foreign brands (21.1%) is also coming into consumers’ eyes. These consumers consider foreign wine may be purer than that in China. The wine in China culture is underdeveloped, and consumers have limited knowledge about wine, including origins and brands. Most Chinese consumers are novice wine buyers, and tend to buy wines from the famous wine-producing regions they only knows. These consumers believe that wine imported from foreign countries is better than domestic wine (BRUWER, 2002), and they seek little information about other brands. Ital. J. Food Sci., vol 29, 2017 - 386 Figure 4. Consumers preferred wine brands when buying. 3.5. Consumers’ recognition of wine attributes Eight kinds of wine attributes, namely, price, quality, advertisements, recommendation of others, flavor, package, origins and productive years were selected in the current study. These scale items were based on the literature (HALL et al., 2013; KING et al., 2012; GOODMAN, 2009; GERAGHTY and TORRES, 2009). The respondents were asked to rank them according to importance. For the each attribute i (i =1, 2, ..,8, representing the 8 kinds of wine attributes), its final score Si can be mathematically formulized as follows: Where valuej is the importance for attribute i and ranges from 1 to 8 (8-the most important, 7- the second important, 6- the third important, 5- the fourth important, 4- the fifth important, 3- the sixth important, 2- the seventh important, 1- the eighth important), and frequencyj is the times for valuej occurence. n is the number of respondents. The responses indicated that quality obtained the highest score of 5.78, followed by price (5.43), as shown in Fig. 5. The third highest score (5.05) pertained to taste, followed by source (2.72). The result indicated that consumers prioritized wine quality in purchasing wine. This result may indicate the need for the Chinese wine industry to improving wine quality. 3.6. Market segments Before designing the questionnaire, the research staff surveyed two supermarkets and three wine shops in Beijing to acquire information on wine price. After analyzing the price data, the wine price ranges for daily drinking wine and wine for gifts were set in the questionnaire. However, the current study focused only on daily drinking of wine. In the analysis of market segmentation, 121 questionnaire were abandoned for missing the information about wine price for daily drinking. Fig. 6 shows that for daily drinking wine Ital. J. Food Sci., vol 29, 2017 - 387 (bottle/750 ml), most consumers accepted the price range of CNY 75-CNY 100, and few samples were willing to purchase wine higher than CNY 175. Based on consumers’ preference in wine price for daily drinking, Chinese wine consumers in the current study were categorized into 3 groups (Table 4). Figure 5. Consumers’ recognition of wine attributes. Notes: 8-most important, 7-second important, 6-third important, 5-fouth important, 4-fifth important, 3-sixth important, 2-seventh important, 1-eigth important. Figure 6. Consumers’ acceptable price for wine for daily drinking. These groups included low spenders, moderate spenders and high spenders. After the groups were identified, chi-square analysis was employed to identify the significant difference between the groups. The variables in the analyses were demographics, which include sex, age, income, level of education, and job category. The statistic results are shown in Table 5. Ital. J. Food Sci., vol 29, 2017 - 388 Table 4. Clustering Chinese wine consumers according to consumer preference in wine price for daily drinking Segments wine price for daily drinking Low Spenders Below CNY 75 (16.6%) Moderate Spenders Between CNY 75-100 (24.2%) High Spenders Above CNY 100 (59.2%) Table 5. Demographic differences among price segments. Variables Groups (N=825) Chi-square Test Low Moderate High F Sig. Gender Male 69 98 259 1.049 0.592 Female 68 102 229 Age Under 25 70 87 211 26-35 42 77 176 6.492 0.592 36-45 13 22 70 46-55 9 11 27 Above 55 3 3 6 Income a Below 2000 41 39 98 2001-3000 22 29 54 3001-4000 29 47 69 4001-5000 16 30 80 35.297 0.000*** 5001-7000 18 24 85 7001-10000 9 25 69 Above 10000 2 6 35 Educational background Junior high school and below Senior high school Junior college Undergraduate and above 3 23 30 81 1 29 32 138 13 29 79 369 28.469 0.000*** Job category Government official 5 12 39 Staff in enterprises Liberal profession 52 18 49 32 126 147 Farmers 1 1 2 Education departments 4 50 26 114.014 0.000*** Students 39 48 110 Unemployed/retired 6 5 10 Else 12 3 10 Notes: aAverage monthly income per capita. N, the total sample number of respondents, *** Significant at 1% level respectively. Ital. J. Food Sci., vol 29, 2017 - 389 3.7. Demographic differences among the three segments Chi-square Test was conducted to distinguish the significant consumer demographics, As Table 5 shows, three variables of income, educational background and job category are significant. Low spenders have a significantly lower income than the other two groups. Therefore, this may explain why more people prefer to middle-priced wine, after all, wine is also only a decoration in consumers’ life. Wine is not like food, consumers are conscious about the price of wine. When wine price exceeds their psychological expectation, consumers are bound to reduce their wine consumption (MU and FENG, 2010). Income is an important predictor of wine involvement bacause consumers need to have the financial resources to support their interest in wine (BRUWER, 2013; Barber et al., 2006). The group composed of undergraduate and above dominated among the three groups. The education standard of high spenders was higher than that of the two other groups. The number of undergraduate and above respondents accounted for 75.3% of the high spenders, 59.1% of the moderate spenders and 69.0% of the low spenders. According to LIU et al. (2014), consumers are frequently exposed to information about wine because of the rapid development of the wine market in China, thus resulting in more mature consumers. THATCH (2008) argues that the actual tasting of wine helps consumers develop an educated palate. Job category did significantly differ among the three groups. High spenders were likely to be liberal professions. Consumers who worked at education departments were likely to be moderate spenders, and staffs in enterprises mainly comprised the low spender group. Occupational difference reflects the status of consumers in society somehow. Researchers have verified that wine is considered as a symbol of social status and sophistication (FAN, 2007; LIU and MURPHY, 2007). No significant difference was observed among the three groups in terms of gender and age. Previous research indicated that wine is not a luxury goods for consumers with improved standard of living, and that people can significantly benefit from wine (PETTIGREW and CHARTERS, 2010). Thus, people prefer wine to other drinks, such as beer and liquor (ZHOU et al., 2011). Income, education background and job category may influence the three segments. To distinguish the actual factors that influence the three segments, a principal component analysis was conducted. Table 6 shows that the data passed the Kmo and Bartlett’s test of sphericity. The eigenvalues of income and educational background were all above one, and the overall contribution rate of the three segments accounted for 82.829%. These findings indicate that income and educational background can be regarded as main factors affecting the preferred wine price of consumers for daily drinking. Table 6. Principal component analysis of income, educational background and job category. Component Total % of Variance Cumulative KMO and Bartlett’s test of sphericity (%) F Sig. Income a 1.484 49.482 49.482 Educational background 1.000 33.347 82.829 0.499 0.000*** Job category 0.515 17.171 100.000 Notes: aAverage monthly income per capita, ***Significant at 1% level respectively. Ital. J. Food Sci., vol 29, 2017 - 390 3.8. Behavior and recognition differences in the three segments For behavior, preferred purchase places and favorite brands were examined. First, the places where the three segments purchased wine were observed. Five common wine selling places were selected in the current study, and the respondents were asked to choose the places they usually purchase wine. The results were listed in Table 7. Table 7. Behavior and preferences differences among the three segments. Variables Groups (N=825) Chi-square Test Low Moderate High F. Sig. Purchase places Supermarkets 94(68.6%) 135(67.5%) 278(57.0%) Exclusive Shop 21(15.3%) 37(18.5%) 110(22.5%) Online shopping 17(12.4%) 15(7.5%) 49(10.0%) 20.287 0.009*** Wine stores 1(0.7%) 11(5.5%) 39(8.0%) Else 4(2.9%) 2(1.0%) 12(2.5%) Favorite bands Changyu 53(38.7%) 76(38.0%) 194(39.8%) GREATWALL 41(29.9%) 70(35.0%) 112(23.0%) DYNASTY 6(4.4%) 3(1.5%) 9(1.8%) WILON 3(2.2%) 2(1.0%) 15(3.1%) 25.459 0.013** SUNTIME 4(2.9%) 3(1.5%) 7(1.4%) Foreign brands 24(17.5%) 38(19.0%) 138(28.3%) Else 6(4.4%) 8(4.0%) 13(2.7%) Notes: N, the total sample number of respondents, *** , ** Significant at 1% and 5% level respectively. Consumers most frequently purchase wine in supermarkets among the three groups. The main factors affecting consumers to select the place to purchase wine are convenience and price (FENG et al., 2012). As various supermarkets, which conveniently supply various and conveniently priced wines for consumers can be found in residential areas in China. Supermarkets are important in making wine accessible to the mainstream market (RICHIE, 2007). Changyu, Greatwall, and foreign bands were preferred by the three segments. The most popular wine brand in China is Changyu, followed by Greatwall (Mu and Feng, 2010). And high spenders and low spenders also preferred Dynasty, Wilon, Sumtine and other brands. The results of Chi-square Test indicated that consumer behavior and recognition of wine showed some statistical significantly difference. It was found that the number of high spenders was higher in other places where wine could be purchased, such as exclusive shops and wineries, than those of the other two groups. Low spenders were more likely to buy wine online shopping and other places, after all, products that sold online were cheaper than other places. For favorite brands, such as foreign brands, there was a great possibility that consumers were high spenders. Ital. J. Food Sci., vol 29, 2017 - 391 3.9. Knowledge and attitude differences among the three segments To examine wine knowledge, consumers’ cognition towards wine origins were analyzed. Table 8 shows that Xinjiang Province and the Ningxia Hui Autonomous Region were considered as the best wine origins by the three groups. High spenders considered Yunnan Province to be the best wine origin. To Hebei Province, which is one of the better wine origins, the number of low spenders is greater than those of the other two groups. Beijing City, Tianjin City and Hebei Province were considered as the worst wine origins in China by the three segments. Northwest areas are the best wine origins (MU and FENG, 2010), the survey result is consistent with this outcome. However, more than 5.0% of the respondents in each group considered Ganus, Shandong, and Jilin Provinces as the best and worst wine origins, thus indicating that consumers’ limited knowledge about wine origins (Mu and Feng, 2010). Table 8. Knowledge differences among the three segments. Groups (N=825) Chi-square Test Low Moderate High F Sig. Best originsa Beijing City 19 26 55 Tianjin City 12 10 36 Hebei Province 27 26 71 Jilin Province 21 30 64 Gansu Province 16 39 77 Ningxia 49 85 165 22.076 0.229 Xinjiang Province 82 128 324 Shandong Province 24 24 81 Yunnan Province 13 18 78 Liaoning Province 11 14 35 Worst originsa Beijing City 49 66 169 Tianjin City 45 61 172 Hebei Province 39 59 149 Jilin Province 21 32 74 Gansu Province 15 32 72 12.179 0.838 Ningxia 12 17 41 Xinjiang Province 5 3 19 Shandong Province Yunnan Province Liaoning Province 23 25 39 24 40 56 85 91 104 Notes: N, the total sample number of respondents, a The question is multiple choice. For deeper statistical analysis of consumers’ origins knowledge above findings, it is found that consumers presented no significant difference in wine origin knowledge. This finding is not surprising. According to Liu et al. (2014) confirmed that a wine culture is lacking in China and many Chinese wine consumers were novice wine buyers. Chinese even have low awareness of grape varietals or food matching (Jenster and Cheng, 2008). Ital. J. Food Sci., vol 29, 2017 - 392 Consumers’ recognition of wine attributes was examined to determine the differences in attitude of the three segments. Table 9 shows that significant difference exists among the respondents in terms of the importance of wine attribute. Table 9. Attitude differences among the three segments. Attributes Groups (N=825) Chi-square Test Low Moderate High F. Sig. Most important Price 64(46.7%) 68(34.0%) 134(27.5%) Quality 35(25.5%) 57(28.5%) 140(28.7%) Advertisement 3(2.2%) 8(4.0%) 13(2.7%) Recommended by friends 6(4.4%) 5(2.5%) 36(7.4%) Flavor 16(11.7%) 38(19.0%) 93(19.1%) Package 2(1.5%) 5(2.5%) 6(1.2%) 29.534 0.009*** Origins 6(4.4%) 12(6.0%) 42(8.6%) Productive year 5(3.6%) 7(3.5%) 24(4.9%) Second important Price Quality 30(21.9%) 47(34.3%) 33(16.5%) 68(34.0%) 52(10.7%) 159(32.7%) Advertisement 13(9.5%) 7(3.5%) 23(4.7%) Recommended by friends 7(5.1%) 22(11.0%) 46(9.5%) 30.602 0.006*** Flavor 30(21.9%) 38(19.0%) 118(24.3%) Package 4(2.9%) 11(5.5%) 26(5.3%) Origins 3(2.2%) 13(6.5%) 38(7.8%) Productive year 3(2.2%) 8(4.0%) 24(4.9%) Third important Price 19(13.9%) 36(18.0%) 78(16.0%) Quality Advertisement Recommended by friends 18(13.1%) 12(8.8%) 20(14.6%) 23(11.5%) 14(7.0%) 25(12.5%) 58(11.9%) 23(4.7%) 42(8.6%) Flavor 36(26.3%) 49(24.5%) 121(24.8%) 16.144 0.305 Package 9(6.6%) 15(7.5%) 42(8.6%) Origins 16(11.7%) 26(13.0%) 74(15.2%) Productive year 7(5.1%) 12(6.0%) 50(10.2%) Forth important Price 9(6.6%) 26(13.0%) 82(16.8%) Quality 6(4.4%) 14(7.0%) 33(6.8%) Advertisement 7(5.1%) 11(5.5%) 24(4.9%) 25.874 0.027** Recommended by friends 23(16.8%) 21(10.5%) 39(8.0%) Flavor 14(10.2%) 22(11.0%) 41(8.4%) Package 21(15.3%) 18(9.0%) 44(9.0%) Origins 24(17.5%) 48(24.0%) 111(22.7%) Productive year 33(24.1%) 40(20.0%) 114(23.4%) Notes: N, the total sample number of respondents, ***, ** Significant at 1% and 5% level respectively. Ital. J. Food Sci., vol 29, 2017 - 393 Low spenders (46.7%) and moderate spenders (34.0%) considered price as the most important attribute, followed by quality, flavor, and origins. High spenders (28.7%) prioritized quality, followed by price, flavor, and source. The results of the three groups were also inconsistent regarding the second most important wine attribute. High spenders focused on the intrinsic attributes (quality, flavor, and source). The three segments did not significantly differ in the third most important attribute of wine, as flavor dominated all the three segments. Low and high spenders chose productive year as forth most important attribute, and moderate spenders favored source. This segment result was similar to that of Liu et al. (2014), in which consumers were categorized to intrinsic and extrinsic cluster. The intrinsic cluster mainly sought information about wine quality, price, flavor, and brand name, which are considered as the intrinsic attributes of wine. The extrinsic cluster pertained to wine package, and foreign wine, which are regarded as the extrinsic attributes of wine. According to Liu et al. (2014), Chinese wine consumers with a high income were likely to belong to the intrinsic cluster, whereas consumers with a low income belonged to the extrinsic cluster. The previous analysis indicated that high spenders have a higher income than the other two groups. Therefore it was persuasive that why High Spenders prioritized quality and low spenders and moderate spenders considered price as the most important factor when purchase. Attitude positively affects the identification of factors influencing the purchasing behaviors of consumers. The result indicates that quality, price, flavor and source are four most important wine attributes. Batt (2000) verifies that price is the most important factor affecting the purchasing behavior of consumers. Gunay and Baker (2011) demonstrate the importance of price, origin, and quality, and finds that quality and price are the most important factors affecting the purchase decision of consumers. 3.10. Gift-oriented acceptable price of wine difference among the three segments The previous analysis showed that a relatively high number of consumers purchased wine for daily drinking and gifts, which were quite different in terms of motivation, preferred price and brand preference (PAN, 2012). Consumers are willing to pay a premium for wine as gifts because wine is a status symbol (Liu and Murphy, 2007). Thus, the accepted price of wine as gifts may differ. Therefore, this study separately analyzed consumers’ price acceptability of wine for daily drinking and that of wine for gifts. Before designing the questionnaire, the research staff surveyed two supermarkets and three wine shops in Beijing to acquire information about wine price. The price data and wine price ranges were set for wine as gifts in the questionnaire. Table 10 shows that the acceptable prices for gift-oriented wine differed among the three segments. High spenders tend to buy high priced wine. Respondents with an income of CNY 100- CNY 150(30.7%) dominated the low spenders. Respondents with an income of CNY 150- CNY 200(26.0%) comprised the largest proportion in the moderate spenders group, followed by those with an income of CNY 200-CNY 250(21.0%). The price of the overall tendency of high spenders was higher than that of the other two groups. A consumer who buys a bottle (750 ml) of wine with a price above CNY 250 may be to be a high spender (77.7%). A Chi-square Test was conducted to identity statistical significant. The results verify that there is great significant difference in preferred price of wine as gifts. Income is an important predictor of wine involvement because consumers need to have the financial resources to support their choice (BARBER et al., 2006). Ital. J. Food Sci., vol 29, 2017 - 394 Table 10. Gift-oriented wine price preference difference among the three segments. Variables Groups (N=825) Chi-square Test Low Moderate High F. Sign. Price for gift-oriented wine(CNY) Under 50 3(2.2%) 0(0.0%) 2(0.4%) 50-100 23(16.8%) 5(2.5%) 1(0.2%) 100-150 42(30.7%) 28(14.0%) 19(3.9%) 150-200 30(21.9%) 52(26.0%) 63(12.9%) 200-250 16(11.7%) 42(21.0%) 69(14.1%) 250-300 3(2.2%) 17(8.5%) 55(11.3%) 262.813 0.000*** 300-350 6(4.4%) 16(8.0%) 73(15.0%) 350-400 4(2.9%) 18(9.0%) 45(9.2%) 400-450 4(2.9%) 9(4.5%) 57(11.7%) Above 450 6(4.4%) 13(6.5%) 104(21.3%) Notes: N, the total sample number of respondents, ***Significant at 1% and 5% level respectively. The previous analysis shows that income and education background significantly differ among the three segments. High spenders are likely to have a high income and educational level. Sending expensive wine is an act of politeness and respect (SUN et al., 2009). Therefore, high spenders had sufficient money and were more likely to choose expensive wine than other types of spenders. 3.11. Profiles of the three consumers segments Creating profiles of the three different price segments based on the significant variables identified in this research is required. Table 11 illustrates the major differences among the segments. Table 11. Profiles of Chinese consumers by price. Significant Variables Groups Low Moderate High Demographics Income Below 2000 3001-4000 5001-7000 Educational background Undergraduate and above Undergraduate and above Undergraduate and above Job category Staff enterprises Education department Liberal profession Behaviors supermarkets, supermarkets, Wine stores, Purchase places exclusive shops, online shopping exclusive shops, online shopping supermarkets, exclusive shop, online shopping Preferred brands Changyu,GreatWall Changyu, GreatWall Changyu, foreign brands Attitude to wine attributes Price, quality, flavor, productive year Price, quality, flavor, origins Quality, price, flavor, productive year Accepted price for gift- oriented wine CNY 100-150 CNY 150-200 Above CNY 250 Ital. J. Food Sci., vol 29, 2017 - 395 4. CONCLUSIONS The frequency of wine consumption is affected by age. The 26-35 age group dominates the often purchase cluster, and under 25 age group is mostly in the purchase occasionally. These results show that more young are involved in wine consumption. Wine consumption occurs in many occasions, mainly at home, during holidays; most people also send wine as gifts. A significant difference is found between consumers’ income and consumption occasion. Consumers with low income tend to consume wine on special occasions, such as the Chinese traditional holiday. And on occasions that require the consumption of wine such as business banquets, consumers whose income is CNY 7001- CNY 10000 (28.8%) seem to have much money and more chances to consume wine. Consumers are familiar with three brands, namely, Greatwall, Changyu, and Dynasty. This research introduces the Chinese wine consumer price segments, namely, low spenders, moderate spenders, high spenders, and explores their differences in demographics, behaviors and preferences, and attitudes. High spenders have incomes higher than those of the other two groups, and they may be liberal professions. Conversely, lower spenders may work in enterprises and moderate spenders can be in the education department. All three groups choose supermarket as their first wine purchase place, and a considerable number of high spenders purchase wine in wine stores. Excluding Changyu, the favorite wine brands of high spenders are foreign brands, and the favorite wine of the other two groups is Greatwall. Knowledge about wine origins of the three segments is limited, and their cognition of the importance of wine attributes has a significant difference. High spenders look for quality, price, flavor, productive year; moderate spenders look for price, quality, flavor, source; and low spenders look for price, quality, flavor, and productive year. A significant difference is observed in the acceptable wine price for gift-oriented. Low Spenders prefer wines priced at CNY 100-CNY 150, and moderate spenders are inclined to buy wine priced at CNY 150-CNY 200. High spenders prefer wines with the highest price at above CNY 250 among the three segments. However, the investigation was only conducted in Beijing because of the limitation of time and energy and this is also the limitation of this paper. In order to better understand Chinese wine consumer consumption characteristics, survey in more provinces of China is the focus of the authors in the near future. ACKNOWLEDGEMENTS This research is supported by the Chinese Agricultural Research System (CARS-30). REFERENCES Barber N., Almanza B.A. and Donovan J.R. 2006. Motivational factors of gender, income and age on selecting a bottle of wine. International Journal of Wine Marketing 18(3):218-232. Batt, P.J. and Dean A. 2000. Factors influencing the consumer's decision. Australian & New Zealand Wine Industry Journal 15(4):S34-S41. Brida J.G., Scuderi R. and Seijas M.N. 2012.Segmenting cruise passengers visiting Uruguay: A factor cluster analysis. International Journal of Tourism Research 16(3):209-222. Brunner T.A. and Siegrist M. 2011. A consumer-oriented segmentation study in the Swiss wine market. British Food Journal 113(3):353-373. Ital. J. Food Sci., vol 29, 2017 - 396 Bruwer J., Li E. and Reid M. 2002. Segmentation of the Australian wine market using a wine-related lifestyle approach. Journal of wine research 13(3):217-242. Bruwer J. and Buller C. 2012. Consumer behavior insights, consumption dynamics, and segmentation of the Japanese wine market. Journal of International Consumer Marketing 24(5):338-355. Bruwer O.A.O.J. 2013. Enduring involvement with wine: predictive model and measurement. Journal of Wine Research 24(3):210-226. Chen L.C., Lin S.P. and Chunmin K. 2013. Rural tourism: marketing strategies for the bed and breakfast industry in Taiwan. International Journal of Hospitality Management 32(1):278-286. China Daily. 2012. China’s wine consumption grows at 25-30% annually, expecting to excel England. Available: http://sp.chinadaily.com.cn/wine/20120619/28954.html. Christodoulidou N. and James M.S. (2011). Factors influencing wine consumption in southern California consumers. International Journal of Wine Business Research 23(1):36-48. Costa C. and J.C. 2013. Price formation and market segmentation in seaside accommodations. International Journal of Hospitality Management 33:446-455. Cohen S.A., Prayag G. and Moital M. 2014. Consumer behaviour in tourism: concepts, influences and opportunities. Current Issues in Tourism 17(10):872-909. Dolnicar S. 2002. A review of data-driven market segmentation in tourism. Journal of Travel & Tourism Marketing 12(1):1-22. Euromonitor International. 2013. Wine in China. Available: www.euromonitor.com/wine-in-china/report. Fan M. 2007. Wine seen as social status in China. The Washington Post, 30 January. Feng H., Feng J., Tian D. and Mu W. 2012. Consumers' perceptions of quality and safety for grape products: A case study in Zhejiang Province, China. British Food Journal 114(11):1587-1598. Flynn M., Elliott G. and Ritchie, C. 2010. Buying wine on promotion is trading‐up in UK supermarkets. The traditional formal logic :.Methuen 102-121. Forbes S. 2012. The influence of gender on wine purchasing and consumption: an exploratory study across four nations. International Journal of Wine Business Research 24(2):146-159. Geraghty S. and Torres A.M. 2009. The Irish wine market: a market segmentation study. International Journal of Wine Business Research 21(2):143-154. Gergely S. and Dieter H. 2014. Consumer segmentation based on usage of sales channels in the German wine market. International Journal of Wine Business Research 26(1):27-44. Goodman S. 2009. An international comparison of retail consumer wine choice. International Journal of Wine Business Research 21(1):41-49. Guillet B.D. and Kucukusta D. 2016. Spa market segmentation according to customer preference. International Journal of Contemporary Hospitality Management 28(2):418-434. Gunay G.N. and Baker M.J. 2011. The factors influencing consumers’ behaviour on wine consumption in the Turkish wine market. EuroMed Journal of Business 6(3):324-341. Hall J., Lockshin L. and Mahony G.B.O. (2013). Exploring the links between wine choice and dining occasions: factors of influence. International Journal of Wine Marketing 13(1): 36-53. Jenster P. and Cheng Y. 2008. Dragon wine: developments in the Chinese wine industry. International Journal of Wine Business Research 20(3):244-259. Johnson T. and Bastian S. 2009. A preliminary study of the relationship between Australian wine consumers’ wine expertise and their wine purchasing and consumption behaviour. Australian Journal of Grape and Wine Research 13(3): 186-197. King E.S., Johnson T.E., Bastian S.E.P., Osidacz P. and Francis I.L. 2012. Consumer liking of white wines: segmentation using self-reported wine liking and wine knowledge. International Journal of Wine Business Research 24(1):33-46. Liu F. and Murphy J. 2007. A qualitative study of Chinese wine consumption and purchasing: Implications for Australian wines. International Journal of Wine Business Research. 19(2):98-113. Ital. J. Food Sci., vol 29, 2017 - 397 Liu H.B., Mccarthy B., Chen T., Guo S. and Song X. 2014. The Chinese wine market: a market segmentation study. Asia Pacific Journal of Marketing and Logistics 26(3):450-471. Lockshin L. and Corsi A.M. 2012. Consumer behaviour for wine 2.0: A review since 2003 and future directions. Wine Economics and Policy 1(1):2-23. Marshall G. and Johnston M. 2010. Marketing Management, McGraw-Hill Higher Education, London. Molina, A., Gómez, M., González-Díaz, B. and Esteban, A. 2015. Market segmentation in wine tourism: strategies for wineries and destinations in Spain. Journal of wine research 26(3):1-33. Mu, W. and Feng, J. 2010. “The research on economic issues of grape industry in China”. Beijing, China. OIV Statistics. 2015. International Organization of Vine and Wine website. Available: http://www.oiv.int/oiv/info/enconjoncture (accessed 4 March 2016). Olsen J.E., Atkin T., Thach L. and Cuellar S.S. (2015). Variety seeking by wine consumers in the southern states of the US. International Journal of Wine Business Research 27(4):260-280. Pan J. 2012. Research on Influencing Factors of Wine Purchase Based on the Discrete Choice. China Agricultural University. Pettigrew S. and Charters S. 2010. Alcohol consumption motivations and behaviors in Hong Kong. Asia Pacific Journal of Marketing 22(2):210-221. Prayag G., Cohen S.A. and Yan H.L. 2014. Potential Chinese travellers to Western Europe: segmenting motivations and service expectations. Current Issues in Tourism 18(8):1-19. Richie C. 2007. Beyond drinking: the role of wine in the life of the UK consumer. International Journal of Consumer Studies 31(1):534-540. Rondancataluña F.J. and Rosadiaz I.M. 2014. Segmenting hotel clients by pricing variables and value for money. Current Issues in Tourism 17(1): 60-71. Simpson K. and Bretherton P. 2004. Market segmentation by consumer lifestyle in a wine tourism setting. In New Zealand Tourism and Hospitality Research. Sun H., Goodman S., Yu Y., Chen S. and Ma H. 2009. Chinese choices: a survey of wine consumers in Beijing. International Journal of Wine Business Research 21(2):155-168. Tang V.C.M., Tchetchik A. and Cohen E. 2015. Perception of wine labels by Hong Kong Chinese consumers. Wine Economics & Policy 4(1):12-21. Thatch E. 2008. How American consumers select wine. Wine Business Monthly 15(6):66-71. UN Comtrade. 2015. United Nations Commodity Trade Statistics Database website. Available: http://comtrade.un.org. Vigarellis D. 2015. Consumer Knowledge: A New Basis for Segmentation? Journal of General Management. Voorhees C., McCall M. and Calantone R. 2011. A new look at the benefits of improving segmentation efforts with reward programs. Cornell Hospitality Reports 11(11):4-14. Vinexpo. 2013. More and more Chinese professionals plan to come to Vinexpo Bordeaux. Available: www.vinexpo.com/en/press/archives/detail/vinexpo-2013/more-and-more-chinese-professionals-plan-come- vinexpo-bordeaux. Wang Y. 2006. Wine consumption of urban residents in China - Taking Beijing city as an example. China Agricultural University. Wine Institute. 2012. Per capita consumption by country. Available: http://www.wineinstitute.org/files/2012_World_Per_Capita_Consumption_by_County_cCalifornia_Wine_Institute.pd f. Zhou Z., Liu H., Tian W., Wang J. and Cao L. 2011. Food consumption trends in China, Department of Agriculture, Forestry & Fisheries, Canberra, ACT. Available: www.daff.gov.au/__data/assets/pdf_file/0006/2259123/food- consumption-trends-in-china-v2.pdf. Paper Received September 26, 2016 Accepted January 25, 2017