Market Segmentation in the K-League: An Analysis of Spectators of the Korean Professional Soccer League
Kim, Sangho, Yoo, Euidong, Pedersen, Paul M., International Journal of Sports Marketing & Sponsorship
This study involved analysis of the consumption behaviours of spectators in the K-League (South Korea). Its dual purpose was to cluster spectators into homogeneous groups on the basis of attitudes towards game attendance, and to define the segments obtained on the basis of the demographic and lifestyle profiles of the spectators. Multiple steps were taken to analyse the data from a survey of 967 spectators. This revealed four distinct groups--promotion-concerned, place-concerned, price-concerned and indifferent.
The Korean K-League commenced operations nearly a quarter of a century ago. Since its inception in 1983, this professional soccer league has been a financial failure for all but two (1983 and 1984) of its regular seasons. Although the 13 teams in the league have not been financially solvent, they have provided added value to their parent companies through their ability to provide those companies with positive communication and effective advertising. Because of this unique situation, the practitioners employed by the K-League have not realised the need for a systematic marketing plan. The failure to recognise the benefits of such a critical plan has limited the growth and financial viability of the sports entities within the league.
In an attempt to introduce the concept of market analysis to the Korean professional soccer industry, this study was designed to provide marketing information and analysis concerning spectators at K-League games. The study, which included a survey of spectators (N=967) who attended three league home games across three different major cities in Korea, involved two investigative aspects. The first part of the study included the clustering of spectators into homogeneous groups that had similar characteristics on the basis of attitudes towards game attendance. The second part of the analysis involved defining the segments obtained on the basis of the demographic and lifestyle profiles of the spectators.
The data analysis involved three steps. First, an exploratory factor analysis was used to determine the 19 attendance items that explained why spectators attend a game. Second, a cluster analysis was utilised to segment spectators based on the identified factor groupings. Third, discriminant and cross-tabulation analyses were conducted to determine any significant segment descriptor among the identified factors and the demographic and lifestyle variables.
The results of a factor analysis produced a solution of four factors with eigenvalues greater than unity: a) place, b) product, c) price and d) promotion. A four-cluster solution was identified through the two types of cluster analyses. Then, the result of a multiple discriminant analysis showed that the three functions accounted for 46.9%, 29.0%, and 24.1% of the variance in the clusters. Lastly, the demographic and lifestyle profiles for each cluster were examined using cross-tabulation. The chi-square analysis revealed that four out of the five demographic variables and all five lifestyle variables showed statistically significant differences relating to the four clusters.
Four distinct segments were identified and named as a result of the multiple analytical steps. These four segments include the promotion-concerned group, place-concerned group, price-concerned group and the indifferent group. By implementing the results of this study, sports marketing practitioners in the K-League can use these segments to understand the needs and wants of their target markets, and to meet the satisfaction demands of their various stakeholders more effectively.
The commencement of professional soccer in Korea occurred nearly a quarter of a century ago. In 1983, two professional teams (Hallelujah and Yukong) and three amateur teams (POSCO, Daewoo and Kookmin Bank) united to form the basis of what is now referred to as the K-League. Within its first four years, the K-League grew to include five professional soccer teams (Daewoo, POSCO, Yukong, Hyundai and LG). The addition of five more teams over the next 10 years resulted in the league's movement to a more professional approach to operations. Professional soccer in Korea developed with the adoption of a professional structure and business outlook in the mid-1990s. It was during this time that the teams changed their names to incorporate their new home-bases throughout Korea creating team names such as the Busan Daewoo, Ulsna Hyundai and Anyang LG. Today, there are 13 professional teams competing in the K-League.
Despite its growth, the K-League has been a financial failure throughout most of its existence. The only two years in which the league was profitable were its first two seasons in 1983 and 1984. Yoo (2004) has noted two major reasons why the league consistently fails to be financially solvent. The first is because the 13 teams in the league have mainly relied on (and simply consider themselves to be promotional extensions of) their parent companies. The second is that much of the spectator focus and interest has been concentrated on international competitions such as The World Cup, The Olympic Games and the FIFA Youth Championship. The major international soccer events feature no star players from the team rosters for the K-League.
As a result, those employed in the K-League have failed to realise and appreciate the need for a systematic marketing plan. This failure to consider the advantages of implementing marketing strategies has limited the enhancement and viability of the league and its associated teams. Several scholars (Chae & Lee, 2000; Choi, 1999; Kim, 1998; Kim & Cho, 2003) have illustrated the lack of systematic marketing concepts for professional sports industries in Korea. There is now a need to devise specific and systematic marketing strategies to revitalise the league and Korean professional soccer. This research, which focused on market segmentation based on factors that affect attendance, can be used to initiate the development of marketing strategies for the professional soccer industry in Korea. Therefore, this study was designed to provide information about the spectators of K-League games. Specifically, there was a dual purpose behind this study. First, the study sought to cluster spectators into homogeneous groups with similar characteristics. Second, the study attempted to define the spectator segments on the basis of the attitudes and demographic profiles of the spectators. This research approach was implemented to understand the needs and desires of the target markets and to improve the customer (i.e. spectator) satisfaction.
Professional team sports contain a number of salient characteristics which lead to segmentation approaches. Several scholars have illustrated these characteristics (Mullin et al, 2000; Park et al, 2003). For example, one characteristic is that sports consumers are often highly involved with the sports product. This is due to sport's salience as well as strong personal identification. Another characteristic of sport is that consumer demand tends to fluctuate widely. The sports product is inconsistent and unpredictable, sport has almost universal demographic appeal worldwide, and the sports marketer has little or no control over the core product (i.e. the game on the field). Another characteristic is that sport is an example of a highly intangible and personalised service. This simply means that each spectator or viewer takes a unique set of benefits from the sports experience. Segmentation is often used in professional team sports because of the prevalence of these and other salient characteristics.
Spectator attendance has also been extensively researched to determine the association between on-field success in team sports, and the attraction and retention of spectators (Bauer et al, 2005; DeSchriver & Jensen, 2002). Most studies have identified a strong positive association between spectator attendance and on-field success. Thus, an understanding of the nature of spectators is essential for the development of marketing strategies in team sports. For example, soccer spectators in previous decades were often quite homogeneous. Russell (1997) noted that until the late 1980s, spectators of soccer in London were mainly males from the skilled working class and the lower middle class demographic. In Korea, the soccer industry attracted a mainly male audience before the 2002 Korea-Japan World Cup (Kim & Cho, 2003). Since that international event there has been both an increase and a diversification of soccer spectators and participants in Korea.
Market segmentation approaches
To understand the nature of spectatorship and the distinguishing characteristics among spectators, a number of scholars have investigated a variety of segmentation approaches (Abell & Hammond, 1979; Croft, 1994; Dibb & Simkin, 1996; Myers, 1996; Tapp & Clowes, 2002). Specifically, according to the study by Kim and Cho (2003), characteristics of spectators in professional sports have become heterogeneous. Furthermore, these researchers noted that sports marketers have realised that often a more meaningful analysis can be achieved by disaggregating the research results and using a technique called market segmentation to organise various important spectator groups. Therefore, this study used the market segmentation approach to secure valuable spectator information and to assist the K-League in the development of systematic marketing strategies.
In the consideration of the market segmentation approach, Kotler (2000) has suggested four major segmentation variables: geographic (i.e. region, city or metro size, density, climate); demographic (i.e. age, gender, income, education); psychographic (i.e. lifestyle, personality); and behavioural (i.e. attitude, benefits, user status). The most prevalent segmentation approach over the last few decades has been based on demographic variables (Woo, 1998). However, Sharma and Lambert (1994) have indicated that the demographic profile does not suggest how the marketing strategies should be formulated because the demographic variables such as gender and age are unlikely to reflect the behavioural patterns of consumers directly.
Therefore, recent research on market segmentation has been focused on psychographic and behavioural variables in diverse areas. Many of these studies have focused on the sports industry and include analyses of the service quality associated with sports clubs (Dale et al, 2005), attitudes and motives of fans of clubs in the English Premier League (Tapp & Clowes, 2002), lifestyles and risk perceptions of tourists (Dolnicar, 2005; Gonzalez & Bello, 2002), benefits and lifestyles associated with the service industry (Ehrman, 2006), lifestyles and shopping styles of females (Jackie & Susan, 1998), attitudes toward agrifood (Walley et al, 2000), perceptions of the FA Premiership teams (Richardson & O'Dwyer, 2003), and attributes and benefits of fans of teams in the National Football League and NCAA Division I college football (Funk, 2002).
Individuals respond in myriad ways when they think about certain objects or concepts in a specific situation. Numerous scholars and practitioners have attempted to predict the behaviour of people by investigating their attitudes in specific situations. The word 'attitude' surfaced in the subject of psychology in 1862. Since this time significant research attention has been given to this word by the academic community. Godin (1993) has examined 33 academic studies concerning the relationships among attitude, behavioural intention, and behaviour toward sports participation (i.e. jogging, mountain climbing, biking and aerobic activity). The result of these investigations is that the attitude towards sports participation has been positively related to behavioural intention or actual behaviour. It can be said that attitude is an important indicator in predicting intention or behaviour. Thus, in this study, attitudes towards attending K-League games are used to cluster spectators who attended the professional soccer games.
Numerous studies that have investigated the various factors for fan attendance have been published across a range of sports contexts. Kolbe and James (2000) identified several dimensions of individuals becoming fans of the National Football League, and what people and things affected this process. Three of their key identified dimensions relating to becoming a fan were influential people, events and their influence, and the importance of community. Bristow and Sebastian (2001) showed that compared to less loyal fans, die-hard fans demonstrated different attitudes and behaviours regarding the Chicago Cubs with respect to loyalty, knowledge, exposure, attendance, and purchasing club products. These researchers concluded that creating die-hard fans is much more difficult than it looks, and thus long-term strategies should be developed.
Madrigal (2000) showed that fans of college football buy more club products as their identification with the team increases. The survey was administered to measure team identification by assessing five major items. These items included perceptions of the importance of being a fan of the team, the importance of a winning team, the extent to which spectators saw themselves as fans of the team, the degree to which friends believed the respondents were fans of the team, and how often the individuals displayed or wore items that identified the team. In an examination of college students and fan attendance, Laverie and Arnett (2000) developed a model including factors that directly and indirectly influence fan attendance. These factors include situation involvement, attachment, enduring involvement, identity salience, and satisfaction. Dale et al (2005) extended the research into fan attendance by assessing the impact of several pre-selected factors involved with increasing attendance at home games. The items were clustered into three groups that included public relations issues (i.e. advertising, marketing, profile) and success of the team, pre- and post-match and children's activities, and entry fees and special/large group/children's tickets.
Four other studies examined factors for fan attendance. To segment a team's consumers, Funk (2002) used 13 items (eight attributes and five benefits) that attract individuals to various sporting events. These items included star player, fan identification, nostalgia, product delivery, success, escape, logo design, pride in place, management, venue, head coach, peer group, and tradition. Baade and Tiehen (1990) conducted research to obtain knowledge about the determinants of fan attendance in professional baseball. Some of the factors involved the city's population, per capita income, ticket price, age of the stadium, number of star players, team's winning record, and team's league standing. Krohn et al (1998) examined why people attended sporting events. They distinguished the different factors that influenced spectators and determined that the key factors were personal objectives, excitement and escape, inspiration, personal grievances, and fan identification. Hansen and Gauthier (1989) selected four general categories of factors affecting attendance at professional sports events. These factors included economic considerations (i.e. ticket price, income, forms of entertainment, television effects, other sports attractions), socio-demographic variables (i.e. population size, population ethnicity, geography), attractiveness of game (i.e. closeness of the pennant race, promotions and special events, star players), and residual preferences (i.e. scheduling of games, fan accommodations, facility and weather conditions).
Sports segmentation in Korea
Several scholars have conducted research on the professional soccer industry in Korea. For example, Kim and Cho (2003) examined how loyalty to soccer teams was affected by the various factors in the marketing mix (i.e. product [facility, star player, coaching capability, stadium location, parking convenience], price, place [accommodation, convenience, ease of access] and promotion). Most researchers, however, have illustrated the importance of, and suggested the need for, determining why spectators attend the soccer games in the K-League. For instance, Kim (2004) analysed six attributes of the K-League. These attributes (e.g. service, licensing, facility, ability of players, ability of team, and team image) were then analysed to determine their relationship with game attendance. The four attributes with statistical significance were service, licensing, facility, and team image. Rhee et al (2004) conducted a similar study when they investigated the relationship between spectators' attitudes towards attending K-League games and their intention to revisit. Four factors; facility, game conditions, promotions and cost, were analysed to determine the attitude of spectators. The researchers found that the facility and game conditions were the most influential factors associated with intention to revisit soccer games.
As illustrated above, there has been limited research conducted in the area of sports marketing in Korea. A couple of the studies have provided insight into the variables that affect attendance at various professional sporting events. However, the extent to which the findings of previous studies are applicable to professional soccer events in Korea is unknown. To develop marketing strategies appropriate for the K-League industry, it is necessary to identify and explore the major variables (noted above) that affect game attendance. Therefore, the current study examines the four categories of factors developed by Kim and Cho (2003) and Hansen and Gauthier (1989), and applies them to the K-League.
Sample and data collection
Each of the 13 soccer clubs in the K-League represents a particular city in Korea. For this study, the three cities that were selected through convenience sampling were Po-Hang, Dae-Jeon and Seong-Nam. For each city, one home game in August or September was selected during the 2004 season. The participants in the study were the spectators at the three games. A combined total of nearly a thousand spectators (N=968) were surveyed before the start of the games selected. There were 326 responses from Po-Hang, 353 from Dae-Jeon and 289 from Seong-Nam. One subject provided insufficient information. Therefore, the responses of 967 subjects were examined in this study. The samples from the three cities were compared across different demographic variables such as age, sex, highest educational degree, and household monthly income. Regional bias did not appear to be a concern as there were no statistically significant differences found across the three separate samples.
The subjects ranged in age from 15 to 61 years old (M=25.66; SD=9.542). There were 510 males (52.7%) and 458 females (47.3%). The educational level of subjects included 23 middle school students (2.4%), 301 high school students (31.1%), 115 high school graduates (11.9%), 203 collegiate students (21.0%), 274 college or university graduates (28.3%), 44 graduate students (4.5%) and eight individuals who did not provide an answer as to their educational status (0.8%). The household monthly income of the subjects revealed:
129 earned less than $1,000 (13.3%)
212 earned between $1,000 and $1,999 (21.9%)
232 earned between $2,000 and $2,999 (24.0%)
162 earned between $3,000 and $3,999 (16.7%)
66 earned between $4,000 and $4,999 (6.8%)
101 earned more than $5,000 (10.4%)
66 provided no answer for monthly income (6.8%).
The questionnaire consisted of two parts: factors affecting attendance in K-League games and demographic and lifestyle profiles of the soccer spectators. The first part of the questionnaire was based on interviews with K-League team managers and the attendance factors developed by Hansen and Gauthier (1989; 1992) and Kim and Cho (2003). Nineteen items were used to determine the factors that affected attendance. These factors were based on the dimensions of place, price, product and promotion. Respondents were asked to rate on a scale of one (not at all) to five (very much) the degree of their attitudinal perceptions towards their attendance at the K-League regular season games. Table 1 illustrates the mean and standard deviation scores and the ranking according to mean scores. Based on the ranking, the product is the factor that has the greatest impact among the four factors on home game attendance. This finding supports the models developed by Pan et al (1999), Laverie & Arnett (2000) and Dale et al (2005).
The second part of the questionnaire contained 10 additional items relating to demographic and lifestyle profiles. These items included demographic variables (e.g. age, gender, educational level, occupation, household monthly income, and lifestyle variables (e.g. frequency of games watched, transportation to the stadium, types of companions at the stadium, media used to obtain information about game schedules, purchasing experience of team products). These two variables (demographic and lifestyle) are discussed in more detail below. Overall, the questionnaire consisted of 29 items.
The Statistical Package for the Social Sciences (SPSS) was used to analyse the data secured from the responses of the 967 subjects. Multiple steps were taken to analyse the data. First, an exploratory factor analysis was conducted to explain the 19 variables in terms of their common underlying dimensions (i.e. the factors that affected game attendance in the K-League). Second, a cluster analysis was used to segment subjects into homogeneous groups that had similar attendance patterns. Third, a multiple discriminant analysis (MDA) was employed to determine which attendance variables best discriminated among the clusters identified in the second step. Fourth, a cross-tabulation of the clusters with demographic variables was used to identify any crucial segment descriptors.
Factor analysis As the first step in the analysis of the data, a principal component factor analysis with VARIMAX orthogonal rotation was conducted to explain the 19 attendance variables in terms of their common underlying dimensions (i.e. the factors that affected game attendance in the K-League). The choice of an orthogonal rotation--rather than an oblique rotation--was based on the finding provided by Mitchell (1994) that an orthogonal rotation should be performed when the purpose of the factor analysis is to predict or include values in subsequent analyses such as cluster analysis or discriminant analysis. Furthermore, Hair et al (1987) noted that the orthogonal rotation is a better method than any other rotation to reduce a large set of variables to a smaller number of uncorrelated variables. Therefore, a principal component factor analysis with VARIMAX orthogonal rotation was regarded as the preferred method. Following this factor analysis, regression factor scores, which have a mean of zero and variance equal to the squared multiple correlation between the estimated factor scores and the true factor values, were calculated and used to cluster the spectators at the games. This was done because, according to Punj and Stewart (1983), factor scores reduce the extremity of outliers in that the performance of some algorithms, which are sensitive to outliers, might be improved.
Cluster Analysis The objective of a cluster analysis was to classify the spectators into relatively homogeneous groups. The two general types of clustering algorithms that are used most often are hierarchical and non-hierarchical. Punj and Stewart (1983), in their evaluation of a dozen studies that utilised different clustering methods, illustrated the way in which researchers should determine the most appropriate algorithm for their study. They noted that three procedures warrant special consideration. These three procedures are Ward's minimum variance, average linkage, and K-means. Ward's method appeared to outperform the average linkage method except in the presence of outliers. K-means appeared to outperform both Ward's method and the average linkage method if a non-random starting point is specified. Based on the findings of Punj and Stewart (1983), this study's sample of 967 spectators was randomly divided into two sub-samples, and Ward's method was applied to sub-sample to determine an initial clustering solution. The number of clusters was obtained on the basis of agglomeration coefficients and dendrogram from the two sub-samples. Then, the number of clusters obtained was used as inputs to K-means because the number of clusters must be specified beforehand in the non-hierarchical method.
Multiple Discriminant Analysis (MDA) In an attempt to examine whether the clusters that were identified in the second step (cluster analysis) were differentiated in terms of the 19 items that affect game attendance, MDA was conducted. To run this analysis, the total sample of respondents (N=967) was randomly divided into two groups: the analysis sample (n=486) and the holdout sample (n=481). The analysis sample was used for estimating the discriminant function. The holdout sample was reserved for validating the discriminant function. The validity test was based on hit ratios of two groups of respondents because, according to Kennelly (1999), the hit ratio indicates the percentage of cases correctly classified by the discriminant analysis.
Cross-Tabulation Analysis The demographic and lifestyle profiles for each cluster were identified using a bi-variate cross-tabulation analysis. This was conducted because a dependent variable (cluster) and the independent variables (10 variables) were categorical in constitution. The chi-square goodness-of-fit was used to determine whether there were any statistically significant differences among the defined clusters.
Factor Analysis Four factors were eventually produced based on the eigenvalues and the scree-tail test (used to identify the optimum number of extracted factors). As illustrated in Table 2 and detailed below, four distinct factors emerged from the scale measuring factors affecting attendance in the K-League. Each of the factors represented a different type of characteristic. The first factor contained place-related variables (i.e. convenient facility, proper starting time, proper game duration, cleanliness of facility, ease of facility access). The second factor showed product-related variables (i.e. offensive playing style, team history, record-breaking performances, rivalries, star athletes, and team's place in league standings). The third factor was related to variables associated with what spectators considered a proper price for certain products or services (i.e. food and beverages, merchandise, parking fee, entrance fee). The fourth factor included promotion-related variables (i.e. media advertisements of games, special event occasions, promotional items given at games, no games on television).
The internal consistency of the scale in terms of alpha coefficient (Cronbach, 1951) was calculated for the four sub-scales and the overall scale. Sub-scale reliabilities were acceptable for factor one (.874), two (.801), three (.816) and four (.751). The overall scale reliability was .880, which indicated high reliability. The four factors accounted for 60.38% of the explained variance. Hair et al (1995) noted that any solution with over 60% of the explained variance is considered an acceptable level in social sciences. Therefore, a solution of the four factors showed stability.
Cluster Analysis To identify possible segments of spectators who attended K-League season games, a cluster analysis was performed with factor scores produced from the 19 variables of the four attendance factors. The cluster solution was validated by first using Ward's clustering analysis on two split-half samples. The significant increase of the amalgamation coefficients and the patterns of the dendrograms across two sets of cluster solution were considered for an initial cluster solution. As a result, a solution of four clusters was identified.
The number of clusters was used as an input for the K-mean cluster analysis (non-hierarchical method) on all respondents (N=967). The cluster centres for the four attendance-dimensions (factors) are shown in Table 3 and illustrated below. The first cluster consisted of 228 members. The individuals in this cluster defined themselves by their views that the promotion factor was (and the price factor was not) influential on game attendance in the K-League. The 206 members of the second cluster had a positive view of place factor, but the promotion factor was not influential on game attendance. The 331 members of the third cluster believed that all factors were influential on attending K-League season games. For this third cluster, price was an especially influential factor. The fourth cluster consisted of 202 members. Those in this cluster were of the opinion that all factors were not influential on attending the professional soccer games. Furthermore, they seemed very indifferent to the place factor. As detailed below, these four clusters were then examined by the multiple discriminant and cross-tabulation analyses in order to provide a more detailed examination of each cluster.
Multiple Discriminant Analysis Following the factor and cluster analyses, the 19 attendance variables were employed as predictors in the MDA. The prior four groups of spectators who attended K-League games were used in the same model as a dependent variable. The objective of the MDA was to identify which variable(s) representing factors affecting attendance best discriminated among the four clusters obtained by a cluster analysis. To do this, the total sample of respondents was randomly divided into two parts: the analysis sample (n=486) and the holdout sample (n=481). As is illustrated in Table 4, the three discriminant functions accounted for 46.9%, 29.0%, and 24.1%, respectively, of the variance in the clusters based on the 19 variables.
Based on the standardised discriminant function coefficients shown in Table 5, the most important variables were determined for each function. The eight key variables within function one were the team's place in league standings, rivalry of team and opponent over the course of the season, special event occasions, free promotional items, proper entrance fee, proper parking fee, proper price of food and beverages, and proper price of merchandise. There were nine variables under the second function. These nine variables included the presence of star athletes on the roster of the home team, playing style, team history, record-breaking performances, ease of facility access, cleanliness of the facility, facility convenience, proper game duration, and proper starting time. The two key variables in the third function were the availability of games on television and advertisements for the game.
To assess the validation of the discriminant results, the proportional chance criterion ([C.sub.pro]=.253) was used to determine the classification accuracy (see Table 6). This criterion was calculated by summing the squared proportion that each cluster represented of the sample. Then, based on the requirement that model accuracy is 25% better than the chance criterion (Hair et al, 1987), the chance criterion was multiplied by 1.25. The final standard used to test the model's validity was 31.6%. The criterion was compared with the hit ratio scores generated from the analysis sample and the holdout sample. The hit ratio scores for the analysis group and holdout group were 94.2% and 92.5%, respectively. In other words, the improvement over chance showed about 68.9% for the analysis group and 67.2% for the holdout group. The results confirmed that the discriminant analysis predicted more accurately than what could be accomplished through chance alone.
Cross-Tabulation Analysis As noted above, the four market segments of the professional soccer industry in Korea were identified through the cluster and multiple discriminant analyses. A cross-tabulation method was then used to explain the demographic and lifestyle profiles for each cluster. Also, the chi-square statistics were utilised to determine whether any statistically significant differences existed among the four clusters. The demographic and lifestyle profiles of these four segments are presented in Tables 7 and 8.
Based on the results of a chi-square analysis, the four distinct market segments showed statistically significant differences across four demographic variables. These differences were found with final degree of education ([X.sup.2]=30.25, df=12, p<.01), occupation ([X.sup.2]=25.34, df=15, p<.05), household monthly income ([X.sup.2]=36.50, df=18, p<.01) and age ([X.sup.2]=27.97, df=12, p<.01). Only one demographic variable (gender [[X.sup.2]=2.34, df=3, p=.505]) was not significant.
Also, the five different lifestyle variables employed in this research showed statistically significant differences across the four clusters. These variables included: number of times the spectator watched K-League games ([X.sup.2]=55.83, df=12, p<.001), transportation to stadium ([X.sup.2]=46.00, df=18, p<.001), type of companion attending the game with the spectator ([X.sup.2]=26.24, df=15, p<.05), media used in order to know the game schedule ([X.sup.2]=77.72, df=24, p<.001), and purchasing experience of team products ([X.sup.2]=20.16, df=3, p<.001).
This study was conducted in an attempt to segment K-League spectators into homogeneous groups on the basis of attendance factors. Four separate segments were revealed through the use of factor analysis and cluster analysis. The MDA established the validity of the four segments. This was because the 19 variables used to measure attendance factors significantly differentiated the four segments through a comparison of the analysis sample and the holdout sample. Finally, the 10 demographic and lifestyle variables were examined by using the chi-square goodness-of-fit statistic and the cross-tabulation analysis. Among these 10 variables, all but one (gender) were significantly different on the four segments identified at the significant level of .05. There are several marketing implications that emerge from this analysis of the demographic and lifestyle profiles of game attendees and why these spectators attend K-League games. Based on the results of this research, each distinct segment had its own particular characteristics. Therefore, several aspects of this study need to be accentuated to assist practitioners in the development of marketing strategies in the professional soccer industry in Korea. The following paragraphs will detail the characteristics of each segment along with the marketing implications for that segment.
Profile of each cluster and marketing implications
Cluster 1 This cluster included 228 survey respondents (23.6%). The members of this group had a positive view of the promotion factor when they attended games in the K-League. They also, however, had a negative view of the price factor. Therefore, while they tended to be sensitive to the promotional factor, they seemed to be unconcerned about the price factor. Because of these characteristics, the individuals in this cluster would be placed together in the 'promotion-concerned group'. This group, which consisted of a significantly high percentage of students (60.5%), tended to find game schedules through television consumption (18.9%). The implications of these findings are that if marketers are able to arrange special events to attract spectators' attention, they should emphasise the events in their promotions. Furthermore, as the spectators in this group are less sensitive to price, they may be willing to pay more for events or gifts if marketers are able to provide attractive events and gifts for the spectators.
Cluster 2 The 206 members of this cluster (21.3%) exhibited a very positive view of the place factor but had a very negative view of the promotion factor. While they seemed to attend games due to the place factor, they were unlikely to be concerned about the promotion factor. Thus, the individuals in this cluster could be placed together in a group referred to as the 'place-concerned group'. Among the four groups, this group was distinguished by having the highest percentages across seven unique items. These items included undergraduate or higher education (63.1%), white-collar employment (26.7%), 25-to-45 year old age bracket (47.6%), use of the internet to secure game schedules (23.8%), middle-class income level (49.0%), game attendance with family members (30.1%) and the purchase of team products (40.8%). One possible conclusion related to these findings is that the spectators in this group perceive going to a professional soccer venue in Korea as a way to escape from the ordinary activities of everyday life. Thus, if sports marketers are able to provide the spectators with a conveniently accessed and clean facility, they may be able to increase the spectator satisfaction levels for attendees who belong to this group.
Cluster 3 With 331 members, this group is the largest (34.2%) of the four clusters. Although spectators in this group had positive views of all factors, they were especially positive in relation to the price factor. Thus, this cluster may be referred to as the 'price-concerned group'. While this group constituted the largest cluster, 39.6% had monthly incomes of less than $2,000 and 48.0% were infrequent game attendees (i.e. fewer than three games). The spectators in this group exhibited the lowest percentage (23.0%) of the four clusters in terms of purchasing team products. Furthermore, this group showed more of an inclination than the other groups to secure game information and schedules through correspondence with other people (53.8%). Because this group is the largest of the clusters, sports marketers in the K-League should make an effort to attract spectators that fit into this group. Such target marketing is very challenging as the members of this group are assertive on such factors as price, place, promotion and product. Furthermore, they have the lowest monthly income and the weakest spending patterns.
Cluster 4 As the smallest (20.9%) of the four clusters, with only 202 members, the spectators in this group showed a negative view relative to all of the factors. This means that these spectators attended a professional soccer game without regard to any of the distinct factors. Therefore, the members of this cluster could be termed the 'indifferent group'. It is significant in this group that 21.8% fit into the highest monthly income level (i.e. more than $4,000) and 31.0% had purchased team-related products. Furthermore, members of this group often attended games with their friends (57.4%) and gathered information regarding games and schedules through correspondence with other people (53.0%). This group should be considered a key target market as the spectators in this cluster possessed the highest income of the four clusters. Furthermore, they had relatively strong spending patterns. To understand the spectators in this group more precisely, the sports marketers in the K-League should attempt to determine other crucial factors besides price, place, promotion and product that might influence their attendance patterns.
Overall, this study was designed to provide marketing information and analysis concerning the spectators of professional soccer in Korea. The specific focus of this marketing analysis was the K-League. The study, which was based on the survey responses of K-League spectators (N=967), involved two investigative aspects. The first part of the study included the clustering of spectators into homogeneous groups that had similar characteristics on the basis of attitudes towards game attendance. The second part of the analysis involved defining the segments obtained on the basis of the demographic and lifestyle profiles of the spectators. Four distinct segments were identified and named as a result of the multiple analytical steps. These four segments include the promotion-centre group, place-concerned group, price-concerned group and the indifferent group. Sports marketing practitioners in the K-League can use the results and analysis provided in this study to assist them in their development of a systematic marketing plan that effectively and efficiently satisfies the needs and desires of their target markets.
Sangho Kim is an assistant professor in the Business School at Kyungdong University in South Korea. He teaches sports management--specialising in consumer behaviour, public relations and marketing research methods. He has contributed to several research projects and published widely in the field of sports marketing.
Euidong Yoo is a senior researcher in the Department of Sport Policy and Development of the Korea Institute of Sport Science in South Korea. His primary research interests encompass sports management, sports marketing, economic impacts of sports events and policy-making for sports industry promotion.
Paul M. Pedersen, an associate professor of sport communication at Indiana University, received his PhD in sports management from Florida State University. Pedersen has published three books and more than 30 peer-reviewed articles in national or international journals. He is also an editorial review board member for five sports journals.
Abell, D.F. & Hammond, J.S. (1979) Strategic Market Planning: Problems and Analytical Approaches. New Jersey: Prentice-Hall.
Baade, R.A. & Tiehen, L.J. (1990). An analysis of major league baseball attendance, 1969-1987, Journal of Sport and Social Issues 14 (1), 14-32.
Bauer, H.H., Sauer, N.E. & Exler, S. (2005) The loyalty of German soccer fans: does a team's brand image matter?, International Journal of Sports Marketing and Sponsorship 7 (1), 14-22.
Bristow, D.N. & Sebastian, R.J. (2001) Holy cow! Wait 'til next year! A closer look at the brand loyalty of Chicago Cubs baseball fans, Journal of Consumer Marketing 18 (3), 256-275.
Chae, H. & Lee, J. (2000) An investigation of factors related to sport fan behavior influencing attendance at Korean professional sports, Korean Journal of Sport Management 5 (2), 137-154.
Choi, Y. (1999) An examination of marketing strategies for professional soccer in Korea Unpublished Master's thesis, Kyunghee University, Seoul, Korea.
Croft, M.J (1994) Market Segmentation. London: Routledge.
Cronbach, LJ. (1951) Coefficient alpha and the internal structure of tests, Psychometrika 16 (3), 297-333.
Dale, B., Iwaarden, J., Wiele, T. & Williams, R. (2005) Service improvement in a sports environment: a study of spectator attendance, Managing Service Quality 15 (5), 470-484.
DeSchriver, T.D. & Jensen, P.D. (2002) Determinants of spectator attendance at NCAA division II football contests, Journal of Sport Management 16, 311-330.
Dibb, S. & Simkin, L. (1996) The Mmarket Segmentation Workbook: Target Marketing for Marketing Managers. London: Routledge.
Dolnicar, S. (2005) Understanding barriers to leisure travel: tourist fears as a marketing basis, Journal of Vacation Marketing 11 (3), 197-208.
Ehrman, C. (2006) On using benefit segmentation for a service industry: a study on college career counseling services, The Journal of American Academy of Business 8 (2), 179-185.
Funk, D.C. (2002) Consumer-based marketing: the use of micro-segmentation strategies for understanding sport consumption, International Journal of Sports Marketing and Sponsorship 4 (3), 231-256.
Godin, G. (1993) The theories of reasoned action and planned behavior: overview of findings, emerging research problems and usefulness for exercise promotion, Journal of Applied Sport Psychology 5, 141-157.
Gonzalez, A.M. & Bello, L. (2002) The construct "lifestyle" in market segmentation: the behavior of tourist consumers, European Journal of Marketing 36 (1/2), 51-85.
Hansen, H. & Gauthier, R. (1989) Factors affecting attendance at professional sport events, Journal of Sport Management, 3 (1), 15-32.
Hair, J.F., Anderson, R.E. & Tatham, R.L. (1987) Multivariate Data Analysis With Readings (4th edn). New Jersey: Prentice-Hall.
Jackie, L.M. & Susan, H.C. (1998) Research note: the psychographic segmentation of the female market in Greater China, International Marketing Review 15 (1), 61-77.
Kennelly, A. (1999) 'Discriminant analysis' in N.K. Malhotra (ed.), Marketing Research: An Applied Orientation (3rd edn, 559-584). New Jersey: Prentice-Hall.
Kim, J.H. (1998) Ticket sales, Sport Business 4, 34-39.
Kim, J.K. & Cho, K.M. (2003) The effects of marketing mix factors on team loyalty of pro-soccer spectators, Korean Journal of Sport Management 7 (2), 113-126.
Kim, Y.M. (2004) The relationship among attributes of a professional football team, team identification, and team loyalty, Korean Journal of Physical Education 43 (6), 623-631.
Kolbe, R. & James, J.D. (2000) An identification and examination of influences that shape the creation of a professional team fan, International Journal of Sports Marketing and Sponsorship 2 (1), 23-37.
Kotler, P. (2000) Marketing Management (5th edn), New Jersey: Prentice-Hall.
Krohn, F.B., Clarke, M., Preston, E., McDonald, M. & Preston, B. (1998) Psychological and sociological influences on attendance at small college sporting events, College Student Journal 32 (2), 277-287.
Laverie, D.A. & Arnett, D.B. (2000) Factors affecting fan attendance: the influence of identity salience and satisfaction, Journal of Leisure Research 32 (2), 225-246.
Madrigal, R. (2000) The influence of social alliances with sports teams on intentions to purchase corporate sponsors' products, Journal of Advertising 29 (4), 13-24.
Mitchell, V.W. (1994) How to identify the psychographic segment (Part 1 and Part 2), Marketing Intelligence and Planning 12 (7), 4-17.
Mullin, B.J., Hardy, S. & Sutton, W.A. (2000) Sport Marketing (2nd edn). Champaign, IL: Human Kinetics.
Myers, J.H. (1996) Segmentation and Positioning for Strategic Marketing decisions. Chicago: American Marketing Association.
Pan, D.W., Zhu, Z., Gabert, T.E. & Brown, J. (1999) Team performance, market characteristics, and attendance of major league laseball: a panel data analysis, The Mid-Atlantic Journal of Business 35 (2/3), 30-39.
Park, S.H., Chun, H.M. & Kim, Y.M. (2003) Sport Marketing (2nd edn). Seoul: Hakhyun.
Punj, G. & Stewart, D.W. (1983) Cluster analysis in marketing research: review and suggestions for application, Journal of Marketing Research 20, 134-148.
Rhee, H.S., Cho, K.M. & Chin, J.H. (2004) A comparative analysis of pro-soccer and pro-baseball spectators' characteristics, Korean Journal of Physical Education 43 (6), 689-701.
Richardson, B. & O'Dwyer, E. (2003) Football supporters and football team brands: a study in consumer brand loyalty, Irish Marketing Review 16 (1), 43-53.
Russell, D. (1997) Football and the English. London: Carnegie.
Sharma, A. & Lambert, D.M. (1994) Segmentation of markets based on customer service, International Journal of Physical Distribution and Logistics Management 24 (4), 50-58.
Tapp, A. & Clowes, J. (2002) From 'carefree casuals' to 'professional wanderers': segmentation possibilities for football supporters, European Journal of Marketing 36 (11/12), 1248-1269.
Walley, K., Custance, P. & Parsons, S. (2000) UK consumer attitudes concerning environmental issues impacting the agrifood industry, Business Strategy and the Environment 9, 355-366.
Woo, K.S. (1998) Using quality perceptions to segment customers in services, Marketing Intelligence and Planning 16 (7), 418-424.
Yoo, E. (2004) Application of Pro-Sports to Enhance Regional Economic Conditions. Seoul, Korea: The Korea Institute of Sport Science.
Department of Sport Management in School of Business, Kyungdong University (217-600) Kangwon-do Sokcho-si Sokcho PO Box 57-ho, South Korea
Tel: + 11 82 33 639 0356
Research Fellow, Department of Policy Development, Korea Institute of Sport Science
Paul M. Pedersen
Associate Professor, Sport Communication & Sport Marketing, Indiana University, USA
TABLE 1 Items influencing attendance MEAN STANDARD RANKING BY MEAN FACTORS SCORE DEVIATION SCORES PLACE CONVENIENT FACILITY 2.89 1.27 8 PROPER STARTING TIME 3.04 1.26 5 PROPER GAME DURATION 3.03 1.28 6 CLEANLINESS OF FACILITY 2.84 1.28 10 EASY ACCESS TO FACILITY 2.81 1.35 11 PRODUCT TEAM HISTORY 3.05 1.25 4 OFFENSIVE PLAYING STYLE 3.10 1.18 3 RECORD-BREAKING PERFORMANCE OF 3.11 1.22 2 HOME TEAM AND ATHLETES RIVALRY OF TEAM AND 2.91 1.16 7 OPPONENT OVER SEASON STAR ATHLETES ON HOME TEAM ROSTER 3.13 1.28 1 TEAM'S PLACE IN LEAGUE STANDINGS 2.86 1.24 9 PRICE PROPER PRICE OF FOOD AND BEVERAGES 2.11 1.24 19 PROPER PRICE OF MERCHANDISE 2.16 1.28 18 PROPER PARKING FEE 2.23 1.38 17 PROPER ENTRANCE FEE 2.63 1.27 13 PROMOTION MEDIA ADVERTISING GAMES 2.66 1.23 12 SPECIAL EVENT OCCASIONS 2.39 1.21 15 NO GAMES ON TV 2.51 1.33 14 LEAGUE'S GRATITUDE GIFTS 2.36 1.18 16 TABLE 2 Results of the factor analysis FACTOR VARIANCE RELIABILITY FACTORS LOADING EIGENVALUE (%) ALPHA PLACE 6.098 32.095 .874 CONVENIENT FACILITY .830 PROPER STARTING TIME .808 PROPER GAME DURATION .786 CLEANLINESS OF FACILITY .781 EASY ACCESS TO FACILITY .667 PRODUCT 2.528 13.305 .801 TEAM HISTORY .734 OFFENSIVE PLAYING STYLE .731 RECORD-BREAKING PERFORMANCE .718 OF HOME TEAM AND ATHLETES RIVALRY OF TEAM AND OPPONENT .660 OVER SEASON STAR ATHLETES ON HOME TEAM .638 ROSTER TEAM'S PLACE IN LEAGUE .623 STANDINGS PRICE 1.796 9.451 .816 PROPER PRICE OF FOOD AND .866 BEVERAGES PROPER PRICE OF MERCHANDISE .849 PROPER PARKING FEE .735 PROPER ENTRANCE FEE .507 PROMOTION 1.051 5.529 .751 MEDIA ADVERTISING GAMES .745 SPECIAL EVENT OCCASIONS .684 NO GAMES ON TV .671 LEAGUE'S GRATITUDE GIFTS .652 OVERALL SCALE 60.380 .880 TABLE 3 Results of the cluster analysis--final cluster centres CLUSTERS CLUSTER CLUSTER CLUSTER CLUSTER FACTORS 1 2 3 4 F-VALUES P-VALUES PLACE .09 .86 .08 -1.21 6.75 .000 PRODUCT .24 .04 .05 -.39 319.98 .000 PRICE -.73 -.47 1.06 -.42 437.24 .000 PROMOTION .98 -.88 .25 -.62 359.64 .000 NUMBER OF CASE 228 206 331 202 N.B. The factor scores are standardised with a mean of zero and a standard deviation of unity. F-tests should be used only for descriptive purposes because the clusters have been chosen to maximise the differences among cases in different clusters. The p-value of .000 means p<.001. TABLE 4 Results of the Multiple Discriminant Analysis--discriminant functions PERCENTAGE CANONICAL WILK'S FUNCTION EIGENVALUE VARIANCE CORRELATION LAMBDA CHI-SQUARE DF 1 1.980 46.9 0.815 .075 1227.776* 57 2 1.225 29.0 0.742 .223 710.696* 36 3 1.016 24.1 .710 .496 331.928* 17 *significant at p<.001 TABLE 5 Results of the Multiple Discriminant Analysis--standardised canonical discriminant function coefficients STANDARDISED COEFFICIENTS FUNCTION FUNCTION FUNCTION ATTENDANCE VARIABLES 1 2 3 TEAM'S PLACE IN LEAGUE STANDINGS .013* .021 .127 STAR ATHLETES ON HOME TEAM ROSTER .030 .186* .071 OFFENSIVE PLAYING STYLE -.036 .042* -.064 RIVALRY OF TEAM AND OPPONENT OVER SEASON .052* -.191 .052 TEAM HISTORY -.083 -.041* .014 RECORD-BREAKING PERFORMANCE OF HOME TEAM .074 .039* .011 AND ATHLETES NO GAMES ON TV .160 -.012 .553* SPECIAL EVENT OCCASIONS .075* -.178 .396 GAME ADVERTISEMENT .217 .132 .439* LEAGUE'S GRATITUDE GIFTS .120* -.014 .269 PROPER ENTRANCE FEE .125* -.058 -.060 PROPER PARKING FEE .248* -.216 -.261 PROPER PRICE OF FOOD AND BEVERAGES .202* -.336 -.292 PROPER PRICE OF MERCHANDISE .437* -.194 -.238 EASY ACCESS TO FACILITY -.007 .199* -.173 CLEANLINESS OF FACILITY .153 .143* -.210 CONVENIENT FACILITY -.108 .360* -.057 PROPER GAME DURATION .172 .387* -.011 PROPER STARTING TIME .186 .258* -.085 * the most important variables for each function on the basis of 'Structure Matrix' TABLE 6 Results of the Multiple Discriminant Analysis--classification matrix PREDICTED GROUP MEMBERSHIP ACTUAL GROUP SAMPLE SIZE C1 C2 C3 C4 TOTAL C1 114 103 6 4 1 124 (90.4%) (5.3%) (3.5%) (.9%) C2 113 2 102 6 3 97 (1.8%) (90.3%) (5.3%) (2.7%) C3 141 1 1 139 0 131 (.7%) (.7%) (98.6%) (0.0%) C4 118 3 0 1 114 131 (2.5%) (0.0%) (.8%) (96.6%) TOTAL 486 109 109 150 118 486 Classification based on weighted group probabilities; classification accuracy 94.2%, [C.sub.pro] = .253 ([C.sub.pro] = (.235)[.sup.2]+ (.233)[.sup.2]+(.290)[.sup.2]+(.243)[.sup.2] = .253) TABLE 7 Demographic profiles of clusters (all values are percentages) CHARACTERISTICS CLUSTER 1 CLUSTER 2 CLUSTER 3 CLUSTER 4 GENDER MALE 53.5 56.8 50.8 50.5 FEMALE 46.5 43.2 49.2 49.5 FINAL DEGREE OF EDUCATION** MIDDLE SCHOOL 4.4 3.9 0.6 1.5 HIGH SCHOOL 42.5 32.0 47.7 47.0 UNDERGRADUATE 48.2 56.3 48.3 44.6 GRADUATE AND ABOVE 3.5 6.8 2.7 6.4 OCCUPATION* STUDENT 60.5 47.1 57.4 52.5 WHITE-COLLAR 21.5 26.7 23.0 25.2 SERVICE 3.1 4.9 3.3 7.4 FREELANCER 4.8 4.9 5.7 4.0 HOUSEWIFE 6.1 12.1 5.4 4.5 OTHERS 3.9 4.4 5.1 6.4 HOUSEHOLD MONTHLY INCOME ** LESS THAN 1,000 11.4 9.2 17.2 13.4 1,000 TO LESS THAN 2,000 23.2 18.0 22.4 23.8 2,000 TO LESS THAN 3,000 23.2 25.2 24.8 22.3 3,000 TO LESS THAN 4,000 18.9 23.8 14.2 11.4 4,000 TO LESS THAN 5,000 3.5 8.7 7.6 6.9 5,000 AND OVER 12.7 9.7 6.6 14.9 AGE** LESS THAN 25 65.8 48.1 63.1 64.4 25 TO LESS THAN 35 16.2 31.6 18.1 20.3 35 TO LESS THAN 45 14.9 16.0 13.9 13.9 45 TO LESS THAN 55 3.1 3.9 4.2 1.5 55 AND OVER 0.0 0.5 0.6 0.0 * p<.05; ** p<.01; *** p<.001 TABLE 8 Lifestyle profiles of clusters (all values are percentages) CLUSTER CHARACTERISTICS CLUSTER 1 CLUSTER 2 CLUSTER 3 4 NUMBER OF K-LEAGUE GAMES*** LESS THAN 3 38.2 26.3 48.0 39.7 3 TO LESS THAN 6 25.0 30.2 26.7 19.1 6 TO LESS THAN 9 7.0 3.9 7.0 3.0 9 TO LESS THAN 12 16.2 17.6 10.0 16.1 12 AND OVER 13.6 22.0 8.2 22.1 TRANSPORTATION TO STADIUM*** CAR 47.8 60.7 44.4 45.0 SUBWAY 1.8 2.9 1.2 3.5 BUS 32.5 20.9 34.7 29.7 TAXI 5.7 10.2 6.3 11.4 WALKING 9.2 3.4 11.2 7.4 TRAIN 3.1 1.5 1.5 1.0 OTHER 0.0 0.5 0.6 2.0 TYPE OF COMPANION IN STADIUM* FAMILY 23.2 30.1 21.5 17.8 COMPANY COMPANION 8.3 9.2 8.8 10.4 FRIEND 53.5 49.0 57.4 57.4 PERSON WHO IS IN LOVE 7.5 6.3 5.7 3.5 SELF 5.7 4.9 3.6 5.0 OTHER 1.8 0.5 3.0 5.9 SOURCE FOR GAME SCHEDULES*** NEWSPAPER 3.1 2.9 3.3 3.0 RADIO 0.4 0.5 0.9 0.0 TELEVISION 18.9 9.7 13.6 5.0 INTERNET 15.4 23.8 11.2 16.3 OTHER PEOPLE 38.6 33.5 53.8 53.0 BOOKLET 4.8 4.4 5.7 6.4 PREVIOUS GAME 14.9 19.4 7.9 8.9 OTHER 3.9 5.8 2.7 5.9 PURCHASE TEAM PRODUCTS*** YES 27.6 40.8 23.0 31.7 NO 72.4 59.2 77.0 68.3 * p<.05; ** p<.01; *** p<.001…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Market Segmentation in the K-League: An Analysis of Spectators of the Korean Professional Soccer League. Contributors: Kim, Sangho - Author, Yoo, Euidong - Author, Pedersen, Paul M. - Author. Journal title: International Journal of Sports Marketing & Sponsorship. Volume: 8. Issue: 2 Publication date: January 2007. Page number: 141+. © 2003 International Marketing Reports Ltd. COPYRIGHT 2007 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.