The scale and influence of Japanese travel agencies in the global tourism business are significant. Data from the Japanese Ministry of Land, Infrastructure, Transport and Tourism, as of May 17, 2012, showed that 59 Japanese travel agencies handled travel products worth 6,049,036,397,000 yen (about $60 billion) from April 2011 to March 2012. These Japanese travel agencies are top global players in terms of customer service, travel products, overall travel service delivery, etc. However, competition among these agencies is getting intense due to limited new travel products, the stagnant number of total travelers, emergence of new competitors, difficulties in differentiating travel services, and the like. In addition, travel agencies face many risk factors related to the increased social status of travel customers [1]. Travel agency managers in Japan are currently feeling a sense of crisis.

There have been numerous previous studies on travelers and marketing of travel agencies using psychological and environmental factors that influence traveler satisfaction [2], past behavior [3], and travelers’ intention to revisit places based on their expectations and past travel experience [4].

This stream of research has led to studies on various expectations of travelers when they select certain travel products or tourist destinations. One study identified the opportunity of a travel product involving festivals which provide an experience of a local culture [5]. Other studies explored how to use newspaper advertisements to attract tour groups from the travel agency’s marketing strategy perspective [6] and analytical evaluation of the importance of marketing mix on traveler satisfaction [7]. However, the ‘traveler decision making process’ and its impact on satisfaction with travel and post behavior are at the core of most studies. Confirmatory or exploratory studies on major factors or motivators related to the traveler’s product purchase or travel decisionmaking are conducted for a variety of regions, research targets and destinations. These studies demonstrate causal relationships between influencing factors related to travel decision making and post behavior based on satisfaction with travel products [8,9].

However, additional studies are needed to identify measurements that can provide effective information for the travel agency’s strategy. Tasks performed in additional studies will shed new light on the concept of selection attributes of travel agencies that has been conducted by existing studies and allow for changes in evaluation variables. One of the important tasks is to step forward to develop studies in similar fields by comparing with the existing causal relationship model and adding new factors to the existing independent variables. This study is expected to support the travel agency for more effective product development and management strategy based on the recognition that the most important strategy for travel agencies under the current fierce competitive environment is to secure customers with a high level of loyalty.

This study begins with a thorough review of current literature regarding travel product selection and the purchase decision-making process, as well as post-purchase behavior. After this, the hypotheses to be tested are presented. The next section outlines the survey’s methodology, followed by the results. Finally, the results are analyzed and a conclusion is reached for all hypotheses.

Selection of tour products

Previous studies on purchasing travel products through a travel agency have focused on ‘travel decision-making’. However, the importance of selection criteria of tour products and their influence need to be studied from the business perspective. The selection criteria of tour products involve the way to offer more professional product compositions from a marketing perspective based on the selection criteria of tour destinations. Haahti [10] presented 10 selection criteria for tour destinations including currency value, accessibility, night life and recreation. Crompton and Ankomah [11] broadly classified selection criteria of tour destinations into the needs satisfaction level, social level and tour ability level. They further developed details of each selection criterion from the psychological perspective of tourists. In addition, the priority and importance of selection criteria in purchasing a package tour product provided by a travel agency were explored by other researchers [12,13].

Selection criteria of tour products presented by various studies include selection factors of tour destination, common attributes (environment of destination, tour activities, accommodation, transportation, food and beverage, shopping, etc.), newly added attributes of tour products (travel agency, tour guide, product price, schedule and considerations before purchasing a tour product). Thus, a review of new attributes that consider the current travel trend and style change is needed to measure selection criteria of tour products. In this regard, recent studies on selection criteria of tour products focus on core attributes from the business perspective. Bowie and Chang [14] suggested travel procedure and service offered by a provider (travel agency), including tour guide and assistance service, are core factors in selecting tour products, which are important for satisfaction with travel. Jin et al. [15] argued for the importance of option production in selection criteria of tour products and its implications from the perspective of a travel agency’s marketing, based on a comparative study of preferences for various options including upgrades or downgrades for the same tour product.

Travel decision-making and post behavior

According to the early ‘travel decision-making model’, a traveler is more influenced by psychological and environmental factors including traveler’s awareness and characteristics of tour activities than by economic factors [2]. Travelers tend to evaluate the overall travel experience after they travel with their own travel plans or with a tour product purchased through a travel agency. Many studies have measured the traveler’s evaluation of a tour based on the level of satisfaction with travel. Regarding studies on travelers’ satisfaction, the main focus has been on causal relationships among decision-making procedure of the traveler, satisfaction and the intention to revisit the tourist destination (re-purchase). Other studies explored the difference between those who were satisfied with their tour destination and those who were not in terms of the possibility of revisiting the destination [3] and the impact of differences between the traveler’s expectations before traveling and the level of satisfaction after coming back from the visit [4]. However, as Geva and Goldmand [16] claimed, satisfaction with travel is not necessarily related to the possibility of revisit as time passes after travel, contrary to studies that suggested the causal relationship between satisfaction and post-travel behavior [17].

Previous studies related to the ‘travel decision-making model’ focused on the demonstration of a structural causal relationship among image of tourist destination, travel quality, recognized value, satisfaction and revisit. Chen and Tsai [8] found that the positive recognized value of a tour leads to satisfaction with travel which in turn leads to revisit. In addition, attributes of friendly service, good quality accommodation and the opportunity to spend time with family or friends were identified as positive factors for satisfaction if the tour destination is a recreational place like a resort [18]. In addition, the satisfaction level with a tour guide service in a tour package has no significant relationship with the overall satisfaction even though the tour guide service has a positive impact on satisfaction with each travel service [19].

Studies have also investigated the causal relationship between satisfaction with travel and revisit as post behavior. This stream of research searches for various kinds of information to develop marketing strategies using loyalty, a useful concept to measure complex psychology of travelers. The traveler’s loyalty to a tour destination has a positive impact on satisfaction, while Push Motivation has a positive and Pull Motivation a negative impact on satisfaction [20]. In addition, a study on the causal relationship among the image of tour destination, tour attributes, satisfaction with travel, and loyalty to the tour destination found that a positive image of a tour destination has a positive impact on satisfaction and a high level of satisfaction leads to loyalty toward tour destinations [21]. Another study on personal loyalty of travelers showed positive results on expectations (image, experience, communication and tangible) have positive impacts on the satisfaction level that leads to one’s loyalty in a positive way [22]. After analyzing structural relationships by adding Disconfirmation, Positive Emotion, Negative Emotion, Destination Image, this research results showed that positive emotion and negative emotion have a positive and a negative impact (respectively) on satisfaction, demonstrating that satisfaction with travel has a positive causal relationship with traveler loyalty [23]. In addition, Lee et al. [24] showed that the traveler’s expectation and motivation have a negative impact on complaints in the structural relationship between satisfaction, complaints and loyalty. The study also found that a traveler with more complaints tended to have difficulty in establishing loyalty.

Post behavior and evaluation of a traveler agency

This study is concerned with how to develop a travel agency’s marketing strategy for the desired post behavior of a traveler. In other words, this study intends to confirm the significance and implication of travel agency evaluation by applying the causal relationship model. The concept of ‘evaluation of the travel agency’ is developed by this researcher to assess the impact of one’s purchase and travel experience with a certain travel agency for considering whether to use the same travel agency for future travel or tour product purchase. The ‘evaluation of the travel agency’ is composed of travel agency selection and travel agency brand.

First, the selection of a travel agency is not usually based on evaluation of physical attributes (facility, location, service, etc.) or information attributes (level of information, reliability, and accessibility) [25]. Rather, it would be better to identify and measure such attributes as image, competitiveness, kindness, responsiveness, accessibility, etc. [26]. This study uses four evaluation items including reliability of travel agencies, convenience(accessibility), evaluation from others, and reasonable price so that it can have the most general and comprehensive information among evaluation items for selecting a travel agency applied by previous studies. Second,

For travel agency brands, this study intends to identify these sets of attributes,

(1) five factors: Awareness, Loyalty, Perceived Quality, Association, and Market Behavior [27];

(2) 10 factors: Leadership/Popularity, Perceived Value, Organizational Association, Brand Personality, Price and Distribution Indices [28]; and

(3) 2 factors: Brand Awareness and Brand Image [29].

Hypotheses development

Selection attributes: Based on the review of previous studies, [10,11,15] the following hypothesis is proposed:

The results of this study showed that a traveler’s decision-making (purchase of tour products) depends on multiple service attributes expected from product components along with numerous selection attributes (physical attributes of destination) provided by travel agencies.

Hypothesis 1: Desirable selection attributes of overseas tour products have positive impacts on purchase decision-making on travel.

Tour product decision making: The causal relationship between tour product decision-making and post behavior (satisfaction, repurchase) is developed based on similar studies of Moscardo [3], Mazursky [4], Geva and Goldmand [16], Chen and Tsai [8], Bowie and Chang [14], Huang et al. [19] and Meng et al. [18].

Post behaviors include satisfaction or dissatisfaction with tour products, recommending others to purchase applicable tour product and repurchase of similar tour products. In previous studies, the casual relationships between satisfaction with travel after purchase of tour products and repurchase decision-making showed positive relationships.

Hypothesis 2: Tour product decision-making has a positive impact on post behavior (satisfaction, re-purchase).

Hypothesis 3: A high level of satisfaction as a post behavior has a positive impact on re-purchase of tour products.

Travel agency brand: Finally, the causal relationship between post behavior and evaluation of a travel agency (selection of a travel agency, travel agency brand) is based on related studies of Yoon and Uysal [20], Chi and Qu [21], Rodríguez del Bosque et al [22], Rodríguez del Bosque and San Martín [23] and Lee et al. [24]. The following hypothesis is suggested:

In other words, previous studies showed that post behaviors including satisfaction with travel, decision-making for repurchasing of applicable tour product, etc. had an impact on the traveler’s loyalty to a tour destination. This study is expected to show that post behaviors will have a positive impact on re-selection of travel agency and its brands at the end.

Hypothesis 4: Favorable post behavior (satisfaction, re-purchase) has a positive impact on selection of a travel agency.

Hypothesis 5: Favorable post behavior (satisfaction, re-purchase) has a positive impact on travel agency brand.

Measurements

Based on a thorough review of relevant literature, the measurements for each construct were determined. Given the fact that the previous studies referenced earlier to identify measurement variables were in English, the contents had to be translated into Japanese and adjusted to make them appropriate for this study. In addition, words were changed into terms easy to be understood by Japanese readers considering cultural and social differences. The double translation protocol was used. The questionnaire with all the parameters (variables) was prepared in English by the researcher. Then, it was translated into Japanese. Then, bilingual faculty members translated it back into English. There was no significant difference between the two English versions.

The ‘selection attributes of an overseas tour product’ were based on the following studies: Haahti [10]; Crompton and Ankomah [11]; Kale et al. [12]; Stovall [13]; Jin et al. [15]. A pilot test was conducted on 27 identified measurement variables, considering various settings (timing, target group, etc.). Next, the measurement variables for tour product decision-making were based on ‘traveler decisionmaking process model’ theory presented in Mathieson and Wall [2] describing the impact (level) of selection attributes of a tour product considered before traveling. In addition, post behavior parameters were based on the precedent studies of Moscardo [3] and Mazursky [4]. Measurement variables were divided into ‘the overall satisfaction (level) after traveling’ and ‘willingness to purchase the overseas tour product again (level)’. Finally, four variables for selecting a travel agency, reliability, convenience, evaluation from others, and reasonable price were measured based on the studies of Kendal & Booms [25] and Leblanc [26]. Four other variables of image, service quality, product differentiation, and price premium were derived from the studies of Aaker [27,28] and Keller [29] for travel agency brand as an asset.

Survey

The first pilot test was conducted with a total of 50 subjects to check feasibility and reliability of the developed questionnaire. After the satisfactory review on reliability and validity of survey items, issues of survey questions, etc. (details of evaluation variables applied by this study are check-up items written in English for the first time. After the first review of vocabulary and contextual flow of survey details required for carrying out a survey targeting Japanese people, a second review on issues of overall reliability and feasibility was conducted based on the results of the Pilot Test), the final version of questionnaire was prepared and used for the main study from July 30, 2012 to Aug. 17, 2012 [30]. The questionnaire consisted of a Likert Scale (7 point scale) from a major evaluation concept perspective, and nominal and ordinal scale was used for other general survey items. The surveyors, students from the Department of Commercial Science of a large Japanese university, were trained to assist in this study. The survey location was Kansai International Airport in Osaka, Japan, where surveyors explained the purpose of the study to randomly selected travelers waiting in the departure lounge, targeting only those who have overseas travel experience during the last 3 years. Out of 480 copies distributed, 435 copies (90.6% of recovery rate) were collected. After removing copies that were not suitable for analysis, 385 copies (response rate of 80.2%) were used for analysis.

Analysis

SPSS 17.0 for Windows was used for data analysis in this study. Analytical methods used include Factor Analysis, Validity and Reliability Analysis, and Multiple Regression Analysis.

Characteristics of target group

Demographic characteristics of survey participants are presented in Table 1. There were 158 males (41%) and 226 females (58.7%) in the sample. The number of university graduates is the highest at 174(45.2%). In terms of age, the number of those between 21 and 30 was the highest at 162(42.1%) and 58.2% (224) were not married. Among those who were married, 17.7% of respondents had two children. The number of those who earned less than 3 million yen (about $30,000) per year was the highest at 218 (56.6%).

Table 1: Demographic Characteristics of Survey Participants.

Validity and reliability analysis

Principal Component Analysis (PCA) and reliability analysis were conducted to check validity and reliability of measurement variables for major constructs in this study. PCA (Principle Component Analysis) and Varimax were used for extraction from factors, and results of reliability analysis were checked through Cronbach ɑ. The analysis results are presented in Table 2. PCA results on 27 variables identified as selection attributes of tour products found the KMO value (.860) with which validity can be determined, and Bartlett’s Sphericity (4113.005) showed the level of significance at 000 (ρ<.01) and factor loadings (higher than 4) and Cronbach’s α (higher than .6) showed the level of effectiveness. PCA results showed that 27 variables were clustered into 6 components which were labeled as destination information, variety of tour activities, differentiation of product, convenience, shopping, uniqueness of the place, and environment of the place.

Table 2: Result of Validity and Reliability Analysis on Selection Attributes of Tour Products.

PCA results on the four selection attributes of the travel agency are presented in Table 3. The KMO value (.668) with which validity can be determined and Bartlett’s Sphericity (359.667) showed the level of significance at .000 (ρ<.01) and factor loadings (higher than 4) and Cronbach’s α (.721) showed the level of effectiveness.

Table 3: Result of Validity and Reliability Analysis regarding Selection of a Travel Agency.

Finally, PCA results on the four measurement variables related to brand of the travel agency are presented in Table 4. The KMO value (.746) and Bartlett’s Sphericity (415.726) showed the level of significance at 000 (ρ<.01) and factor loadings and Cronbach’s α (.774) showed the level of effectiveness.

Table 4: Result of Validity and Reliability Analysis regarding Brand of the Travel Agency.

Test of hypotheses

The results of multiple regression analysis to test hypotheses are presented in Table 5. Here are the test results. Given that this study tested relationships among selection attributes of tour products, decision-making for purchase of tour products and evaluation of travel agencies in an exploratory manner, it was determined that the partial significance level (partially supported) was p<.1 which is a relatively positive one compared to that of p < .05 which is generally used.

Table 5: Summary of Hypothesis Test Results.

For Hypothesis 1, R2 on the impact of six selection attributes of tour products as independent variables on one’s willingness to purchase a tour product as a dependent variable was 0.48. The F value which supports the hypothesis was 3.090 with the significance level of ρ<.01. Regression analysis results showed that independent variables which have positive impacts on one’s willingness to buy a tour product within the level of statistical significance were travel information, differentiation of travel product, convenience, shopping and uniqueness of the place. Two other factors of tour activity diversity and environment of the place showed no significance. Travel information had the significance probability of .073 (ρ<.1, β=.082) when t value was 1.800, differentiation of travel products had the significance probability of .016 (ρ<.05, β=.110) when t value was 2.415. Convenience of travel and shopping had the significance probability of .074 (ρ<.1, β=.081) when t value was 1.791 and uniqueness of place showed the significance probability of .014(ρ<.05, β=.112) when t value was 2.473. Hypothesis 1 has six sub-hypotheses. Four were supported and two were not. So, overall, H1 was supported. The result of the Hypothesis 1 test found that the factor that had the greatest impact on one’s willingness to buy a tour product was uniqueness of the place, followed by differentiation of the tour product, travel information, and convenience of travel and shopping.

The test result for Hypothesis 2 showed the R2 value between one’s willingness to buy a tour product as an independent variable and satisfaction with travel was .084. The F value was 35.044 at the significance level of .000. H2 is thus supported. In addition, the R2 value which shows goodness of fit between one’s willingness to buy a tour product as an independent variable and repurchase of the tour product as the dependant product was .085. The F value was 35.568 at the significance level of .000. Regression analysis result showed that one’s willingness to buy a tour product had the significance probability of .000 (ρ<.01, β=.254) when t value was 5.920 and repurchase of tour product had the significance probability of .000 (ρ<.01, β=.286) when t value was 5.964 indicating that both had positive impacts.

The test result for Hypothesis 3 showed an R2 value of 0.343 regarding the impact of satisfaction with travel as an independent variable on repurchase of a tour product as the dependent variable. The F value was 199.197 with the significance probability of .000. Regarding the satisfaction with travel and one’s willingness to purchase a tour product, t value was 14.114 and the significance probability was .000 (ρ<.01, β=.654). H3 is also supported.

The test result for Hypothesis 4 had an R2 of 0.75 regarding the impact of satisfaction with travel and repurchase of the tour product as independent variables on selection of the travel agency as the dependent variable. H4 is thus supported. The F value was 15.548 with the significance probability of .000. The result showed that satisfaction with travel had the significance probability of .003 (ρ<.01, β=.230) when t value was 2.964 and repurchase of the tour product showed significance probability of .037 (ρ<.05, β=.146) when t value was 2.094, indicating that both had positive impacts on selection of the travel agency.

The test result for Hypothesis 5 showed an R2 value of 0.141 regarding the impact of satisfaction with travel and repurchase of the tour product as independent variables on the travel agency’s brand. The F value was 31.211 with the p-value of .000. H5 is also supported. The analysis result showed that satisfaction with travel had the significance probability of .052 (ρ<.1, β=.148) when t value was 1.953 and repurchase of the tour product had significance probability of 0.000 (ρ<.01, β=.339) when t value was 5.016 indicating that both had positive impacts on the travel agency’s brand.

Discussion

Multiple regression analysis was conducted to test hypotheses and the results are presented in Table 5. Among the six selection attributes identified through PCA for Hypothesis 1, destination information, differentiation of the product, convenience and shopping, uniqueness of the place were found to have statistically significant impacts on one’s willingness to purchase a tour product. The factor with the highest influence was uniqueness of the place. The result of the Hypothesis 2 test showed that the impact of purchasing a tour product that resulted in satisfaction with the overall travel on repurchase of a tour product was significant. For Hypothesis 3, the level of satisfaction with travel had a great impact on re-purchase of the product. This result supported the significant causal relationship among purchase of a tour product, satisfaction and repurchase, as predicted in previous studies. For Hypotheses 4 and 5, it was found that post behavior had statistically significant impacts on selection of a travel agency and travel agency brand. In particular, satisfaction with travel had a great impact on selection of a travel agency and willingness to purchase another tour product again on travel agency brand. These test results present implications for the ‘travel decision-making model’ presented in previous studies and marketing travel agencies.

Implication

The result of hypotheses tests found that four factors out of six selection attributes of tour products had significant impacts on one’s willingness to purchase a tour product. In particular, uniqueness of the place showed the highest level of influence. It means that travelers considering overseas travel pursue new experiences and value in selecting tourist destinations. From the business perspective, the study results imply that travel agencies need to emphasize differentiated characteristics of the tourist destination in the process of making and planning tour product.

The causal relationship tested between one’s willingness to purchase a tour product and repurchase of that tour product supported by previous studies was found to be significant. In particular, the causal relationship between satisfaction and repurchase of a tour product was very strong. In this study, the causal relationship of the experience of purchasing a tour product, post behavior and evaluation on the travel agency had a positive impact on selection of a travel agency and travel agency brand. In particular, a high rate of repurchase and good evaluation of travel agency brand showed a strong causal relationship. Through these results it is demonstrated that one’s satisfaction with the tour experience leads not only to willingness to purchase a tour product again but also to the possibility of pursuing the same travel agency for future travels and a positive impact on the brand of the travel agency.

In conclusion, customers considering various factors in selecting tour products include not only evaluation of the overall tour experience after coming back from a tour but also evaluation of the travel agency which provided the tour product. Therefore, travel agencies should have a ‘loyal customer management system’ that considers comprehensive brand management strategies from the initial tour product development to securing loyal customers. Recently, many travel agencies are conducting various promotion activities as part of their efforts to attract new and loyal customers through websites. In addition, with the emergence of online travel agencies, the need to utilize websites and their importance to travel agencies are obvious [31]. In addition, travel agencies need to identify demographic and economic characteristics of senior groups to provide more professional and specific tour products to increase customer loyalty as the number of seniors is increasing worldwide [32]. As such more age-group specific strategies should be developed.

Limitation of the study and research needs

This study applied the evaluation of a travel agency as the primary influencing variable on causal relationships among tour products, purchase and post behavior among Japanese overseas travelers. However, this study has the following limitations. First, even though this study presented a new concept developed by this researcher (concerning evaluation of the travel agency), it did not present a structural model that could present a more specific causal relationship. In regards with development of the study model and verification of hypotheses conducted in an exploratory manner, this study showed limits on making objection to generalized results by comparing with the existing statistical significance level and then by partially supporting figures (p<.1) of a relatively positive significance level. In addition, the study did not consider demographic characteristics, lifestyles and travel patterns of respondents that could provide more specific marketing implications. These limitations will be considered in future research through a more comprehensive survey.