SEGMENTATION OF TRADITIONAL MARKET SHOPPERS BASED ON STORE IMAGE IN BANDUNG, INDONESIA

Traditional market is a place for buying and selling, also part of micro, small medium enterprises (MSMEs) that play an important role in the regional economy. Besides having economic value, the traditional market in Bandung City is also one of the hereditary cultures that must remain preserved. Unfortunately, the condition of most traditional markets in Bandung has been deemed unfeasible. The image that the markets are dirty, smelly and other bad things have attached to most of the traditional markets in the city. To minimize the bad sentiments, the government is carrying out market revitalization to change the market to be in accordance with the wishes of the customers. In other words, this revitalization goal is to make the market image in accordance with the wishes of the customer. Segmentation needs to be done to be able to know the profile of each traditional market customer group. Thus, the policy makers could determine the appropriate target market. This research used quantitative method with distributing questionnaires toward 400 respondents. The data then analyzed using two-step cluster analysis and the help of software SPSS 20.0. The results show that there are 2 dominant segments among all respondents. The first segment is apathetic shoppers, those who rarely visit the market and have a more negative attitude towards traditional markets. And the second segment is enthusiast shopper, those who visit traditional markets more often and give more positive attitude towards traditional markets.


Introduction
In this era, the development of modern retailers both owned by domestic and foreign-owned companies are increasing in Indonesia. The development of modern retail has increasingly eroded the existence of traditional markets in Indonesia. In 2013, the survey institute of AC Nielsen, showed a significant continuous decline from 2007 to 2011. In 2007, the number of traditional market is 13,550, while in 2009, it shrank to 13,450, and in 2011 it was 9,950. When compared to the modern market, traditional markets shrank 8.1% and modern markets grew 31.4% (Maharani, 2014).
The decline number of traditional markets is happened continuously. The situation is detrimental because traditional markets are one main medium for Micro, Small and Medium Enterprises (MSMEs) (Sumrahadi et al., 2020). The MSMEs have important role in driving the growth of the Indonesian economy. The presence of the MSMEs make unemployment in this country is reduced. The sector has also proven to be a formidable economic pillar. The Minister of Finance in 2015 stated that during the economic crisis in 1998, the MSMEs sector was still holding out (Ma'ruf, 2005). The traditional retail market share tends to decrease, along with the increasing number and capitalization of modern retail businesses (Haryotejo, 2014). Buyers prefer to shop at modern retail compared to traditional markets because the image of modern retail is better, cleaner, more comfortable etc.
Poor traditional markets must receive serious handling from the government. The government must encourage traditional markets to revise their services like modern retail, thus, that they are not eliminated. Improving traditional markets to become a comfortable, attractive and positive shopping place is a formidable challenge that must be sought by the government as a sense responsibility to the public (Malano, 2011). One of the programs being run by the government in responding to this is revitalizing traditional markets. This program is carried out to create a better market image by eliminating the impression of slums and dirty traditional markets in the eyes of customers. This revitalization is not only aimed at fixing the physical condition but will also rearrange the market structure to fix the market management system (Juliarta & Darsana, 2015). Traditional market managers have to determine their target market. Before determining the target market, the thing to do is segmentation. Segmentation is done by identifying and compiling profiles of customer groups who have different needs and desires (Kotler & Keller, 2016).
As previously discussed, it is the bad market image that makes market customers switch to modern retail. Therefore, market revitalization is carried out to change the market to be in accordance with customer wishes. In other words, this revitalization wants to make the market image in accordance with customer desires. To be right on target, segmentation must be carried out so that managers can determine the target market of the traditional market. Different customer groups generally have different views about the importance of a store attribute compared to other store attributes. Ideally, a store will emphasize the priority of store attributes in accordance with the order of the target group of customers that are considered important (Sezhiyan et al., 2010).

Literature Review
"Marketing is about identifying and meeting human and social needs". The statement explains that marketing has a goal to be able to identify and meet human and social needs. Marketing must identify in advance to find out the business opportunities that exist in meeting social and human needs (Kotler & Keller, 2016). Market segmentation divides a market into clearly defined parts. A market segment consists of groups of buyers who have similar needs and desires.
The task of marketers is to identify the appropriate number and natural market segments and determine which segments will be targeted (Kotler & Keller, 2016). Market segmentation as a process of dividing consumers or potential customers in a market into different groups or segments. The essence of segmentation is to group consumers who have similarities or similarities into a group and separate them from other consumers who have different characteristics (McDonald & Dunbar, 2012).
The American Marketing Association defines consumer behavior as "The dynamic action of affect and cognition, behavior, and the environment by which human beings conduct the exchange aspects of their lives." According to the AMA, consumer behavior is a dynamic interaction of influence and awareness, behavior, and the environment in which living things exchange various aspects in their lives. From this definition, we can conclude that consumer behavior changes rapidly due to the influence of various things in their lives that can affect their thoughts and decisions in interacting and carrying out a series of consumption processes as consumers (Peter & Olson, 2019).
The concept of store image was first introduced in 1958 by Martineu. Then the concept was developed by subsequent researchers and focused on the store image dimension (Shamer, R, 2016). Image depicts the views of consumers and other parties to a retailer. Citra stop the position of retail in the eyes of consumers when compared with competitors. Therefore, retailers must be able to manage their business well to get a good image in society (Berman & Evans, 2010).
Different consumer groups will have different preferences and opinions about the level of importance of each attribute in the store image, so that no store can reach everyone. In carrying out its marketing strategy, a store might develop several attributes of its store in accordance with the target market. So in developing the store's store image attributes, the marketing strategy builder must be able to know the store attribute preferences that are important for the target market segment (Chandrasekar et al., 2013). a. Merchandise A series of activities, ranging from obtaining certain goods and or services to making these goods and or services available in place, time, price and quantity that allows retailers to achieve their goals. (Berman & Evans, 2010). Merchandise includes price, quality, diversity of categories, item availability (color, size, type) (Sopiah & Syihabudhin, 2008). Merchandise can be measured using indicators of quality, diversity and full shelf display of goods (Watanabe et al., 2013).

b. Facilities
The existence of public facilities such as toilets and places of worship are part of the attributes of traditional market store images (Prabowo & Rahadi, 2015). Facilities include gift wrappings, food courts, toilets and parking, home service (Sopiah & Syihabudhin, 2008).

c. Service
Service is an in-store retail service that aims to facilitate shoppers when shopping at the store (Ma'ruf, 2005). Some services that can be an indicator of store image in a retail store are choice of how to pay, the availability of food corner, home delivery services for certain products (such as refrigerators) and catalogs (Sopiah & Syihabudhin, 2008).

d. Store Atmosphere
The atmosphere is one of the components forming a store image which is a psychological feeling that consumers feel when they visit a retail store (Berman & Evans, 2010). Store atmosphere is measured using temperature indicators, air clean system, smell, cleanliness (Watanabe et al., 2013). Cleanliness of the store must be considered both in terms of space, shelves where goods, merchandise and air cleanliness (Sunyoto, 2012). e. Personnel Personnel are people whose job is to assist customers in shopping. Behavior in serving (friendly, polite, alert, efficient), knowledge of the product and an adequate number of workers as indicators of personnel in the store image (Sopiah & Syihabudhin, 2008). The quality of the sales force is measured using indicators of friendliness and courtesy as well as knowledge of the goods sold and is very helpful. Besides that, it can also measure the attributes of personnel with indicators of appearance, care, friendliness, knowledge of goods sold and also the amount sufficient to serve customers (Watanabe et al., 2013).
f. Price One of the keys to success in doing retail business is to provide superior value for certain price orientations chosen by consumers (Berman & Evans, 2010). The price of goods that are too expensive and not affordable by consumers can damage the store's image (Sopiah & Syihabudhin, 2008). Price can be measured using indicators Competitive, prices, Good prices, Price/ quality (Charity, 2017).

g. Convenience
Convenience is the ability that can reduce non-monetary costs (such as time, energy and effort) consumers in making purchases or using goods and services (Fazlzadeh et al., 2012) h. Location Location is a crucial factor in starting and developing a retail business. In choosing a location, retailers should pay attention to several things, including visible, heavy traffic, directions to go home, close to other public facilities, acquisition costs, permits and access (Gusway F, 2009). Customers want locations that are easily accessible, safe and located in shopping centers or close to retail stores (Sopiah & Syihabudhin, 2008).

Research Methods
This study used quantitative method. The Bernouli formula with an unknown population and a 95% confidence level. Thus, it was found that 385 samples were rounded up to 400 people.
The sample will fill the questionnaire using 5 Likert scale ranges. From the data obtained, SPSS 20.0 software and Microsoft Excel 2010 are used to process the data.
The analysis used in this study is cluster analysis. Cluster analysis is an analysis with a multivariate approach to identify objects or individuals who have similarities to each other in several ways (Alamanda & Ramdhani, 2018). Cluster analysis classifies the individuals or objects into a small number of complete and independent groups. Objects or individuals will be placed in groups so that there will be many similarities within a group and little similarity compared to other groups. Clusters must have high homogeneity within their clusters and high heterogeneity with other clusters.
This study uses two-step clustering which combines Hierarchical and Partitioning cluster analysis approaches. In the first stage, the algorithm used in the procedure is like the partition method especially k-means. And from the results of the first process will undergo a hierarchical procedure. This technique can process categorical and continuous variables simultaneously and can determine the number of clusters and the maximum number of clusters flexibly according to statistical evaluation criteria. This method was developed to deal with problems in analyzing variables that have different levels of scale (Sarstedt & Mooi, 2014).

Result and Discussion
Colinearity test using Bivariate correlation is used in this paper to test whether there are indicator variables that have a correlation number of more than 0.90. Based on the results of the collinearity, there is no correlation was found between the indicator variables that were more than 0.90. The highest correlation of 0.827 is between the indicators of cleanliness and odor that does not interfere in the store atmosphere attribute. While the lowest correlation with the number -0.033 is between the indicator of the variety of goods in the merchandise variable with the debt indicator in the service attribute. Then it can be concluded that all indicator variables used will not cause problems in this segmentation process and are feasible to use.
Then a two-step cluster analysis is carried out and produces two clusters. the first cluster is 49.8% or 199 respondents and the second cluster is 50.3% or 201 respondents. The ratio of the second cluster (the largest) compared to the first cluster (the smallest) produces a number 1.01 which indicates that the size of the first and second cluster is not too far away. This analysis also shows 10 indicators that are considered the most important in the castering process. Namely the availability of adequate toilets, information from traders, availability of adequate places of worship, cleanliness, the absence of disturbing odors, availability of adequate parking spaces, quality of goods offered, distance between stalls that are free, price conformity with quality and response from traders.
In order to test whether the result of the cluster partition distinguishes data well, cluster centroids must be tested. This stage is very important because it shows whether the clusters formed are different from each other or not. therefore, an independent sample T test and ANOVA were performed to compare the average of the variables between clusters (Sarstedt & Mooi, 2014) This writing produces two clusters so that the writer will do the Independent sample T test. From the test results, the writer will compare the significance number with the p-value with decision criteria: If the p-value is α, then H0 is rejected If p-value> α, then H0 is accepted Because this writing uses a two-tailed test, the value of α will be divided by 2. The hypothesis used is: H0: There is no difference based on the store image between the first cluster and the second cluster. H1: there is a difference based on the store image between the first cluster and the second cluster Twenty-four of twenty-five indicators have a significance value of 0.000 which is smaller than the α / 2 value of 0.025. therefore sig. (0.000) <0.025 then H0 is rejected and H1 is accepted. So it can be concluded that there are differences based on these indicators between the first cluster and the second cluster. While the indicator "close to other crowds" of the location attribute which is the last indicator number 25 has a value of 0.065 which is greater than the α/ 2 value of 0.025, so sig (0.065)> 0.025 then H0 is accepted and H1 is rejected. Therefore, it can be concluded that there is no difference based on these indicators between the first cluster and the second cluster.
The first cluster was named apathetic Shoppers consisting of 49.8% of the total respondents, namely 199 people. Although this cluster is dominated by female respondents, it also represents most respondents with male gender. The age of this cluster member is younger, and the level of education is higher when compared to the second cluster. Most of the cluster members work as students or college students.
They visit the traditional market less often and spend less money on each visit than the second cluster. The purpose of their shopping is for daily needs (Prabowo & Rahadi, 2015) The result of their shopping at traditional markets is also consumed by fewer people than the second cluster. The first cluster consists of people who are less interested in traditional markets and tend to be more negative than the second cluster. However, more than a few of them still gave positive statements on several store image attributes such as price, merchandise and location. For other attributes, namely personnel, convenience, service, facilities, and store atmosphere, although less than half, there are still members of this cluster who give positive values.
The second cluster was named enthusiasts Shoppers consisting of 50.3% of the total respondents, namely 201 people. This cluster is dominated by respondents with female sex with an older age than the first cluster. Based on the level of education, respondents in this cluster have a lower level of education. The work of most cluster members is as a housewife and entrepreneur (Prabowo & Rahadi, 2015). They more often visit traditional markets and spend more money each visit. Their purpose in visiting traditional markets is not only to shop for daily necessities, but also to shop for goods to be resold in the same form and goods to be processed and resold (Alamanda et al., 2019). These cluster members shop for more people than the first cluster.

Conclusion
Overall, members of the second cluster appear to be more interested and respond more positively to all statements in the questionnaire than the first cluster. more than most of them give a positive value to the attributes of price, merchandise, personnel, convenience, location and facilities. There are only two attributes that get a positive response from less than half of the cluster members namely service and store atmosphere. That means only the two attributes that are the rarest reasons for the second cluster to visit traditional markets.

Recommendations
The government needs pay more attention to the management of traditional markets, both in physical and management terms. It is intended that people who rarely visit traditional markets and tend to be apathetic as the first clerks feel are interested in visiting traditional markets more often. Besides maintaining loyal visitors such as the second cluster also to be able to maximize the profits that can be received by traditional market traders.
After realizing the many shortcomings in this study, the researcher wanted to provide suggestions for further research. Namely: a. Conduct further research, namely "targeting" so that the results of this writing can provide practical benefits directly or indirectly; b. Conduct similar research but use different data analysis and sampling techniques; c. Do a factor analysis first before segmenting.