Applying Data Mining To Customer Segmentation Marketing Essay

ABSTRACT. The intent of this survey was to happen factors that affect clients ‘ purpose to revisit eating houses utilizing informations excavation. A sum of 390 useable questionnaires were used in the information analysis. AnswerTree, a information excavation package, was used as a major analytical method. AnswerTree enables a research worker to place and aim desirable client groups and therefore is suited to utilize in placing differences between one group with a revisit purpose and one group with no revisit purpose. Study consequences indicated that different constituent factors affected clients ‘ revisit purpose between the two groups. Three factors ( recommendation, populating country, and figure of dining occasions per hebdomad ) most affected a group ‘s revisit purpose. On the other manus, three factors ( recommendation, business, and most frequent dining finish ) most affected the group with no revisit purpose. Study consequences provide meaningful information for selling schemes that can be successfully used by full-service eating house operators.

Hire a custom writer who has experience.
It's time for you to submit amazing papers!


order now

KEYWORDS. Customer cleavage, full-service eating house, information excavation, AnswerTree

Introduction

With increasing competition in the eating house concern, selling schemes need to be more important to guarantee client satisfaction ( Kara et al. , 1997 ; Kivela, 1997 ; Murphy et al. , 1996 ) . Since client satisfaction affects concerns ‘ ability to prolong their position in a competitory eating house market, many research workers have studied client satisfaction ( Almanza et al. , 1994 ; Andaleeb & A ; Conway, 2006 ; Barsky & A ; Labagh, 1992 ; Domingo, 2002 ; James, 1995 ; Johns & A ; Tyas, 1996 ; Oh, 1999, 2000 ; Oliver, 1980, 1981 ) . Previous research workers have queried eating house clients to place the grounds for their determinations to revisit and urge eating houses to other possible visitants, thereby proposing effectual selling schemes. Although there is no confidence that clients will do a return visit ( Dube et al. , 1994 ) , restaurateurs assume that satisfied clients will return while clients who had a hapless experience will non come once more. Therefore, understanding both satisfied and unsated clients is an of import procedure in doing appropriate selling schemes. In order to section clients harmonizing to their properties, cognition of those properties is an effectual tool in developing appropriate selling schemes ( Bowen, 1998 ; Gregoire et al. , 1995 ; Klosgen & A ; Zytkow, 2002 ; Reid, 1983 ; Richard & A ; Sundaram, 1994 ; Swinyard & A ; Struman, 1986 ; Woo, 1998 ; Yuksel & A ; Yuksel, 2002 ) . Customer cleavage can be determined by client properties such as demographic and behavioural features: purchasing forms, attitude and usage or response to a merchandise ( Kotler et al. , 2005 ) . In this survey, the term client properties was defined as clients ‘ demographic and behavior forms, attitude and usage or response to a merchandise via recommendation purpose.

To place client cleavage, informations excavation was applied in this survey. Data excavation is the procedure of happening or sorting tendencies and forms in informations to analyse clients ‘ past behaviours ( Adriaans & A ; Zantinge, 1997 ; Chatfield, 1995 ; Fadairo & A ; Onyekelu-Eze, 2008 ; Kudyba & A ; Hoptroff, 2001 ; Lovell, 1983 ; Pyle, 1999 ; Thuraisingham, 1999 ) . If eating house directors clearly know their mark client properties, they can develop a better service scheme or do up for their weak points.

The purpose of this research was to place factors impacting clients ‘ purpose to revisit a eating house. Study consequences provide meaningful information to the procedure of developing selling schemes that can be successfully used by full-service eating house operators. Further, survey consequences will inform restauranters about of import and non-important factors in return purpose and enable them to choose those factors on which to concentrate on and better.

REVIEW OF THE LITERATURE

Customer Segmentation

Customer cleavage is a important portion of today ‘s excessively competitory eating house concern. In the mid-1950s, Wendell R. Smith, an American seller, foremost introduced the construct of client cleavage ( Nairn & A ; Berthon, 2003 ) . Russell Haley so developed cleavage theory in 1968. Customer cleavage can be defined as a manner to divide clients into groups for determination devising intents or to back up effectual direction in geting or maintaining clients ( Bowen, 1998 ; Bahn & A ; Granzin, 1985 ; Chen et al. , 2007 ; Yuksel & A ; Yuksel, 2002 ) . Since client cleavage can assist eating houses increase gross, a selling scheme based on client cleavage can be more powerful and effectual ( Auty, 1992 ; Bojanic & A ; Shea, 1997 ; Chen et al. , 2006 ) .

Customer cleavage has encouraged research workers to take a closer expression at client cleavage as portion of an effectual selling scheme. Lewis ( 1981 ) used discriminant analysis to place the differences between departers and non-goers with nutrient quality, bill of fare assortment, monetary value, atmosphere, and convenience. Bahan and Granzin ( 1985 ) investigated four client sections: wellness, epicure, value, and unconcerned. They reported that each group had different penchants for service quality. Auty ( 1992 ) divided respondents into three client groups ( pupil, comfortable middle-aged people, and older people ) and examined eating house image and atmosphere. Oh and Jeong ( 1996 ) revealed the features of four client sections: orderly service searcher, convenience searcher, authoritative dinner searcher, and apathetic dinner searcher. Bojanic and Shea ( 1997 ) sought differences between business district diners and suburban diners. Yuksel and Yuksel ( 2002 ) identified the properties of five client sections ( value searchers, service searchers, adventuresome nutrient searchers, atmosphere searchers, and healthy nutrient searchers ) based on nine factors ( service quality, merchandise quality and hygiene, adventuresome bill of fare, monetary value and value, atmosphere, healthy nutrient, location and visual aspect, handiness of nonsmoking country, and visibleness of nutrient readying country ) . In this survey, the research worker divided clients into two groups based on revisit purpose in order to place which factors most affect these two groups. The sum-up of the literature reappraisal is presented in Table 1.

Table – 1 about here

Data Mining

To split respondents into two groups based on revisit purpose, unlike old surveies, informations excavation was used to place revisit purpose in the eating house concern. Data excavation was introduced in the late eightiess and developed in the 1990s. Its beginning is in the Fieldss of statistics and a specialised country within unreal intelligence ( AI ) that is portion of computing machine scientific discipline ( Ogut et al. , 2008 ; Roiger & A ; Geatz, 2003 ) . Harmonizing to Fadairo and Onyekelu-Eze ( 2008 ) , informations excavation is a manner to happen information in a immense database. Such information can be used to place relationships between variables. That is, informations excavation may be utilized in analysing a specific information set with the purpose of placing forms and set uping relationships ; utilizing informations excavation, it is possible to screen through monolithic volumes of informations and detect new information or an analytic method for concluding utile cognition and foretelling future tendencies ( Bolshakova et al. , 2005 ; Chatfield, 1995 ; Chen et al. , 1998 ; Groth, 2000 ; Hand et al. , 2001 ; Koyuncugil, 2004 ; Lovell, 1983 ; Westphal & A ; Blaxton, 1998 ) . Data excavation can execute two basic operations: foretelling client behaviours and placing cleavage ( Lampe & A ; Garcia, 2004 ; Wang et al. , 2008 ) . For that ground, many research workers have attempted to use informations excavation in the concern industry ( Keating, 2008 ; Liu et al. , 2008 ; Rygielski et al. , 2002 ; Zambochova , 2008 ) . Previous surveies have stressed that companies could utilize informations excavation to place client temperaments or tendencies sing the patronizing of certain companies. With this information, companies can concentrate their attempts on good clients from whom they would do the most net income. Further, all industries can take advantage of informations excavation in seeking to understand inconsistent cleavage of their mark clients. In drumhead, informations excavation is a powerful engineering that may be used in support of companies prosecuting in decision-making on issues such as client abrasion, client keeping, client cleavage and gross revenues prognosis ( Ogut et al. , 2008 ) .

CHAID in AnswerTree

To use informations excavation, the AnswerTree plan was used in this survey. AnswerTree, a information excavation package, is a foreseeable theoretical account that shows consequences in a tree theoretical account ( SPSS, 2009 ) . Variables may be analyzed in the AnswerTree plan in three ways: CHAID ( chi-squared automatic interaction sensor ) , CART ( categorization and arrested development trees ) , and QUEST ( speedy, indifferent, efficient, statistical tree ) . Basically, the CHAID method is a more comprehensive method and generates more accurate consequences when utilizing categorical variables, while CART and QUEST are suited when utilizing uninterrupted variables. Since categorical variables were used in this survey, the CHAID method was applied here.

The original CHAID method grew from a 1975 doctorial thesis by Kass, who published a more accessible article four old ages subsequently ( Kass, 1980 ) . Since the CHAID method allows sellers to place sections in relation to a dependant variable holding two or more classs based on the combination of independent variables ( Chen, 2003 ) , the CHAID method has popularly been applied in the consumer research field ( Haughton & A ; Oulabi, 1997 ; Levin & A ; Zahavi, 2001 ) . In the CHAID process, a dependent variable and cardinal independent variables are ab initio chosen. Harmonizing to chi-squared, the dependent variable can be divided by the degrees of a certain independent variable that has the strongest association with the dependant variable. That is, the most of import and related independent variable with a dependent variable becomes the first node. This analysis procedure occurs when one of three standards are met, harmonizing to Berson et Al. ( 2000 ) :

1. The section contains merely one record. ( There is no other inquiry that you can inquire to farther polish a section of merely one. )

2. All the records in the section have indistinguishable features. ( There is no ground to go on inquiring farther inquiries because all the staying records are the same. )

3. The betterment is non significant plenty to justify inquiring the inquiry ( p. 162 ) .

All variables used in this survey were categorical measurings with two or more categorical degrees. The fillet regulations for AnswerTree analyses were a maximal tree deepness of 3, minimal figure of instances of 25 for a given node, and significance degree for splitting of 0.05.

RESEARCH METHODOLOGY

Data Collection and Questionnaire

The informations used for this survey were collected in Miami via face-to-face interviews. The response rate for the face-to-face interview has revealed it to be the best method among assorted study methods ( The Monkey Team, 2008 ) . Surveies were administered from May 1 to May 31, 2007. To increase consequence dependability, we selected respondents who had visited a full-service eating house within the last one month. The selected full-service eating houses offered full table service and the mean guest outgo was at least $ 25 per individual. Of the 414 questionnaires collected, 24 were uncomplete and were eliminated. As a consequence, a sum of 390 questionnaires were used in the information analysis.

Since AnswerTree enables a research worker to place and aim desirable client groups ( SPSS, 2009 ) , it is a suited analysis method for placing differences between groups. Further, AnswerTree is a more robust method than bing statistical methods in placing section features ( Byrd & A ; Gustke, 2006 ) . There were two groups in the dependent variable: one group with the purpose to revisit and one group with no purpose to revisit. Questionnaire points for revisit purpose were rated on seven-point Likert graduated table runing from “ strongly differ ” to “ strongly hold. ” To use CHAID ( Chi-square Automatic Interaction Detection ) analysis, research workers converted the seven-point graduated table into categorical variable ( agree, indifferent, disagree ) . Even though respondents ‘ replies were a five or six on the seven-point graduated table, the information sufficiently ensured positive responses. Finally, those with five, six and seven points were converted into a positive group ( agree ) . On the other manus, those with one, two and three points were converted into a negative group ( disagree ) . Finally, four on the seven-point graduated table was converted into a “ indifferent ” group.

Among respondents ( n=390 ) , 83.33 % ( n=325 ) indicated that they were willing to revisit the eating house ; on the other manus, 11.54 % ( n=45 ) thought that they would non revisit the eating house ; and 5.13 % ( n=20 ) replied “ indifferent ” , which means “ I do n’t cognize. ” Customer cleavage can be sorted by demographic and behavioural features such as purchasing forms, attitude and usage or response to a merchandise ( Johns & A ; Pine, 2002 ; Kotler et al. , 2005 ) . Independent variables were composed of demographic profiles ( gender, age, matrimonial position, business, income, populating country ) and client properties in relation to purchasing behaviour, a form of full-service eating house usage ( how many dining occasions per hebdomad, how much spent at a eating house, frequence of eating house visits, with whom did they dine at that place ) . Finally, to determine a response to a merchandise, recommendation purpose was used in this survey as an independent variable. The sum-up of informations description is presented in Table 2.

Table – 2 about here

Consequence

Figure 1 shows the general theoretical account for revisit purpose. The general theoretical account makes it easy for readers to calculate out how many nodes are included in the consequences and shows the full theoretical account design. In this survey, 13 nodes were used to explicate the factors impacting a group who intended to revisit and a group with no purpose to revisit. For the dependant variable, originally there were three groups: “ yes, ” “ indifferent, ” and “ no ” . However, because of little sample size, the “ indifferent ” group was non divided by form. The dependent variable was divided by five forms: “ urge purpose, ” “ living country, ” “ how many times to dine per hebdomad, ” “ business, ” and “ when frequently dine. ”

Figure – 1 about here

In footings of “ yes ” cleavage on revisit purpose, in Figure 2, the first split was “ recommend: yes ” ( I‡2=376.4356, d.f.=4 ; p=.000 ) . In Node 3, 98.10 % ( n=309 ) of respondents thought that they had a revisit purpose. Node 3 was divided into two groups: Node 7 and Node 8. The 2nd split was based on the variable of “ Living country: South Florida, other Florida, and other U.S provinces ” ( I‡2=68.5075, d.f.=2 ; p=.000 ) . Node 7 was divided into two groups: Node 11 and Node 12. 98.69 % ( n=302 ) of respondents ( Node 7 ) who lived in “ South Florida, other Florida, and other U.S provinces ” showed that they were willing to return to the eating house. The last split was “ how many times dine per hebdomad: over 3 times ” ( I‡2=19.7972, d.f.=1 ; p=.000 ) . In Node 12, 100.00 % ( n=254 ) of respondents who visited the eating house over 3 times per hebdomad indicated that they had a revisit purpose. In drumhead, there were three forms as follows: “ recommend: yes, ” “ life: South Florida, other Florida, and other U.S provinces, ” and “ how many times to dine per hebdomad: over 3 times. ”

Figure – 2 about here

Sing “ no ” cleavage on revisit purpose, in Figure 3, the first split was “ recommend ( no ) ” ( I‡2=376.4356, d.f.=4 ; p=.000 ) . In Node 1, 81.13 % ( n=43 ) of respondents indicated that they would non see the eating house once more. Node 1 so was divided into three groups: Node 4, Node 5, and Node 6. The 2nd pruning tree was based on the “ business ” variable: office worker ( professional, salesman, and freelance ) ” ( I‡2=20.1046, d.f.=4 ; p=.000 ) . In Node 4, 94.87 % agreed that they were non willing to return to a eating house. Node 4 was divided into two groups: Node 9 and Node 10. The last split was “ when frequently dine: tiffin ” ( I‡2=25.2973, d.f.=1 ; p=.000 ) . In Node 9, all office workers ( 100.00 % , n=36 ) who frequently visited the eating house at lunch period did non hold a recommendation every bit good as revisit purpose. In brief, three forms split the node: “ recommend: no, ” “ business: office worker ( professional, salesman, and freelance ) , ” and “ when frequently dine: tiffin. ”

Figure – 3 about here

Figure 4 nowadayss drumhead statistics. The saloon graph makes it easy for readers to understand which node most represents the dependant variable and provides the fluctuation for each dependant variable. In this survey, the dependant variable, revisit purpose, was classified by a maximal tree deepness of 3, minimal figure of instances of 25 for a given node, and significance degree for splitting of 0.05. The saloon graph for AnswerTree showed that the peculiar nodes most frequently represent groups ‘ purpose or non-intention to revisit, severally ( Node 12: “ hold ” group, Node 9: “ disagree ” group ) .

Figure – 4 about here

Table 3 nowadayss a gain chart of “ yes ” sections. A addition chart is a tabular array that summarizes the full theoretical account descriptively. In the addition chart, we can see the per centum representation of each node for the dependant variable. In the instance of the “ yes ” section, the root node was 83.33 % ( n=325 ) . Node 12 was computed by taking 100.00 % ( Derive: % computed from Node: N divided by Resp: N ) and so spliting it by 83.33 % ( root node ) . The consequence was 120.00 % , the index mark for Node 12. That is, Node 12 ( recommend: yes, living country: south Florida, other Florida, and other U.S provinces, how many times dine per hebdomad: over 3 times ) represents a root node about 1.2 times. Therefore, in the instance of “ yes ” cleavage, three variables ( urge, life, and how many times to dine per hebdomad ) are of import factors in spliting respondents into groups that answered “ yes ” sing revisit purpose.

Table – 3 about here

Table 4 nowadayss a addition chart for “ no ” section. The root node was 11.54 % ( n=45 ) .

Node 9 was computed by taking 100.00 % ( Derive: % ) and so spliting it by 11.54 % ( root node ) . The consequence was 866.66 % , the index mark for Node 9. That is, Node 9 ( recommend: no, business: professional, salesman, and self-employed, when frequently dine: tiffin ) represents a root node about 8.6 times. Therefore, in the instance of “ no ” section, three variables ( urge, business, when frequently dine ) are of import factors in spliting a group that answered “ no ” with regard to revisit purpose.

Table – 4 about here

Table 5 offers a hazard chart bespeaking the clearcutness of the categorization. It resembles the per centum of classified respondents in the discriminant analysis. The hazard estimation predicted the hazard incurred due to misclassification of the respondents in the AnswerTree analysis. A lower hazard estimation indicates a more exactly classified theoretical account. Harmonizing to the consequences of the appraisal of revisit purpose, the hazard estimation was 0.0615385. This means that the preciseness of sorting respondents in the AnswerTree analysis was 99.9384615 % . That is, approximately 99.93 % of the respondents were classified accurately on split nodes.

Table – 5 about here

DISCUSSION AND IMPLICATION

In the eating house concern, client cleavage enables eating house directors or sellers to develop effectual selling schemes. The intent of this survey was to place factors that affect purpose to revisit a full-service eating house. To determine differences between group purpose to revisit and group non-intent to revisit, informations excavation was used. There has been small usage of informations excavation in the cordial reception field. Because information excavation is one manner to make decision-making theoretical accounts that predict future behaviour based on analyses of past activity, utilizing collected informations from section targeting is the best manner to make suited selling schemes ( Lampe & A ; Garcia, 2004 ; Wang et al. , 2008 ) . Among the respondents ( n=390 ) , 83.33 % ( n=325 ) indicated that they were willing to see the eating house once more while 11.54 % ( n=45 ) thought that they would non see the eating house once more and 5.13 % ( n=20 ) were “ indifferent. ” As mentioned earlier, due to the little sample size, the “ indifferent ” group was non divided.

The AnswerTree consequences revealed different constituent factors between the two groups. In the instance of the “ yes ” section, there were three forms: “ recommend: yes ” ( I‡2=376.4356, d.f.=4 ; p=.000 ) , “ living country: south Florida, other Florida, and other U.S provinces ” ( I‡2=68.5075, d.f.=2 ; p=.000 ) , and “ how many times dine per hebdomad: over 3 times ” ( I‡2=19.7972, d.f.=1 ; p=.000 ) . Analysis consequences revealed that the more clients like to dine out, the more they intend to revisit a eating house. In other words, people who have a great involvement in dining are more likely to be loyal clients. Furthermore, as they intend to urge the eating house to other people, maintaining such a client means other possible clients could be positively affected. Harmonizing to Soderlund ( 1998 ) , “ Word-of-mouth is defined here as the extent to which a client informs friends, relations and co-workers about an event that has created a certain degree of satisfaction ” ( p. 172 ) . As word-of-mouth can be a important determiner of behavioural purposes, recommendation purpose greatly affects a eating house concern ‘s gross revenues ( Babin et al. , 2005 ; Edwards & A ; Meisleman, 2005 ; Mangold et al. , 1999 ; Mattila, 2001 ; Soderlund, 1998 ) . This survey ‘s consequences supported the determination that loyal clients are more likely to promote other people to besides hold their exceeding experience, which is consistent with old surveies ( e.g. , Bowen, 1998 ) . It is less dearly-won to maintain an bing loyal client than to pull a new client. Besides, loyal clients return to do more repetition purchases at the eating house. Therefore, placing loyal clients is an of import portion of the eating house concerns ( Bowen & A ; Shoemaker, 1998 ; Fierman, 1994 ; Jang & A ; Mattila 2005 ; O’Brien & A ; Jones, 1995 ; Orr, 1995 ; Schneider et al. , 1998 ) . From a managerial point of view, the consequences of this survey based on informations mining aid restaurateurs in placing the features of a loyal client section. In the instance of the “ no ” section, three forms split the node: “ recommend: no ” ( I‡ 2=376.4356, d.f.=4 ; p=.000 ) , “ business: office worker ( professional, salesman, and freelance ) ” ( I‡2=20.1046, d.f.=4 ; p=.000 ) , and “ when frequently dine: tiffin ” ( I‡2=25.2973, d.f.=1 ; p=.000 ) . The consequences of this analysis revealed that office workers who frequently visited a eating house at tiffin did non mean to revisit. In other words, office workers, salesmen, and the freelance were less satisfied with tiffin at the full-service eating house. In general, few full-service eating houses concentrate on tiffin for office workers, who normally eat lunch off from the office. Even though some tiffin bill of fares do concentrate on the office worker, offering this repast is normally expensive and time-consuming. However, as office workers tend to eat tiffin in a brief span of clip, they prefer to eat tiffin at fast nutrient eating houses instead than at a full-service eating house. From a managerial point of view, if full-service restauranters could supply a tiffin bill of fare with low monetary values every bit good as speedy service, they could obtain portion of the office worker market presently traveling to fast nutrient eating houses. The sum-up of sections is presented in Table 6.

Insert Table 6

The intent of client cleavage is to aim a certain type of client when developing a selling scheme. If a eating house can non develop correct and appropriate selling schemes, they may non prolong their being in this extremely competitory concern. As properties have changed and become more complicated, client cleavage is going more of import in supplying basic beginning stuff for selling schemes. In order to react rapidly to altering client properties, eating house sellers require rapid entree to information on assorted client properties. In this context, eating house sellers need to be able to place clients ‘ past behaviours in order to foretell future inclinations. This ability can be provided and maintained in the eating house concern by utilizing informations excavation engineerings. Data excavation enables restaurant sellers to pull information more efficaciously from databases ( Bolshakova et al. , 2005 ; Chatfield, 1995 ; Chen et al. , 1998 ; Groth, 2000 ; Hand et al. , 2001 ; Koyuncugil, 2004 ; Lovell, 1983 ; Westphal & A ; Blaxton, 1998 ) . Through the effectual usage of informations excavation, directors can more rapidly analyse client property in the eating house concern.

This survey of client revisit purpose occurred in the Miami country merely. Therefore, findings might non be generalized to other countries. Another restriction is that survey consequences can non be applied to all eating house services, because our focal point was on full-service eating houses merely. Therefore, findings must be applied to other eating houses with due cautiousness. Last, this survey did non utilize a big adequate sample size for informations excavation. Although we used 390 samples to place and aim client groups, informations excavation is typically used with a big database. Therefore, future research may utilize informations excavation with a larger sample size. Data excavation was a utile method in foretelling eating house clients ‘ purposes to revisit. Unfortunately, really few surveies have used this method. Thus, farther research in other cordial reception Fieldss would profit from informations excavation.

x

Hi!
I'm Heather

Would you like to get such a paper? How about receiving a customized one?

Check it out