EVALUATING THE SERVICE QUALITY OF THIRD- PARTY LOGISTICS SERVICE PROVIDERS USING THE ANALYTIC HIERARCHY PROCESS

Revista de Gestão da Tecologia e Sistemas de Iformação Joural of Iformatio Systems ad Techology Maagemet Vol. 3, No. 3, 2006, p. 261-270 ISSN olie: 1807-1775 EVALUATING THE SERVICE QUALITY OF THIRD- PARTY LOGISTICS SERVICE PROVIDERS USING THE ANALYTIC HIERARCHY PROCESS Soo-hoo So Korea Callceter Idustry Research Ceter, Choam Natioal Uiversity, Korea JaeJo Kim KiJu Cheog Geo Cho BK21 Biz Covergece Team, Maagemet Research Istitute, Choam Natioal Uiversity, Korea ABSTRACT I this study we apply the aalytic hierarchy process (AHP) to evaluate the service quality of third-party logistics (3PL) service providers. We first coceptualize five dimesios of 3PL service quality (i.e. tagibles, reliability, resposiveess, assurace ad empathy). We the apply the AHP method to determie the relative weights of the five service quality dimesios ad evetually select the best 3PL service provider. To implemet this idea i practice, we coduct a empirical case study o four compaies providig 3PL services i Korea. The results idicate that Resposiveess out of the five service quality dimesios is the most importat factor i the perceptio of 3PL customers. Keywords: Third-Party Logistics, Aalytic Hierarchy Process, Service Quality, Iteret Shoppig Mall, Logistics Outsourcig Recebido em/mauscript first received: 27/07/2006 Aprovado em/mauscript accepted: 06/09/2006 Soo-hoo So is curretly workig as a vice-director of Korea Call Ceter Idustry Research Istitute. He received his Ph.D. degree i Maagemet Iformatio Systems, specialized i Supply Chai Maagemet, from Choam Natioal Uiversity, Korea. E-mail: swso@choam.ac.kr JaeJo Kim is a Professor of Iformatio Systems at the School of Busiess Admiistratio, Choam Natioal Uiversity, Korea. He received his Ph.D. degree i Computer Iformatio Systems from Arizoa State Uiversity i the U.S.A. E-mail: jaejo@choam.ac.kr KiJu Cheog is a Professor of School of Busiess Admiistratio, Choam Natioal Uiversity, Korea. He received his Ph.D. degree i Marketig from Alabama Uiversity-Tuscaloosa. He taught at Purdue Uiversity for 4 years. E- mail: kcheog3@choam.ac.kr Geo Cho is a associate professor of School of Busiess Admiistratio, Choam Natioal Uiversity, Korea. He received his Ph.D. degree i Operatios Research from Purdue Uiversity. E-mail: gcho@choam.ac.kr ISSN olie: 1807-1775 Publicado por/published by: TECSI FEA USP 2006

262 So, SH., Kim, JJ., Cheog, KJ., & Cho, G. 1. Itroductio The cocept of third-party logistics (3PL) has become a growig reality i Korea i recet years, allowig shippers to outsource logistical activities that had previously had to be coducted i house. I fact, over 60 percet of Fortue 500 firms have at least oe cotract with a 3PL service provider (Outsourcig Logistics Report 2006). As a cosequece of the rapid growth of 3PL service applicatios ad the abudace of service providers, the customer (i.e. the shipper ) has bee faced with icreasig difficulty whe tryig to select a appropriate service provider. I such scearios, service quality becomes a bechmark to differetiate services ad providers. Therefore, i order to deliver good-quality service ad gai customer loyalty, 3PL service providers eed to uderstad how customers perceive ad evaluate service quality. I this paper, we focus o evaluatig the quality of service delivered to customers by 3PL service providers. For this, we first coceptualize 3PL service quality as a secod order costruct, with five dimesios: tagibles, reliability, resposiveess, assurace ad empathy. We the apply the aalytic hierarchy process (AHP) method to determie the relative importace of five service quality dimesios, ad evetually prioritize 3PL service providers i the order of the overall service quality scores. To implemet this idea i practice, we coduct a empirical case study o four compaies providig 3PL services i Korea. This paper is orgaized as follows. First, we preset the basic cocept of 3PL ad the model we used to measure its service quality. The, we briefly outlied the AHP method adopted i this study, ad the applied it to a practical case study to evaluate service quality of 3PL service providers. Fially, we have cocluded with a discussio of the implicatios for maagemet of our fidigs ad thoughts o future research issues. 2. Evaluatio Criteria for 3PL Service Quality The outsourcig of logistics fuctios to 3PL service providers has ow become a commo practice i may idustries. 3PL refers to the use of subcotracted specialized logistics compaies to perform logistics fuctios that ca ecompass the etire logistics process or selected activities withi that process ad that have traditioally bee performed withi a orgaizatio (Lieb et al., 1993). Accordig to a survey o 3PL users (Lagley et al., 2005), the activities most frequetly outsourced to 3PL service providers are outboud trasportatio(north America, 78%; Wester Europe, 88%, Asia-Pacific, 96%; ad Lati America, 84%) ad warehousig(north America, 63%; Wester Europe, 72%, Asia-Pacific, 88%; ad Lati America, 55%), followed by customs clearace/brokerage, iboud trasportatio, freight forwardig, freight bill auditig/paymet, cross-dockig/shipmet cosolidatio, ad order dispatch/distributio. Most compaies cite greater flexibility, operatioal efficiecy, improved customer service levels, ehaced supply chai performace, ad better focus o their core busiesses as part of the advatages of egagig the services of 3PL service providers (Sahay ad Moha, 2006). Revista de Gestão da Tecologia e Sistemas de Iformação/Joural of Iformatio Systems ad Techology Maagemet

Evaluatig the service quality of third- party logistics service providers usig the aalytic hierarchy process 263 As the competitio i the 3PL market has itesified, service quality has become a importat differetiator amog service providers. The cocept of service quality goes beyod the techical aspects of providig the service. It icludes customers perceptio of what the service should be ad how the service is to be coveyed (Tsaur et al., 2002). Therefore, 3PL service providers should uderstad how customers perceive ad evaluate service quality, because service quality is related to customer satisfactio, which i tur iflueces the performace of their orgaizatios. Util recetly, the SERVQUAL istrumet suggested by Parasurama, Zeithaml ad Berry (1988, 1991) has bee the most widely used to measure customers perceptio of service quality ot oly i academic research but also i practice. Accordig to SERVQUAL, perceived service quality is measured alog five geeric dimesios such as tagibles, reliability, resposiveess, assurace, ad empathy. However, despite the wide spread of SERVQUAL, empirical research that uses this scale i the 3PL services cotext is very scarce. Therefore, we have attempted to use the five dimesioal structure of SERVQUAL to measure the quality of service provided by 3PL service providers. The five service quality dimesios are: Tagibles: the physical facilities, appearace of persoel, tools or equipmet used to provide service Reliability: the ability to perform the promised service depedably ad accurately Resposiveess: the willigess to help customers ad provide prompt service, respods immediately to customer request ad site problems Assurace: the skill, kowledge ad courtesy of service providers ad the level of cofidece that they covey to customers Empathy: the care ad persoalized attetio the firm provides for its customers. 3. Aalytic Hierarchy Process: A Multiple Criteria Decisio Makig Approach The AHP developed by Saaty (1980) has bee widely used i multi-criteria decisiomakig situatios ad has bee applied by a umber of researchers ad practitioers. Some of its applicatios iclude trasportatio problems, corporate plaig problems, ad problems cocered with marketig strategies, budget allocatio, project selectio, ad so o. The AHP methodology compares criteria, or alteratives with respect to each criterio, i a atural, pairwise mode to determie the priority weights. I additio, it makes it possible to select the sigle alterative that best meets a particular decisio criterio. I this paper, we utilize the AHP method to solve a 3PL service provider selectio problem because it is well suited to our research purpose of determiig the relative importace of the five service quality dimesios ad choosig the best 3PL service provider. The use of the AHP method, whe used to determie how customers evaluate 3PL service providers service quality ad make a selectio, eables the maagers to uderstad more clearly which service quality factors are more importat, thus providig isights that ca help with strategy formulatio ad the delivery of ehaced performace. Vol.3, No.3, 2006, p. 261-270

264 So, SH., Kim, JJ., Cheog, KJ., & Cho, G. I the AHP method, there are three basic priciples, amely decompositio, comparative judgmet, ad sythesis of priorities (Saaty, 1994). Based o AHP priciples, the recommeded procedure for usig the AHP methodology ca be summarized as follows: Step 1: Set up the decisio hierarchy by decomposig the problem ito a hierarchy of iterrelated elemets. The top level of the hierarchy represets the ultimate goal, while the lowest level is composed of all possible alteratives. Oe or more itermediate levels cotai the evaluatio criteria ad sub-criteria. Step 2: Costruct a set of pairwise compariso matrices for each level of the hierarchy, ad make all the pairwise comparisos. The pairwise compariso matrix is deoted by A = [a ] where a implies the relative importace (or preferece) of elemet i over elemet j. Each etry i matrix A is positive (a > 0) ad reciprocal (a = 1/a ji for all i, j = 1,2,,). To fill the matrix of A, Saaty proposed a 9-poit pairwise compariso scale as show below. For example, if a = 5, this meas that a i is strogly more importat tha a j. 1/9 1/7 1/5 1/3 1 3 5 7 9 extremely very strogly strogly moderately equally moderately strogly very strogly extremely less importat more importat (1/8, 1/6, 1/4, 1/2, 2, 4, 6, 8 : Itermediate values betwee the above adjacet values) Step 3: Calculate the priority vector ad cosistecy ratio for each pairwise compariso matrix. The priority vector w = (w 1, w 2,, w ) T is geerated by ormalizig the pricipal eigevector w of the matrix A: A w = λ max w where λ max is the largest eigevalue of the matrix A. To check the cosistecy i pairwise compariso judgmet, the cosistecy ratio (CR) is calculated usig CR = CI / RI where cosistecy idex (CI) is give by CI = (λ max ) / ( 1) ad radom Idex (RI) is obtaied by from the followig radom idex table. If the value of CR is 0.1 or less, the the judgmet is cosidered to be cosistet ad acceptable. Order of matrix () 1 2 3 4 5 6 7 8 9 10 RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 Step 4: Sythesize the priority vectors of each level obtaied i the Step 3, to produce the overall score for each alterative. The overall score for the i th alterative (S i ) is computed as follows: = i c j j= 1 S w where c j is the vector of priorities associated with the j th elemet of the hierarchy ad w is the vector of priorities derived from comparig the alteratives i each criterio. Revista de Gestão da Tecologia e Sistemas de Iformação/Joural of Iformatio Systems ad Techology Maagemet

Evaluatig the service quality of third- party logistics service providers usig the aalytic hierarchy process 265 4. A Case Study of Korea 3PL Service Providers To demostrate the practical applicatio of the idea discussed above, a empirical case study was coducted with a Iteret shoppig mall i Korea, which wated to cotract a appropriate 3PL service provider. Four compaies were cosidered as potetial 3PL service providers, which were represeted as 3PL A, 3PL B, 3PL C ad 3PL D. 4.1 Structurig the decisio hierarchy We first structured the decisio hierarchy i which all decisio elemets were classified ito four levels as show i Figure 1. The highest level (Level 1) of the hierarchy stads for the ultimate goal that was to evaluate 3PL service quality ad select a appropriate 3PL service provider. The five service quality dimesios idetified to achieve this goal are located at the secod level (Level 2). At the third level (Level 3), the absolute measuremet mode of the AHP (i.e., a five-poit ratig scale of outstadig, good, average, fair ad poor) is applied to rate each alterative accordig to each criteria i the level directly above. I our case, sice the decisio maker has ot eough kowledge or experiece about all the alteratives, it is quite difficult to directly compare the alteratives with each other. Therefore, i a deviatio from the usual AHP approach, a ratig scale was used to elimiate these difficulties ad eable the decisio maker to assig a ratig to a alterative without makig direct comparisos. The lowest level (Level 4) cotais the alteratives to be evaluated, amely four differet 3PL service providers. Level 1: Ultimate Goal Evaluatig the Service Quality of 3PL Service Providers Level 2: Dimesios Tagibles Reliability Resposiveess Assurace Empathy Level 3: Ratig scale Outstadig Good Average Fair Poor Level 4: Alteratives 3PL A 3PL B 3PL C 3PL D Figure 1. Decisio hierarchy for evaluatig 3PL service quality Vol.3, No.3, 2006, p. 261-270

266 So, SH., Kim, JJ., Cheog, KJ., & Cho, G. 4.2 Makig the pairwise compariso matrix Oce the decisio hierarchy is costructed, pairwise comparisos are made betwee the elemets at each level of the hierarchy with respect to the coected elemet i the level above. I order to make pairwise compariso betwee five service quality dimesios at level 2, a questioaire was desiged ad set out to the customers of the four alterative 3PL service providers i February 2005. The target respodets were the geeral maagers or logistics maagers of the sampled shipper compaies. 23 customers of 3PL A, 21 customers of 3PL B, 26 customers of 3PL C, ad 19 customers of 3PL D replied to the questioaires. Of these, 67 idividual pairwise compariso matrices with cosistecy ratio of less tha 0.1 were aggregated usig the geometric mea method. The aggregate pairwise compariso matrix is preseted i Table 1. Table 1. Aggregate pairwise compariso matrix for five service quality dimesios Tagibles Reliability Resposiveess Assurace Empathy Priorities Tagibles 1.000 1.267 0.915 1.036 0.962 0.205 Reliability 0.789 1.000 0.790 1.088 1.005 0.185 Resposiveess 1.093 1.266 1.000 1.272 1.036 0.224 Assurace 0.965 0.920 0.786 1.000 0.859 0.180 Empathy 1.040 0.995 0.965 1.164 1.000 0.205 λ max = 5.011, CI = 0.002, RI = 1.12, CR = 0.002 4.3 Calculatig the priority vector After the pairwise compariso process is completed, the priority vector is calculated by the followig two-stage procedure. First, each etry i colum i of the matrix A is divided by the sum of the etries i colum i. This yields a ormalized matrix Ā which is defied as: Ā = [ā ] where a = a a ik k= 1 for i, j = 1,2,, Secod, the average value of the etries i row i of the ormalized matrix Ā is computed to get the priority weights or eigevector, which is determied by: W = [w k ] where w k = a i= 1 for j, k = 1,2,, Usig the above calculatio procedures, the priority vector for the five service quality dimesios is derived ad the result obtaied is give i the last colum of Table 1. Also, as show below the matrix, the cosistecy ratio is foud to be less tha 0.1, which is typically cosidered acceptable. The priority vector i Table 1 reflects the 3PL customers view that the most importat dimesio i evaluatig the service quality of Revista de Gestão da Tecologia e Sistemas de Iformação/Joural of Iformatio Systems ad Techology Maagemet

Evaluatig the service quality of third- party logistics service providers usig the aalytic hierarchy process 267 3PL service providers is resposiveess (0.224), followed by tagibles (0.205), empathy (0.205), reliability (0.185) ad assurace (0.180). As metioed earlier, we used a five-poit ratig scale to rate the four alterative 3PL service providers accordig to five service quality dimesios i level 2. Usig pairwise compariso alog a five-poit ratig scale, the relative weights of outstadig (O), good (G), average (A), fair (F) ad poor (P) were determied as 0.513, 0.261, 0.129, 0.063 ad 0.034, respectively (Liberatore et al., 1992; Liberatore, 1987). The customers of each of the four 3PL service providers were the asked to assig the five-poit ratig scale to their 3PL service provider with respect to each of the five service quality dimesios i the questioaire. The resultig cosesus ratigs are show i the cells labeled R i Table 2. 4.4 Determiig overall score for each alterative Fially, the overall score for each of the four alterative 3PL service providers is computed for the purpose of selectig the most appropriate 3PL service provider i terms of service quality. The overall score S i for the i th 3PL service provider is computed as follows: i S = w j= 1 j r for i = 1,2,, where w j is the priority weight (local weight) of the j th service quality dimesio i level 2 ad r is the ratig score of the i th 3PL service provider with respect to the j th service quality dimesio. Table 2. Overall scores of four alterative 3PL service providers Dimesios Tagibles Reliability Resposiveess Assurace Empathy Overall Normalized Alteratives L: 0.205 L: 0.185 L: 0.224 L: 0.180 L: 0.205 scores scores 3PL A 3PL B 3PL C R: 0.261 R: 0.129 R: 0.261 R: 0.063 R: 0.129 G: 0.054 G: 0.024 G: 0.058 G: 0.011 G: 0.026 0.174 0.298 R: 0.129 R: 0.063 R: 0.129 R: 0.063 R: 0.063 G: 0.026 G: 0.012 G: 0.029 G: 0.011 G: 0.013 0.091 0.156 R: 0.129 R: 0.261 R: 0.513 R: 0.129 R: 0.261 G: 0.026 G: 0.048 G: 0.115 G: 0.023 G: 0.054 0.266 0.456 R: 0.063 R: 0.034 R: 0.063 R: 0.034 R: 0.063 3PL D G: 0.013 G: 0.006 G: 0.014 G: 0.006 G: 0.013 0.052 0.090 *) R: Ratig Scores, L: Local Weights, G: Global Weights Vol.3, No.3, 2006, p. 261-270

268 So, SH., Kim, JJ., Cheog, KJ., & Cho, G. Based o the (ormalized) overall scores of the four 3PL service providers show i Table 2, we fid that 3PL C has the highest overall score amog four alteratives. Therefore it must be selected as the best 3PL service provider, as it satisfies all the evaluatio criteria for 3PL service quality. 5. Coclusios I this paper, we preseted a efficiet way of applyig the AHP to selectig the best 3PL service provider, based o service quality evaluatio. I order to measure 3PL service quality, we utilized the five geeric dimesios of SERVQUAL. Although the coceptualizatio ad dimesioality of SERVQUAL have bee subjected to some severe criticisms (Buttle, 1996), there is a geeral agreemet that the five dimesios are reasoably accurate predictors of perceived service quality (Sureshchadar et al, 2002). We also applied the AHP approach to solve a 3PL service provider selectio problem. The major advatage of this approach is that it gives a relative importace to the evaluatio criteria ad a assessmet of alteratives based upo each criterio. To explore practically the ideas preseted i this paper, we coducted a empirical case study o four 3PL service providers i Korea. The results show i Table 1 idicate that Resposiveess (the willigess to help customers ad provide prompt service) out of the five service quality dimesios is cosidered as the most importat dimesio perceived by 3PL customers. Furthermore, accordig to the overall service quality scores show i Table 2, 3PL C has bee chose as the best 3PL service provider with respect to service quality. However, i order for our results to be more complete, further research should be udertake to establish a set of metrics to quatify each of the dimesios of 3PL service quality discussed i this paper. Refereces Buttle, F. (1996), SERVQUAL: review, critique, research ageda, Europea Joural of Marketig, Vol. 30, pp. 8-32. Lagley Jr., C.J., Va Dort, E., Ag, A., ad Sykes, S.R. (2005), Third-Party Logistics: Results ad Fidigs of the 10th Aual Study. Liberatore, M.J. (1987), A Extesio of the Aalytical Hierarchy Process for Idustrial R&D Project Selectio ad Resource Allocatio, IEEE Trasactios o Egieerig Maagemet, Vol. EM-34, No. 1, pp. 12-18. Liberatore, M.J., Nydick, R.L., ad Sachez, P.M. (1992), The Evaluatio of Research Papers (Or How to Get a Academic Committee to Agree o Somethig), Iterfaces, Vol. 22, No. 2, pp. 92-100. Lieb, R.C., Mille, R.A., ad Va Wassehove, L.N. (1993), Third Party Logistics Services: A Compariso of experieced America ad Europea Maufacturers, Iteratioal Joural of Physical Distributio ad Logistics Maagemet, Vol. 23, No. 6, pp. 35-44. Parasurama, A., Zeithaml, V.A., ad Berry, L.L. (1988), SERVQUAL: A multipleitem scale for measurig cosumers perceptios of service quality, Joural of Retailig, Vol. 64, No. 1, pp. 22-37, 1988. Revista de Gestão da Tecologia e Sistemas de Iformação/Joural of Iformatio Systems ad Techology Maagemet

Evaluatig the service quality of third- party logistics service providers usig the aalytic hierarchy process 269 Parasurama, A., Zeithaml, V.A., ad Berry, L.L. (1991), Refiemet ad Reassessmet of the SERVQUAL Scale, Joural of Retailig, Vol. 67, No. 4, pp. 420-450. Saaty, T.L. (1980), The Aalytic Hierarchy Process, McGraw-Hill, New York. Saaty, T.L. (1994), The Fudametals of Decisio Makig ad Priority Theory with the Aalytic Hierarchy Process, Volume VI of the Aalytic Hierarchy Process Series, RWS Publicatio, Pittsburgh, PA. Saaty, T.L. (2006), Rak from comparisos ad from ratigs i the aalytic hierarchy/etwork processes, Europea Joural of Operatioal Research, Vol. 168, No. 2, pp. 557-570. Sahay, B.S. ad Moha, R. (2006), Maagig 3PL relatioships, Iteratioal Joural of Itegrated Supply Maagemet, Vol. 2, No. 1/2, pp. 69-90. Sureshchadar, G.S., Rajedra, C., ad Aatharama, R.N. (2002), The relatioship betwee maagemet s perceptio of total quality service ad customer perceptios of service quality, Total Quality Maagemet, Vol. 13, No. 1, pp. 69-88. Tsaur, S.H., Chag T.Y., ad Ye, C.H. (2002), The evaluatio of airlie service quality by fuzzy MCDM, Tourism Maagemet, Vol. 23, pp. 107-115. Outsourcig Logistics Report 2006, http://www.eyefortrasport.com (accessed 21 March). Vol.3, No.3, 2006, p. 261-270