ENHNCING CUSTOMER EXPERIENCE THROUGH BUSINESS PROCESS IMPROVEMENT: N PPLICTION OF THE ENHNCED CUSTOMER EXPERIENCE FRMEWORK (ECEF) G.J. Both 1, P.S. Kruger 2 & M. de Vries 3 Deprtment of Industril nd Systems Engineering University of Pretori, South fric 1 jozine.both@up.c.z, 2 pul.kruger@up.c.z, 3 mrne.devries@up.c.z BSTRCT In 2010 Both, Kruger & De Vries [5] introduced frmework for enhncing customer experience through improved business processes. They developed the Enhnced Customer Experience Frmework (ECEF) by integrting vrious well-known techniques into one comprehensive frmework. The ECEF incorportes three well-known tools: qulity function deployment (QFD), business process re-engineering, nd simultion. QFD is used to determine the reltionship beween business processes nd customer requirements, nd to prioritise business processes from customer perspective. Business Process Re-engineering (BPR), together with the steps ssocited with benchmrking, provides useful process tht my be followed when reengineering business processes to fit customer needs; while simultion modelling is used to test the impct of process improvements on customer experience. This pper ims prtilly to vlidte the ECEF ginst empiricl dt obtined from the telecommunictions industry. OPSOMMING n Rmwerk om kliënte-ervring te verbeter deur die herontwerp vn besigheidsprosesse is in 2010 deur Both, Kruger & De Vries [5] bekendgestel. Die verbeterde-kliënte-ervringrmwerk is ontwikkel deur welbekende tegnieke te integreer tot een omvttende rmwerk. Die verbeterde-kliënte-ervring-rmwerk inkorporeer bekende tegnieke soos gehltefunksie-ontplooiing, besigheidsprosesherontwerp en simulsiemodelering. Gehltefunksieontplooiing word gebruik om die verhouding tussen besigheidsprosesse en kliëntebehoeftes te bepl, sook om die prosesse te prioritiseer vnf n kliëntperspektief. Besigheidsprosesherontwerp tesme met die stppe gessosieer met die gebruik vn n verwysingsnorm, bied n bruikbre proses wt gevolg kn word om besigheids prosesse te herontwerp om n kliëntebehoeftes te voldoen. Simulsiemodelering kn vervolgens gebruik word om die impk vn die prosesverbeteringe op kliënte-ervring te meet. Hierdie rtikel beoog om die verbeterde-kliënte-ervring-rmwerk gedeeltelik te vlideer deur die toepssing drvn op empiriese dt, verkry uit die telekommuniksie industrie. South fricn Journl of Industril Engineering My 2012 Vol 23 (2): pp 39-56
1. INTRODUCTION In the current business environment, customers need to be the centrl concern of mngement [12]. Jmes et l. [12] stte tht when compnies mke customers prmount in their business strtegies, rdicl shift occurs in the wy they mnge nd mesure success. New economics of service demnd innovtive mesurement techniques tht cn help mngers to build customer stisfction nd loylty, nd t the sme time mesure the corresponding impct on profitbility nd growth [12]. Jmes et l. [12] mintin tht the lifetime vlue of loyl customer cn be enormous. This vlue cn be enhnced when referrls re dded to the economics of customer retention nd repet purchses of relted products. From this it is cler tht customer retention is extremely importnt. It is dependent on customer stisfction nd on how customers experience the products nd services of the compny. If compny cn succeed in stisfying customer needs, they will be ble to retin their customers. Customers in the telecommunictions industry cn choose from n rry of products nd services from more thn one telecommunictions compny. South fric hs three mjor cellulr networks: Vodcom, MTN, nd Cell C. prt from the three cellulr gints, there is one fixed line network. ll four compnies compete for voice trffic. Qulity of service is impertive in the industry, s it serves s differentitor [1]. The shift of power from compny to customer forces compnies to focus on customer retention nd loylty through improved customer experience. The Enhnced Customer Experience Frmework (ECEF), introduced by Both et l. [5], my ssist compnies to enhnce their customer experience by delivering qulity service through customer-centred business processes. The ECEF cn be seen in Figure 1. The frmework hs seven stges, which cn be seen s the ECEF process steps followed to build the frmework. They re discussed in detil by Both et l. [5]. In this pper the conceptul frmework developed by Both et l. [5] is prtilly vlidted using dt obtined from the telecommunictions industry. The limittions of this prtil vlidtion re discussed in Section 4 of this pper. 2. THEORETICL BCKGROUND OF THE ECEF The ECEF is bsed on vrious theoreticl concepts tht provide process for building the frmework. Vrious principles nd methods cn be derived from these concepts. The Business Process Re-engineering (BPR) methodology developed by Dvenport & Short [9] provides process nd the principles for the frmework. Business Process Benchmrking, defined by Cmp [7], provides vrious process steps for the frmework. The service design nd mngement model defined by Rmswmy [13] provides insight into how benchmrking cn be used to determine both the current nd the desired performnce level for ech business process. Rmswmy [13] lso provides rtionle for using QFD over other techniques such s the tree digrm. The customer mesurement nd mngement system defined by Gustfsson & Johnson [10] provides the method for gthering customer dt for the QFD digrm. frmework for nlysing the qulity of the customer interfce defined by Bitrn [3] provides further principles ssocited with determining customer requirements tht cn be used for the QFD digrm. The vlue chin nlysis defined by Chse et l. [8] provides method for identifying key business processes tht deliver vlue to the customer. Both et l. [5] motivted their choice of tools nd the design trde-offs for the ECEF by referring to the literture listed bove nd referenced by Both & Vn Rensburg [6]. 40
http://sjie.journls.c.z Figure 1: The Enhnced Customer Experience Frmework (ECEF) (Both et l. [5]) 3. THE TELECOMMUNICTIONS INDUSTRY IN SOUTH FRIC The strtegic position of the telecommunictions industry in South fric my be nlysed using Porter s five forces model. The model emphsises five potentil competitive forces [11]: The thret of new entrnts. The thret of substitute goods or services. 41
The brgining power of customers. The brgining power of suppliers. The degree of rivlry mongst existing competitors. In South fric the telecommunictions industry is fced with the thret of new entrnts s one of the lrgest competitive forces. In 2001 third telecommunictions compny ws introduced in South fric, nd the mrket shre tht hd belonged to only two compnies ws redistributed between three. For ll of them to be competitive, they hd to invest in delivering qulity service to customers. Unfortuntely, s mentioned by Both & Vn Rensburg [6], the telecommunictions industry hs not yet succeeded in giving customers wht they feel they need. Telecommuniction compnies re in dire need of service delivery processes tht not only meet customer requirements but exceed them, nd so deliver n exceptionl customer experience. 4. PRTIL VLIDTION STRTEGY OF THE ECEF The ECEF is prtilly vlidted in this pper by pplying the frmework stges to dt obtined from one of the three mjor cellulr networks in South fric. To mintin confidentility, it will be referred to s Compny. This pper ttempts to enhnce the customer experience of Compny by pplying the ECEF to improve one of their business processes. The other two mjor cellulr networks re referred to s Compny B nd Compny C, nd re the two min competitors of Compny. There re certin limittions to the vlidtion of the ECEF: Due to the unvilbility of dt, not ll the dt used in this pper could be obtined from Compny. Certin dt points re estimted by the uthor for illustrtion purposes, nd the output cn thus not be pplied directly to Compny. For the purpose of the prtil vlidtion, only one wht-if scenrio is determined for ech process, nd is tested through simultion. The min gol of the frmework is to improve business processes nd to mesure the impct on customer stisfction nd consequently on customer experience. The cost involved in implementing the improvements does not form prt of the ECEF, nd so the costing nlysis of improvement inititives does not form prt of the prtil vlidtion. For the purpose of the prtil vlidtion, only one KPI is specified for ech process tht my be ssocited with one of the customer requirements of tht process. The feedbck loop from stge seven bck to stge one is not ddressed in this prtil vlidtion; it will be prt of future reserch. 5. VLIDTION OF FRMEWORK STGES 5.1 Stge One: Develop frmework objectives The first step in the vlidtion of the frmework is to understnd the strtegic plnning of the compny bout mesuring customer experience. The compny s strtegy must be tken into ccount when developing objectives for the compny-specific frmework. The objective of the ECEF remins the sme: the focus is on improving customer experience through improved business processes [5]. The first step in chieving this objective is to identify where the compny went wrong from customer perspective. When identifying problem res, the compny should decide, on strtegic level, which mrket segment to trget first. Gustfsson & Johnson [10] suggest tht customers cn be grouped into segments bsed on customer needs, benefits sought, or personl vlues served. For the purpose of vlidting the ECEF, the post-pid consumer segment is chosen minly due to the number of complints received from post-pid customers, but lso due to the quntity of informtion nd dt vilble for this segment. 42
5.2 Stge Two: Identify key business processes nd define KPIs for ech process ccording to Both et l. [5], it is importnt to identify nd document ll the business processes in the orgnistion tht hve n impct, direct or indirect, on customer experience in selected mrket segment. vlue chin nlysis my be done to identify list of high level business processes for Compny. Chse et l. [8] suggest the vlue chin s structure to cpture the linkge of orgnistionl ctivities tht crete vlue for the customer nd profit for the firm. Vlue chin nlysis ws chosen s the preferred technique for Stge Two of the ECEF, s the vlue chin my be useful to identify ll the key business processes from Compny tht my deliver vlue to the customer. Figure 2 shows typicl vlue chin for compny like Compny. Dt for vlue chin nlysis re obtined from existing sources within Compny. Figure 2: Vlue chin for Compny fter the vlue chin is defined for Compny, business processes my be determined tht deliver ech of the vlue-dding ctivities to the customer. Tble 1 provides list of identified high level business processes for Compny. Process ID B C D E F G H I J K L M N Process nme End to end network roll out Product lifecycle Hndset repir Stock provision Credit vetting nd ctivtions Cmpign (promo) mngement Distribution footprint Post-ctivtion service mngement Customer upgrdes nd retention Number mngement HR (resource recruitment) In-store customer service process Employee trining Billing nd pyment Tble 1: Identified high level business processes for Compny The processes re relted to the ctivities depicted in the vlue chin. Figure 3 illustrtes where ech of the bove-listed processes will tke plce in the vlue cretion process. Figure 3: Business processes delivering vlue in the vlue chin 43
The business objective of the ECEF is to enhnce customer experience. It is vitl to identify ll the business processes tht re currently in conflict with this objective. nlysing dt on customer complints my help to identify the res nd consequently the processes with which customers re not stisfied. Customer complint dt within Compny cn then be used to choose smple of processes tht should be re-engineered first. ccording to relevnt stkeholders, the following four processes re in dire need of improvement from customer perspective: The repir process. The new del process. The upgrde process. The in-store customer service process. These processes my be mpped nd problems surrounding the processes documented. For the purpose of this pper only one of the bove-mentioned processes will be nlysed nd improved. The process chosen for this pper is the In-store customer service process, due to the lrge number of complints received in this re. While mpping the process, the performnce of the process should lso be tken into ccount. Understnding process performnce from customer perspective is vitl for the development of the compnyspecific frmework. Process performnce from the technicl perspective is lso importnt, nd cn sometimes differ substntilly from the perspective of the customer. The current business processes of Compny re designed from the technicl perspective. The gol of the frmework is to link the two perspectives to do the technicl redesign of business processes while considering the perspective of the customer. The technicl performnce of the repir process my be mesured by defining Key Performnce Indictors (KPIs) for the process tht my be ssocited with one of the customer requirements. The KPIs should be specified by the design tem. KPIs re specified for the repir process, bsed on the business objectives of enhncing customer stisfction with regrd to specific customer requirement. Mesuring the KPI will give n indiction whether or not the process stisfies the specific customer requirement. The KPIs serve s technicl evlution for the repir process. For the purpose of frmework vlidtion, the ssumption is mde tht improving the specified KPI will led to the desired outcome of incresed customer stisfction. The ssumption is tested in Stge Seven with the id of simultion modelling. The in-store customer service process mp, existing problems nd identified process KPIs re discussed in the next sections. The in-store customer service process The in-store customer service process hs significnt impct on customer experience. The process encompsses the entire time the customer spends in the system. Unfortuntely the current queue length in mny stores leds to frustrted nd unhppy customers. The min problem to be ddressed is the long witing times experienced by customers. Figure 4 shows the in-store customer service process. Informtion for the process mp ws collected from relevnt stkeholders. The KPI identified for the process is the totl time the customer spends in the queue witing for ssistnce. The time is influenced by vrious fctors, such s: Queue length. Number of sles consultnts. Time customer spends in service. Number of customers entering the store. 44
Customer enters store Wits in queue Receive customer t Enquiries Enquiries ble to resolve query Yes Service t enquiries No Wit in queue Yes Services counter full? Refer to Services counter Yes Require dditionl services? No No Opening t Services counter? Yes ssist customer t Services counter Customer exit No 5.3 Stge Three: Gther customer dt Figure 4: The in-store customer service process In this stge, customer dt should be gthered to obtin ll the customer requirements, together with importnce rtings of the requirements for ech business process identified. The customer dt required for the ECEF my be obtined through the process of CIT reserch s defined by Gustfsson & Johnson [10]. Both et l. [5] concluded tht fce-tofce interviews were the best pproch to obtin customer needs, while mil/telephone surveys could be used to obtin importnce nd stisfction rtings from the customer. It is importnt, however, to identify existing sources of informtion within the compny. Relevnt secondry informtion should be compiled nd ssessed to determine whether the informtion my be used s input for Stges Four nd Five of the ECEF. If relevnt dt exists, it should be used insted of conducting new surveys, since it will sve significnt mount of time nd money. Compny hs lrge quntities of customer informtion from previously conducted surveys. In 2008 Compny gthered vluble informtion from their customers through surveys nd questionnires. Informtion ws gthered from 2,081 rndomly selected customers distributed over Compny, B nd C. The following tble shows the smple selected for ech of the three compnies. Compny nme Number surveyed Compny 1,240 Compny B 730 Compny C 111 TOTL 2,081 Tble 2: Smple surveyed The smple demogrphics for the survey shown in Figure 5 give good indiction of the rndom nture of the smpling. The min gol of the survey conducted by Compny ws to estblish the key influences on customer stisfction nd loylty. The survey ws conducted through rndom smpling nd mde use of individul fce-to-fce interviews. The survey consists of fctors nd ttributes tht drive customer loylty in South fric, nd includes rting on how well the compny nd its competitors stisfy ech specific fctor. The reserch objectives of the survey re s follow: Understnding customers expecttions nd uncovering the reltionship between Compny s products nd services, fctors nd ttributes tht drive customer loylty in South fric. Mesuring the importnce of the fctors nd ttributes for strtegic improvement cross customer segments. 45
Mesuring Compny s performnce in ech fctor, nd ttribute nd clerly understnd the extent to which customers expecttions in the post-pid segment re met. From these objectives it is cler tht the dt gthered in 2008 through the stisfction nd loylty survey cn esily be interpreted to obtin the dt needed for Stges Four nd Five of the ECEF. The fctors nd ttributes surveyed my be interpreted s customer requirements. The importnce nd performnce determined for ech fctor cn therefore be used s importnce nd stisfction rtings for ech requirement. lthough the informtion ws gthered in 2008, nd my be of little vlue to Compny becuse it is outdted, it my be used to vlidte the frmework introduced by Both et l. [5]. By using the secondry dt, the use of the frmework cn be illustrted within rel life orgnistion. Since Compny hs relible nd complete secondry dt, the existing dt sets re used in this pper to vlidte the ECEF. 46 Figure 5: Smple demogrphics 5.4 Stge Four: Link business processes to customer requirements In this stge, the business processes identified in Stge Two re linked to their ssocited customer requirements identified in Stge Three. Both & Vn Rensburg [6] suggest the Qulity Function Deployment (QFD) digrm s suitble tool to link processes to requirements, becuse it cn relte ny identified set of business processes to ny number of requirements. The informtion obtined from QFD is detiled enough to estblish whether the reltionship between given process nd set of requirements is wek or strong. This stge consists of the QFD digrm for Compny with reltionship mpping between identified business processes nd customer requirements. ccording to Both et l. [5], the reltionship mpping is determined by nlysing the extent to which specific business process could techniclly influence the customer requirement. They suggest tht the reltionships re defined using 1-, 3- or 9-point scle, where one represents wek reltionship, three represents moderte reltionship, nd nine represents strong
reltionship. group of people from Compny who hd the necessry knowledge of the defined set of business processes ws interviewed during workshop to obtin the reltionship mpping. The importnce of ech requirement ws determined through the survey on scle from 0 to 100 percent, where 0 percent represents zero importnce nd 100 percent represents bsolute importnce. Figure 6 shows portion of the QFD digrm for Compny tht illustrtes the reltionship between the business processes nd the requirements. Figure 6: QFD relting business processes to customer requirements for Compny 5.5 Stge Five: Prioritise business processes bsed on their impct nd performnce During this stge, the business processes should be ctegorised nd displyed ccording to their importnce nd performnce by mking use of the strtegic stisfction mtrix defined by Gustfsson & Johnson [10]. The mtrix is prticulrly useful in prioritising business processes by dividing them into four strtegic ctegories, depending on their impct on customer stisfction nd their performnce. Mesuring the impct of process on customer stisfction is firly simple process. Once Stge Four is complete nd ll the business processes of Compny re relted to their ssocited requirements, prioritistion of business processes bsed on their impct on customer experience is utomticlly completed. In Figure 6, the row lbelled Rw Score gives the totl score ech process hs obtined, nd directly represents the impct the process hs on customer requirements. The clcultion of the Rw Score is given by eqution (1): Rw Score = n i=1 y i x ji j {1 n} (1) where: 47
y i Importnce rting of requirement i, i {1 n} x ij Reltionship vlue between requirement i nd process j, i {1 n}, j {1 n} In Figure 7 the business processes of Figure 6 re prioritised ccording to impct. Business process importnce from customer perspective 36% 26% 26% 12% Repir process New del process Upgrde process In-store customer service process Figure 7: Relevnt importnce of business processes from customer perspective The performnce of the processes from customer perspective cn be mesured by using the stisfction rtings obtined by the survey. The survey rting obtined on how well the competitors stisfy ech need is not relevnt for this stge, nd is used in Stge Six. Bsed on the stisfction rting specified by the customer (0 100 percent), it becomes possible to identify the process performnce for Compny from customer perspective. Both et l. [5] explin how this cn be done by expnding on the trditionl concept of QFD nd using the tool to obtin quntittive performnce score for ech process. Figure 8 shows the QFD digrm for Compny, where the process performnce of ech business process is determined using Trget score nd Current score clculted with equtions (2) nd (3). The Process performnce is then clculted using equtions (4) s specified by Both et l. [5]. where: Trget Performnce = n i=1 y i x ij z i j {1 n} (2) Current Performnce = n i=1 y i x ij s i j {1 n} (3) Current Performnce Process Performnce = (4) Trget Score z i Trget stisfction rting of requirement i, i {1 n} s i Stisfction rting for the producing compny of requirement i, i {1 n} The clcultions re bsed on formuls tht should form prt of the QFD digrm. The formul for clculting the Trget score contins the vlues specified by Compny s the trget stisfction rtings. The unvilbility of dt ment tht the trget stisfction rtings could not be obtined. For demonstrtion purposes, the trget rtings re set t 100 percent. This exmple is thus only used for demonstrtion purposes, nd the output does not pply in prctice. 48
Figure 8: Prioritising processes ccording to their performnce from customer perspective for Compny The four chosen business processes identified for Compny cn now be prioritised, bsed on the process performnce clculted in this stge. Business process performnce from customer perspective 83% 80% 80% 78% Repirs process New dels process Upgrdes process In-store customer service process Figure 9: Performnce of business processes from customer perspective The processes cn be plotted in the strtegic stisfction mtrix to prioritise the four processes fully, bsed on their impct nd performnce. The mtrix indictes which process 49
should be focused on first. ccording to Gustfsson & Johnson [10], resources should be focused where impct is high nd performnce is wek. Improvements within this ctegory will hve the gretest impct on customer stisfction nd thus on loylty nd profitbility. In this cse, no processes fll within this ctegory. The four selected processes for Compny fll within the upper two qudrnts. The in-store customer service process is highest on the priority list, s it hs the highest impct nd the wekest performnce. From the mtrix it my be ssumed tht customer stisfction will be enhnced when improving the in-store customer service process. Improvement efforts should thus be focused on this process first, nd then on ech of the other processes ccording to their mtrix position. 50 Figure 10: Strtegic stisfction mtrix for Compny 5.6 Stge Six: Estimte the desired performnce level for the business processes In this stge, the desired performnce level from customer perspective should be estimted prior to ttempting process improvements. Both et l. [5] concluded from the literture tht QFD is suitble method to estimte the desired performnce level. In this stge the desired performnce level is estimted by exmining the extent to which the performnce of similr services in the mrket stisfies customer needs. The desired performnce level is thus determined through benchmrking. The first step ssocited with Stge Six is to identify the competitors of Compny tht deliver similr products nd services to the mrket. s discussed in section 2 of this pper, two other telecommuniction compnies deliver similr products nd services to customers; they will be referred to s Compny B nd Compny C. In the stisfction loylty survey done for Compny in 2008, customers were sked to rte the service delivered by the competitors on scle from 0 to 100 percent. Once the stisfction rtings hve been obtined, technicl benchmrk should be specified for ech process. The technicl benchmrks my be obtined by mesuring the KPI specified for ech process for Compny s well s the competitors. The estimted mesurements re bsed on dt obtined from relevnt stkeholders. The KPIs cn be linked to one of the customer requirements ssocited with ech process. The desired performnce level cn thus be estimted by looking t the customer benchmrk nd the ctul performnce or technicl benchmrks of the competitors. The QFD in Figure 11 shows competitor stisfction rtings s well s the technicl benchmrks for ech process. From Figure 11, the totl time spent witing in the queue is estimted to be n hour t most for Compny, while the stisfction rting for the ssocited requirement is 68 percent. The highest stisfction rting is obtined for Compny B, where customers rted the time spent witing in the queue t 69 percent; the corresponding technicl benchmrk for the rting is 30 minutes. The performnce level of 30 minutes is thus closer to the performnce level desired by customers, nd the desired performnce level cn be estimted t less thn 30 minutes. For the purposes of this pper, only the in-store customer service process will be discussed, s this process requires the most ttention.
trget performnce level nd lowest perceived performnce level must lso be specified in order to determine the best nd worst technicl benchmrks for ech process. These performnce levels my be obtined by interviewing customers of Compny. For the purposes of this pper, the trget nd the lowest perceived performnce levels re estimted by the uthor to show how the performnce/stisfction rting my be obtined. The survey done by Compny in 2008 does not contin sufficient informtion on the two mesures, nd so n estimted vlue for both the trget performnce level nd the lowest perceived performnce level is used. The trget performnce level is the level of performnce tht corresponds to 100 percent stisfction rting. The lowest perceived performnce level is the level of performnce tht corresponds to zero percent stisfction rting; it is thus the worst performnce or the lowest perceived performnce where customer stisfction reches zero percent. The difference between the desired performnce level nd the trget performnce level is tht the desired performnce level specifies the minimum performnce level t which Compny must perform to remin competitive. The trget performnce level specifies the best possible performnce tht will gurntee 100 percent stisfction from the customer. The design performnce level will lso depend on the mount of money the compny is willing to spend. In the next stge simultion models re used to test possible improvement inititives for the in-store customer service process. Wht-if nlysis cn be done with improvement inititives to determine how the process my be improved to rech the desired performnce level. 5.7 Stge Seven: Specify improvement inititives nd test their impct on customer experience 5.7.1 Specify improvement inititives In this stge, improvement inititives in the form of wht-if scenrios re defined for the in-store customer service process. The wht-if scenrio is specified with the im of improving the performnce of the process ginst the KPI chosen for the process, nd is shown in Tble 3. The improvement in performnce lso then hs n impct on the stisfction rting of ech ssocited requirement. The wht-if scenrio must be tested with simultion models to determine the extent to which the process my be improved. Vrious wht-if scenrios cn be identified nd tested to determine the best possible improvement inititive for the process. The different improvement inititives my lso be nlysed to determine the cost ssocited with ech lterntive to help choose the best inititive. For the purpose of the prtil vlidtion, only one wht-if scenrio for the process is determined nd tested through simultion. The costing nlysis of the improvement inititives does not form prt of the prtil vlidtion. The following improvement inititive my contribute to higher customer stisfction. Process nme In-store customer service process Key performnce indictor Totl time spent witing in the queue Wht-if scenrio/improvement inititives Increse the number of consultnts Tble 3: Wht-if scenrios/improvement inititives specified for ech process 51
52 Figure 11: Determining the desired performnce level for ech process The bove scenrio my be simulted to identify its impct on customer experience. To cpture ll the necessry detil to nlyse the specified improvement inititive, simultion model is built using REN softwre. The dt for the model ws gthered through one-on-one interview with store mnger from Compny. Dt were lso obtined by questioning selected customers from Compny. ccording to Bnks et l. [2], the Poisson rrivl process hs been employed successfully s model of the rrivl of people t service fcilities such s resturnts nd bnks. In these fcilities, rrivl tkes plce rndomly from lrge clling popultion, from which customers mke independent decisions bout when to rrive. Bnks et l. [2] stte tht the Poisson rrivl process is the most importnt model for rndom rrivls. From this the ssumption is mde tht customers rrive t the service fcility of Compny ccording to rndom distribution.
tringulr distribution for process times is used s n input prmeter for the model, becuse this distribution is typiclly used s subjective description of popultion where only limited smple dt is vilble [4]. The tringulr distribution is bsed on the minimum, mximum, nd most likely guess s to wht the process time might be, nd is very useful for modelling processes where the reltionship between vribles is known but dt is scrce [4]. Simulted scenrio: Increse the number of consultnts n verge of three sles consultnts re presently used to serve customers in delerships. The number is incresed to observe the impct on the queue length, witing time, totl time in system, nd process completion time. Tble 4 depicts the results obtined from the simultion model. The number of sles consultnts is incresed from three to six consultnts. The improvement ffects the queue length nd therefore lso the time spent in the queue. ccordingly, the current mximum witing time is bout one hour nd 24 minutes, nd customers encounter queues with mximum length of 7 persons. If four sles consultnts re employed, the improvement is remrkble: mximum witing time of bout 40 minutes nd queue length of five persons. When six sles consultnts re employed, the mximum witing time is 14 minutes nd the queue length decreses to mximum of three persons. The improvement inititive is worth serious considertion when viewed from customer perspective. Number of consultnts 3 4 5 6 Minimum queue length 0 0 0 0 Mximum queue length 7 5 3 3 verge queue length 0.7437 0.1747 0.031 0.011 Minimum witing time (hours) 0.00 0.00 0.00 0.00 Mximum witing time (hours) 1.43 0.7081 0.4234 0.2417 verge witing time (hours) 0.186 0.0433 0.0081 0.0025 Minimum time in system (hours) 0.1181 0.0961 0.0871 0.0871 Mximum time in system (hours) 2.0239 1.525 1.3608 1.0444 verge time in system (hours) 0.764 0.624 0.5871 0.5782 Tble 4: Results ssocited with incresing the number of sles consultnts 5.7.2 Testing the impct of improvement inititives on customer stisfction rtings Once the scenrio hs been simulted, the impct of the proposed improvement on customer stisfction should be determined. The stisfction rtings of the ssocited customer requirements my be improved due to the reltionship between the technicl benchmrks nd customer requirements. The reltionship my be described by using performnce/stisfction function defined by Rmswmy [13]. The increse in stisfction for ech requirement depends on the functionl form ssocited with the specific KPI nd customer requirement. performnce/stisfction function my be determined for the instore customer service process, bsed on the technicl benchmrk of the process with the ssocited stisfction rting. The performnce level of the KPI is mesured for ll three compnies, nd the trget nd the lowest perceived performnce re determined in section 4.6 for the in-store customer service process. The performnce level is shown in Figure 11 s the technicl benchmrk. The stisfction rting for the ssocited customer requirement is lso shown in Figure 11. The five performnce rtings, together with the five stisfction rtings for the KPI specified for the in-store customer service process, my be plotted grphiclly to determine the performnce/stisfction function with the lest squres regression method. The impct tht n improvement inititive my hve on customer stisfction is demonstrted in the function obtined for the in-store customer service process. 53
Testing the impct of employing more sles consultnts t Compny stores The performnce of the in-store customer service process cn be mesured by mesuring the performnce of the KPI defined for this process. In the previous section the performnce levels of the totl time spent witing in the queue ws mesured. The performnce level of this KPI for Compny ws improved from 60 minutes to 14 minutes. The figure below shows the functionl form ssocited with this KPI. Stisfction rting for time spent witing in the queue 100.00% 80.00% 60.00% 40.00% 20.00% 0.00% y = -0.0069x + 1.0083 0 20 40 60 80 100 120 140 Figure 12: Performnce/stisfction function for time spent witing in the queue From the lest squres method, the functionl form is represented by the following eqution: y = -0.0069x + 1.0083 (9) Using eqution (9), the stisfction rting (y) ssocited with witing 14 minutes in the queue is 91.1 percent. The improvement from 68 percent to 91 percent is significnt, nd Compny cn consider implementing this inititive; but they must consider the cost involved in ppointing three more sles consultnts before they mke their decision. 6. RECOMMENDTIONS ND FUTURE RESERCH Totl time spent witing in the queue (hours) Due to the unvilbility of relible dt, the ECEF is only prtilly vlidted in this pper. The lck of dditionl dt points leds to the use of only five dt points to obtin the performnce/stisfction function of the in-store customer service process. The use of only five dt points my cuse the performnce/stisfction function to be unrelible, nd due to limited dt the mesures of goodness of fit or the correltion fctor could not be determined. To mke the function more relible, customers my be surveyed to obtin forecst stisfction rtings for lterntive performnce levels. The dditionl dt points obtined from the survey my then be plotted together with the five vilble dt points to obtin more relible function. Lest squres regression is used in this pper to cpture the reltionship between process KPIs nd customer stisfction. It is ssumed tht the lest squre method will provide the function tht best fits the dt obtined. It is importnt, however, tht this ssumption be tested. It is recommended tht further empiricl reserch be performed to nlyse the impct of improvement inititives on customer experience. The relibility of the ECEF my be tested through the ppliction of the frmework in different service-oriented environment. It is recommended tht the ECEF is pplied in n environment such s the insurnce or bnking industry to test the relibility of the frmework. 54
7. CONCLUSION 7.1 criticl nlysis of the dvntges nd disdvntges of the ECEF From the ppliction of the frmework, it is concluded tht there re mny dvntges ssocited with the ECEF, mking it n pproprite frmework for enhncing the customer experience. The use of QFD in the frmework cn ssist mngers with the following [14]: To define product specifictions which meet the customers requirements, while pying ttention to the competitors. To ensure consistency between the customers requirements nd the mesurble chrcteristics of the product or service. To inform nd convince ll those responsible for vrious stges of the process of the reltionship between the qulity of the output of ech phse nd the qulity of the finished product or service. To ensure consistency between the plnning nd opertionl processes of the compny. To help minimise mistken interprettions of priorities nd objectives becuse plnning tkes plce t n erlier stge. To trnslte customer requirements into meningful (technicl) requirements t ech stge of the development nd production processes. To bring people together from vrious disciplines nd fcilitte the formtion of tems cpble of meeting customer requirements. By integrting the methodology of BPR into the frmework, compnies will be ble to redesign their business processes by excising non-vlue dding work. BPR helps to mximise customer vlue while minimising the consumption of resources required to deliver their product or service [6]. The redesign of business processes is tested through simultion models, which leds to more dvntges, such s these: Mngers re ble to study the dynmic behviour of business processes. The results obtined re ccurte compred with nlyticl models. Simultion is ble to provide quntittive estimtes of the impct tht process redesign is likely to hve on key performnce mesures. Simultion mkes it esy to perform wht-if nlysis to determine the best improvement inititive. lthough the integrtion of these three tools llows the ECEF to enjoy the combined dvntges ssocited with these tools, it lso leds to certin disdvntges tht my be ssocited with the ECEF. Disdvntges ssocited with using QFD include these [14]: QFD is dependent on ccurte customer dt; if relevnt secondry informtion does not exist, surveys must be conducted. These surveys my be very expensive to conduct in time nd effort. QFD is dependent on the commitment of key people nd senior mngement to contribute to the input. If their commitment cnnot be obtined nd mintined, the technique my hve seriously negtive strtegic implictions. QFD is very dependent on resources such s people, time, nd finnces. QFD nd BPR must be integrted grdully, nd resistnce to chnge must be tken into ccount. Employees concerns must be mnged crefully. Other disdvntges must be considered, such s these: Implementing the ECEF my led to higher demnds on employees; so employees must be motivted to do whtever it tkes to stisfy the needs of the customers. It my be expensive to build simultion models to test the impct of process improvements. It my sometimes be difficult to interpret the results yielded by the simultion models. 55
56 It my not lwys be possible to obtin relible dt s input for simultion models, nd this cn led to inccurte results tht my hve significnt negtive impct on customer experience. 7.2 The vlue of the ECEF in the telecommunictions industry Due to the strict competition fced by the telecommunictions industry in South fric, nd the low differentition between the products they offer, they hve hd to invest in delivering qulity service to their customers. Exceptionl customer service my serve s differentitor between the existing telecommunictions compnies. In this pper, the ECEF is pplied to identify nd prioritise key business processes nd to lign the highest priority process to the needs of the customer. The ECEF is effectively utilised to identify fesible improvement inititive nd to test the impct of the inititive on customer experience. The seven stges outlined in the ECEF re vlidted through the results obtined in this pper. The results prove tht the customer stisfction of Compny customers my be incresed significntly through the ppliction of the frmework. This pper vlidtes the usefulness of the ECEF in helping mngers to stisfy customer needs. The ECEF my be used s vluble frmework for redesigning business processes to deliver products nd services ccording to the needs of the customer. REFERENCES [1] Bnd, C., Pmpllis,., Vn der Wl, R.W.E. 2002. Service qulity in cellulr telecommunictions compny: South fricn experience, Mnging Service Qulity, vol. 12, no. 5, pp. 323-335. [2] Bnks, J., Nelson, B.L. & Nicol, D.M. 2010. Discrete-event system simultion. 5 th ed., Person Eduction, Prentice Hll. [3] Bitrn, G. 1993, Frmework for nlysing the qulity of the customer interfce, Europen Mngement Journl, Vol. 11, No. 4, pp. 385-396. [4] Brighton Webs Ltd. Dt & nlysis services for industry & eduction. http://www.brightonwebs.co.uk/distributions/tringulr.sp. Visited on 12 pril 2012. [5] Both, G.J., Kruger, P.S. & De Vries, M. 2010. Enhncing customer experience through business process improvement: n introduction to the Enhnced Customer Experience Frmework (ECEF). CIE40 Conference proceedings, Jpn 2010. [6] Both, G.J. & Vn Rensburg,. 2010. Proposed business process improvement model with integrted customer experience mngement. South fricn Journl of Industril Engineering, Vol. 21(1), pp. 45 57. [7] Cmp, R. 1995. Business Process Benchmrking: Finding nd Implementing Best Prctices. Productivity Inc. SQC Qulity Press. [8] Chse, R.B. et l. 2006. Opertions mngement for competitive dvntge with globl cses. New York, McGrw-Hill/Irwin. [9] Dvenport, T.H. & Short, J.E. 1990. The new industril engineering informtion technology nd business process redesign, Slon Mngement Review, pp. 11-27. [10] Gustfsson,. & Johnson, M.D. 2000. Improving customer stisfction, loylty, nd profit: n integrted mesurement nd mngement system, Sn Frncisco, Jossey-Bss Inc. [11] Jin, S.C. 1997. Mrketing plnning nd strtegy, 5 th ed., South-Western College Publictions, Cincinnti, OH. [12] Jmes, L.H. et l. 1994. Putting the Service-Profit-Chin to work, Hrvrd Business Review Mrch-pril 1994. [13] Rmswmy, R. 1996. Design nd mngement of service processes: Keeping customers for life. New York: ddison Wesley. [14] Youssef, M.. & Ziri, M. 1995. min pillr for successful totl qulity mngement nd product development, Interntionl Journl of Qulity & Relibility Mngement, vol 12, no. 6, pp. 9-23.