strategic management of logistics service: a fuzzy qfd approach
TRANSCRIPT
ARTICLE IN PRESS
0925-5273/$ - see
doi:10.1016/j.ijp
�Correspondifax: +39521 90
E-mail addre
Int. J. Production Economics 103 (2006) 585–599
www.elsevier.com/locate/ijpe
Strategic management of logistics service:A fuzzy QFD approach
Eleonora Bottani�, Antonio Rizzi
Department of Industrial Engineering, viale delle Scienze 181/A, Campus Universitario, University of Parma, 43100 Parma, Italy
Received 20 April 2005; accepted 15 November 2005
Available online 3 March 2006
Abstract
Logistics and Supply Chain Management literature indicates that customer service management has become a strategic
issue for companies in the new millennium. By improving logistics performances, companies increase customer satisfaction
and gain market shares.
The aim of this paper is to propose an original approach for the management of customer service. The approach is based
on the quality function deployment (QFD), a methodology which has been successfully adopted in new products
development. Specifically, the paper addresses the issue of how to deploy the house of quality (HOQ) to effectively and
efficiently improve logistics processes and thus customer satisfaction. Fuzzy logic is also adopted to deal with the ill-defined
nature of the qualitative linguistic judgments required in the proposed HOQ.
The methodology has been tested by means of a real case application, which refers to an Italian company operating in
the mechanical industry.
r 2006 Elsevier B.V. All rights reserved.
Keywords: Logistics service; Customer service management; Fuzzy QFD; House of quality
1. Introduction
The importance of customer service as a strategicissue has emerged in particular during the lastdecade, due to a twofold reason. First, brand powerhas progressively decreased, making products al-most undifferentiated in terms of trademarks.Customers do not rely on brands except in a fewmarket niches, such as fashion. Second, due totechnology diffusion, the functionalities and tech-nological features of products tend to be the same(Franceschini and Rossetto, 1997). Today, new
front matter r 2006 Elsevier B.V. All rights reserved
e.2005.11.006
ng author. Tel.: +39521 905872;
5705.
ss: [email protected] (E. Bottani).
customers cannot be acquired counting only onbrands or on technical characteristics of products.On the contrary, the breath of logistics servicesrelated to products may play a significant role in thecompetitive scenario (Vandermerwe and Rada,1988; Bailey, 1996). As a consequence of this shifttowards service, customers have become more andmore exacting about logistics performances (Lee-Kelley et al., 2002).
Customer service, hereafter understood to be theservice performance perceived by customers as aresult of logistics processes and activities, has beenwidely recognized as a mean to gain competitiveadvantage. Through customer satisfaction, compa-nies retain their customers and gain new mar-ket shares (Zeithaml, 1988; Christopher, 1998).
.
ARTICLE IN PRESSE. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599586
However, even though new customers are welcomedalmost in every line of business, the main objectiveof companies is to maintain customers for a long-time period. The total value of a lifetime customer isalmost unquantifiable, and allows firms to achieve acompetitive advantage against competitors (Chris-topher, 1998; Bailey, 1996). Bailey (1996), stressesthe significant role of service quality in achievingcompetitive advantage, and, conversely, the weakimportance of sales and profits.
According to the so-called ‘‘disconfirmationparadigm’’ (Philip and Hazlett, 1996; Zeithaml etal., 1990), customer satisfaction is achieved whenlogistics performances delivered by the supply chainmeet customer requirements. To this extent, Ro-bledo (2001), states that customers evaluate serviceby comparing their perceptions of the servicereceived with their expectations; thus, the gapbetween customer expectations and perceptions isa synthetic measure of customer satisfaction. Sincecustomers will be satisfied when perceptions exceedtheir expectations, understanding these require-ments is an imperative for firms.
In addition, when speaking about service man-agement, a dynamic perspective should be adopted.Customer service is not a steady concept, but iscontinually in a state of change, and evolvesthrough a continuous improvement cycle (Morris,1996; Baines, 1996). Therefore, the quantitativemeasure of logistics performances delivered andexpected has to be repeated over time, periodicallyauditing gaps between expectations and percep-tions. When a lack of correspondence occurs, viablelogistics areas and factors of intervention have to beidentified, pondered and ranked in terms ofefficiency and effectiveness. Since interventionsimply costs, before taking steps toward implemen-tation, a costs/benefits analysis is appropriate, inorder to undertake actions starting from thosefactors with the highest impact on customer service.
To conclude, providing logistics service whichmeets customer expectations is a continuous pro-cess, which can be summarized in the followingsteps:
�
understanding the customer’s voice, that isrequirements and expectations in terms ofrelevant logistics performances; � assessing customer’s service perception; � if a gap between perception and requirementsoccurs, identifying viable steps that can be imple-mented to improve customer satisfaction;
�
identifying costs and benefits related to each step;and � implementing the most efficient actions forcustomer satisfaction by means of a cost/benefitanalysis.
A similar process is followed in new productsdevelopment, where customer requirements have tobe engineered into products features. The qualityfunction deployment (QFD) methodology has beenfound as a viable tool which can be successfullyapplied for this purpose (Akao, 1990). QFD hasbeen defined by the American Supplier Institute as‘‘A system for translating consumer requirements into
appropriate company requirements at each stage from
research and product development to engineering and
manufacturing to marketing/sales and distribution’’.As detailed in the next section, by assessing howeach ‘‘how’’ (engineering characteristics) impacts oneach ‘‘what’’ (customer requirements), QFD makesit possible to rank ‘‘hows’’ in terms of efficiency toreach the required ‘‘whats’’.
A preliminary review of the literature has high-lighted only few references where QFD has beenassociated to service assessment, none of which canbe directly related to logistics issues. Lapidus andSchibrowsky (1994), illustrate the QFD applicabil-ity as a method for improving service starting fromcustomer complaints. In their approach, customercomplaints become the ‘‘whats’’ to be considered inthe house of quality (HOQ). Conversely, wepropose a proactive approach to be adopted beforecomplaints occur: thus, ‘‘whats’’ do not emergefrom complaints but from logistics and supply chainmanagement literature.
Behara and Chase (1993), illustrate the QFDprocess in matching customer requirements tospecific topic areas in service management. How-ever, these applications do not provide a generalmethodology to plan and manage the trade-offs andcorrelations associated with customer requirementsand firm viable actions.
Stuart and Tax (1996), propose the QFDapplication to manage the service design phase.They suggest the use of HOQ as an effective meanto plan processes for a successful execution ofservices. Their approach is of general purpose anddepicts the general traits of a QFD approach todesign service strategies. However, the authors donot detail how the approach may be deployed for apractical in-field application. In conclusion, theworks cited above deal with service management
ARTICLE IN PRESSE. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 587
under a general perspective, and do not focus theapproach on service performances which stem fromlogistics processes and activities.
Starting from the work of Stuart and Tax (1996),we develop a tool suitable to be adopted in thelogistics context. Moreover, one of our mainobjectives is to introduce a methodology that couldbe directly adopted by practitioners in the logisticsfield. A cost/benefit analysis is also introduced toidentify and rank the most efficient steps towardimprovement of logistics processes and customersatisfaction. A fuzzy approach is adopted sincethe methodology mainly relays on qualitativejudgments given by panel of experts and bycustomers.
The remainder of the paper is organized asfollows. In the next paragraph, after a briefdescription of the QFD methodology, the approachproposed is detailed. Then, a real case application ofthe QFD methodology is presented. The case refersto an Italian company operating in the mechanicalindustry. Concluding remarks are finally presented.
2. The fuzzy QFD approach
2.1. QFD fundamentals
Quality function deployment originated in 1972in Japan, as a methodology to be adopted toimprove products quality in Japanese firms, such asMitsubishi, Toyota and their suppliers (Hauser andClausing, 1988). QFD methodology has introduceda twofold innovation in traditional product devel-opment processes. First, the application of QFDrequires the careful consideration of customerduring the development process (Akao, 1990).Second, the QFD approach has introduced thecollaboration among different business areas as aprerequisite for product design. This is obtained bysetting up appropriate work groups, whose mem-bers belong to different business units involved inthe product design phase (Bouchereau and Row-lands, 2000).
Two main QFD approaches to production devel-opment emerge from literature analysis (Choen,1995), namely the ‘‘matrix of matrices’’ and the‘‘four-phases model’’. In this paper, we focused onthe four-phases approach to product development,whose steps have been thoroughly described byHauser and Clausing (1988) and Bouchereau andRowlands (2000).
This approach is composed of four successivematrices (the customer requirement planning ma-trix, the product characteristics deployment matrix,the process and quality control matrix and theoperative instruction matrix), which are applied inas many phases of the product design process. Wefocused on the customer requirement planningmatrix which has been used to develop the modelfor strategic customer service management.
The customer requirements planning matrix, alsocalled the ‘‘House Of Quality’’ because of its typicalshape, is the first step in investigating customer’sneeds and requirements. It is composed of two mainparts, related to customer’s requirements (‘‘what’’customer needs) and technical elements (‘‘how’’ theproduct has to be made) respectively. The HOQ isthus adopted by the design work group to transformthe customer’s requirements and needs into productcharacteristics.
The HOQ can be built by following an eight stepsprocess. At the beginning of the process, customer’sneeds and requirements have to be identified.According to Hauser and Clausing (1988), thoseelements are also called ‘‘customer’s attributes’’(CAs), and are generally known as a result ofsurveys or direct questions to customers. CAs arelisted in row in the HOQ; if necessary, they can begrouped into sets that express similar expectations(step 1).
Customer’s attributes are weighted in order toexpress their relative importance. The weight ofeach CA is inserted in a column in the matrix(step 2).
Next, firms have to establish how their productsperform against those of competitors. Generally,the evaluation of a firm product is carried out bydirectly asking customers how products/services arerated in relation to the competition. Benchmarkinganalysis can also aid in this evaluation. The resultsof this step are thus added in a column in the rightside of the matrix (step 3).
In order to develop a new product, CAs must betranslated into ‘‘engineering characteristics’’ (ECs)that probably affect one or more CAs. Engineeringcharacteristics are measurable attributes concerninga firm’s product or service; they are listed incolumns in the HOQ (step 4).
The core element of the matrix is the ‘‘relation-ships matrix’’. In order to complete this part of theHOQ, the relationships between customer’s needsand firm’s ability to meet those needs have to bedetermined. The relationships are expressed with
ARTICLE IN PRESS
Relationships matrix (Rij)
Rel
ativ
e im
port
ance
of C
As
(Wi)
Cus
tom
er’s
attr
ibut
es(C
As)
Correlations matrix
Engineering characteristics (ECs)
Absolute importance of ECs (AIj)
Relative importance of ECs (RIj)
Technical analysis of competitors
Ben
chm
ark
anal
ysis
Target values of ECs
Fig. 1. The house of quality.
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599588
graphic symbols that indicate how and to whatextent each engineering characteristic meets eachcustomer’s attribute. Usually, symbols express threedegrees of strength (weak, medium, strong), whichare translated in an appropriate rating scale, such as1-3-9 or 1-5-9. In general, no specific justificationsshould be provided for the choice of the adoptedrating scale. Absence of symbols means absence ofrelationships (step 5).
In a similar manner, the top side of the HOQ,called the ‘‘correlations matrix’’, is then filled in,expressing how ECs affect each other. A positiverelationship indicates that two ECs can complementor improve each other, while a negative one suggeststhat trade offs are required. Correlations areindicated with graphic symbols that express thedegree of relation between ECs. Symbols are thentranslated into a four-value rating scale (strongnegative, negative, positive, strong positive), such as1-3-7-9 or 1-3-5-9. Again, it is possible to have nocorrelations between ECs (step 6).
Moreover, firm’s products are compared with thoseof competitors. To this extent, the work group carriesout a quantitative benchmark analysis of competitors’engineering characteristics. The results are added in arow in the lower part of the matrix (step 7).
Finally, firms have to introduce a target measurefor each EC in the matrix. The target measuretranslates customer’s expectations into numericalvalues, in order to quantitatively assess firm’sperformances against customer’s requirements.The lower part of the HOQ is therefore completedintroducing the goal measure of each EC (step 8).
The typical structure of the HOQ is shown inFig. 1.
The result of the matrix is the ranking of ECs indescending order of importance. To this extent,either the absolute and/or the relative importance ofeach EC against customer’s requirements have to bequantitatively evaluated.
As stated above, in the traditional QFD applica-tions the generic position Rij in the relationshipsmatrix expresses the relationship between the ithCA and the jth EC with a numeric scale. Therefore,the absolute importance AIj, j ¼ 1,ym of each ECcan be calculated as
AIj ¼Xn
i¼1
W iRij ; j ¼ 1; . . . ;m (1)
being Wi the relative importance of the ith CA, Rij
the numerical value added to the position (i,j) of the
matrix, j ¼ 1,y,m and i ¼ 1,y,n the number ofECs and of CAs respectively.
The relative importance RIj can be derived fromthe absolute importance AIj, through the followingequation:
RIj ¼AIjPmj¼1AIj
; j ¼ 1; . . . ;m. (2)
Literature analysis has pointed out that engineer-ing characteristics are usually ranked based on RIj
rather than on AIj. Thus, the higher the RIj, themore important the engineering characteristic thatshould be incorporated into the product in order toimprove customer satisfaction.
2.2. The proposed methodology
The approach proposed is based on the transla-tion of HOQ principles from product developmentfield to logistics service management. While thetraditional HOQ correlates customer requirements(‘‘whats’’) with engineering characteristics of newproduct under development (‘‘hows’’), in ourapproach customer service requirements in termsof logistics performances (‘‘whats’’) are crossed overwith viable strategic actions, either technical (suchas the adoption of a more performing technology)or managerial (i.e. a reorganization of processes inthe supply chain), that could be undertaken by thefirm’s top management to improve logistics pro-cesses (‘‘hows’’).
ARTICLE IN PRESS
Relationships matrix (Rij)Wha
ts
Serv
ice
Fact
ors
(SF)
Correlations matrix (Tkj)
Hows
Strategic Actions (SAs)
Relative importance of SAs (RIj)
Real importance of SAs (RIj*)
Cost for implementation of SAs (Cj)
Rea
l im
port
ance
of
CFs
(W
i* )
Rel
ativ
e im
port
ance
of
CFs
(W
i)
Benefits of SAs (Uj)
Fig. 2. The house of quality for the strategic management of the logistics service.
Table 1
List of viable indicators for the evaluation of logistics service,
adapted from Franceschini and Rafele (2000)
Service factors
‘‘whats’’
Description
Lead-time Time period elapsing from customer’s
order until receipt
Regularity The dispersion around the mean value of
the delivered lead-time
Reliability Capability to deliver orders within the due
date
Completeness Capability to deliver full orders when
required
Flexibility Capability to modify orders in terms of
due date and quantity when required
Correctness Avoidance of mistakes in orders delivered
Harmfulness Avoidance of damages in orders delivered
Productivity Number of item produced in a given time
period
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 589
The resulting customer service HOQ is shown inFig. 2.
As can be seen from the table, ‘‘whats’’ elementsexpress service factors SFi, i ¼ 1; . . . ; n affectinglogistics service perception. These factors have beenextensively described by logistics and supply chainmanagement literature. The reader will find acomprehensive list of the main criteria that can beused for the evaluation of the logistics service inFranceschini and Rafele (2000). For the sake ofclarity, the service factors proposed by the authorsare shown in Table 1, together with a briefdescription.
Obviously, service factors listed in table do notprovide an absolute description of all viable factorsthat could be considered when perceived service hasto be assessed. Depending on particular circum-stances, factors could be either added or removed.However, our paper strives to introduce a newmethodology for customer service managementrather than to formalize an exhaustive frameworkof factors affecting customer service perception,which have been widely addressed by logistics andSCM literature.
Once customer service has been assessed, viablestrategic actions SAj, j ¼ 1; . . . ;m the firm can
undertake in the logistics field to improve serviceperformances have to be identified and ranked interms of both effectiveness and efficiency withregard to customer service improvement. Thoseactions correspond to ‘‘hows’’ in the proposedcustomer service HOQ. A list of possible ‘‘hows’’
ARTICLE IN PRESS
Table 2
List of viable strategic actions
Strategic actions
‘‘hows’’
Description
Just-In-Time
philosophy
JIT is a philosophy originated in the
manufacturing sector which has been
extended to supply chain management. It
helps to streamline the logistics pipeline
through the efficient flow of materials and
information, i.e. by providing the right
materials, in the right quantities and
quality, in the right place at the right time.
Warehouse
management
optimisation
The efficiency and the effectiveness of the
logistics flows are deeply affected by
optimized warehouse & distribution
centres management policies. Shipping &
receiving, storage, picking activities can
largely benefit from ad hoc optimization
tools.
Transport
management
Transportation has been recognized as a
paramount factor affecting effectiveness
and efficiency of logistics processes.
Through transportations, the product
value is increased by making it available
where and when it is required. However,
transports add significant costs, which
could jeopardize the profitability of the
supply chain. Therefore specific
optimization tools can be considered as
viable actions to improve logistics
performances perceived by customers.
Information
technology
Information technology is generic term
used to include hardware, software and
networking technologies, such as servers,
computer networks, expert systems,
software for communication, such as
Enterprise Requirement Planning (ERP),
Electronic Data Interchange (EDI), etc.
All these tools play a significant role in
synchronizing the flow of goods with the
flow of information, which affects the
logistics performance of the supply chain.
Demand forecasting
methods
Accurate forecasting methods make it
possible to match supply and demand,
smoothing uncertainty, reducing safety
stocks and stock outs. The set up of
collaborative programs, such as CPFR,
VMI, or consignment aimed at reducing
uncertainties may be encompassed in this
category.
Other Depending on the particular
circumstance, other strategic actions in
the logistics field may be considered.
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599590
when customer service performance related tologistics activities has to be improved is shown inTable 2.
Again, Table 2 does not provide a definitiveframework of all the options available, whichdepend on the particular circumstances.
The roof of correlations, the weights Wi [n� 1],the relationships matrix Rij [n�m] and the relativeimportance of SAs vector RIj [1�m] complete theHOQ. It is worth stressing that the weight vector,the correlations matrix and the relationships matrixtranslate linguistic judgments given by humanbeings. Therefore, an effective mean to deal withthem would appear to be fuzzy logic.
The fuzzy set theory was originally introduced byZadeh (1965) to deal with ill-defined problems,characterized by a certain degree of uncertainty andvagueness. The main advantage of the adoption offuzzy logic is the opportunity to express ill-definedjudgments, such as the impact of a SA on a SF.Moreover, the use of fuzzy numbers becomes veryimportant in decision-making problems, wherelinguistic scales are adopted and where a panel ofdecision makers (DMs) is involved in the judgmentprocess. To this extent, fuzzy numbers make itpossible to reproduce the typical subjective way ofthinking of human beings. As an example, let usassume a firm scores 90% in fill rate. In a crispapproach, this value will match one and only onelinguistic value, that is there is a biunivocalrelationship between performances and judgments.Therefore, a performance judgment ‘‘very high’’under ‘‘fill rate’’ factor means a score of 90% andonly 90% for every DM, even though DMs mayhave a different perception of ‘‘very high’’. Con-versely, in a fuzzy approach, DMs may meandifferent performance with the same value, i.e. ifthe firm scores 90% in fill rate, its serviceperformance against fill rate parameter could beconsidered as ‘‘very high’’ by a customer, and‘‘high’’ by another one, with a certain degree ofmembership. This implies that the value ‘‘90%’’should belong to the two categories ‘‘very high’’ and‘‘high’’ at the same time and to a different extent.The degree of membership is assessed through therespective membership functions.
To conclude, fuzzy logic allows to take intoaccount the different meaning that we may give tothe same linguistic expression. As a matter of fact,this is why the fuzzy approach has been so widelyadopted in different research fields, as witnessed bythe massive literature on the subject. The reader
ARTICLE IN PRESSE. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 591
may refer to Zadeh (1965), and Zimmermann (1991),for a complete description of fuzzy numbers andrelated algebra.
From now on, unless specified, all parametersshould be thought of as fuzzy triangular numbers.In our approach, four new fuzzy elements have beenadded to the traditional HOQ, namely:
�
the weighted importance of service factors; � the weighted importance of strategic actions; � the cost for the implementation of strategicactions; and
� the marginal benefit of strategic actions.These elements, as well as their role in rankingSAs, are detailed below.
2.2.1. Weighted importance of service factors
The weighted importance Wi* of SFs is a [n� 1]vector which expresses the real importance of eachSF. The introduction of Wi* is required to weighteach service factor considering not only theimportance the customer gives it, which is expressedby the value Wi, but also the performance deliveredby the firm for that factor. To gain a competitiveadvantage, the firm must provide superior service tothe customers on critical service factors, that iseither those that are perceived as the mostimportant ones or where service perceived isinferior. Conversely, improving service either for afactor whose importance is trivial or where the firmalready delivers a superior service is useless.
The weighted importance Wi* is computed byassessing the distance di between firm performanceand that which is perceived by customers assuperior, the latter being the performance thatallows the firm to achieve customer satisfaction.Both the performance delivered and the targetsuperior value could be retrieved from customerservice surveys by asking the customer directly.Since both performance values are fuzzy, a distancebetween fuzzy numbers has to be assessed. To thisextent, the Hamming procedure is suggested to beadopted (Chien and Tsai, 2000). This procedureidentifies the distance between two fuzzy numbers asthe distance between the centres of gravity of therespective membership functions. From a mathe-matical point of view, given two fuzzy sets A and B,the Hamming distance dðmAðxÞ;mBðxÞÞ between twofuzzy numbers belonging to A and B respectively,
can be computed as
dðmAðxÞ; mBðxÞÞ ¼
ZX
mAðxÞ � mBðxÞ�� ��dx, (3)
where X is the universe of discourse. Due to thecalculation method, the resulting Hamming distanceis a crisp value.
The di parameters are then calculated accordingto Eq. (3). Then, the weighted importance Wi* ofSFs can be derived as follows:
W �i ¼ di �W i; i ¼ 1; . . . ; n. (4)
2.2.2. Weighted importance of strategic actions
This element strives to determine which strategicaction has the highest impact on customer satisfac-tion. It takes into account the weighted importanceof service factors, the relationships matrix and thecorrelations matrix.
As already detailed, the generic position Rij in therelationships matrix expresses the relationshipbetween the jth SA with the ith SF. Again, afuzzy linguistic scale may be usefully adopted byDMs to interpret the vagueness and incompleteunderstanding of the relationships between ‘‘hows’’and ‘‘whats’’.
The importance RIj of each strategic actioncan then be calculated applying the followingequation:
RIj ¼Xn
i¼1
W �i � Rij ; j ¼ 1; . . . ;m, (5)
where Wi* is the fuzzy weighted importance ofith service factor, while Rij is the fuzzy numberexpressing the impact of the jth SA versus the ithSF.
In a similar manner, the generic position Tkj, j,k ¼ 1,ym, k 6¼j, in the correlations matrix expressesthe correlation between the kth and the jth ‘‘hows’’.In order to quantitatively ponder the correlationbetween ‘‘hows’’, we adopt the approach of Tang etal. (2002). According to the authors, the correlationTkj can be interpreted as the incremental changes ofthe degree of attainment of the jth ‘‘how’’ when theattainment of the kth one is unitary increased.Using this definition, the weighted importance RIj*can be computed as follows:
RI�j ¼ RIj �Xk¼j
Tkj �RIk; j ¼ 1; . . . ;m. (6)
ARTICLE IN PRESSE. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599592
2.2.3. Cost and marginal benefit of strategic actions
In order to complete the assessment and rankingof strategic actions, their cost of implementationshould be considered. In this situation fuzzy logicbecomes a fundamental tool in dealing with ill-defined issues such as the evaluation of costs. Whilea DM may find objective difficulties in quantita-tively assessing the costs of implementation ofstrategic actions, he/she can more easily give ajudgment on a linguistic scale, ranging for instancefrom Very High to Very Low. This is why, in thelower part of the HOQ a fuzzy parameter Cj hasbeen added to ponder the cost of implementing thejth strategic action.
The marginal benefit Uj of strategic actions can becalculated through the ratio between benefits andcosts, as expressed by the following equation:
Uj ¼ RI�j �1
Cj
; j ¼ 1; . . . ;m. (7)
Since both RIj* and Cj parameters are fuzzynumbers, Eq. (7) describes an operation betweenfuzzy numbers; the resulting Uj is thus a fuzzynumber. In order to make SAs comparable andrank the results, defuzzified values should becomputed. Due to its simplicity, the Yager method(Yager, 1981) is suggested as a viable tool to adoptin order to obtain final crisp marginal benefits.Starting from a fuzzy triangular number a(l,m,u),the defuzzified value is computed as
l þ 2mþ u
4. (8)
Once crisp values have been computed, SAs canbe finally ranked. In particular, according toTrappey et al. (1996), the greater the crisp Uj
parameter, the higher the implementation priorityof the corresponding strategic action. Strategicaction which scores the highest is the one whichhas the highest impact on customer service, andtherefore whose implementation should be consid-ered by the firm top management to improve thelogistics performance.
3. Application of the methodology
In this paragraph, the methodology developed isapplied to a real industrial case, which refers to amajor Italian company operating in the mechanicalindustry.
The main objective of the application is twofold.On the one hand, it is aimed at assessing its
robustness and consistency, where robustness andconsistency are respectively understood to berelated to the applicability of the methodologyand to the reliability of the result obtained. On theother hand, the application strives to considerpractical implications in managing customer servicethrough a QFD approach.
3.1. The company
Since it was established in 1973 in Northern Italy,the firm has been acquiring specific experience indesigning and manufacturing special piping compo-nents. Specific expertise has promoted high-levelspecialization in stainless and carbon steel manu-facturing, making the firm an industry leader in thedomestic market of special piping parts. Productsare used as components in assembly lines. Theautomobile and mechanical industries are one of themost important markets for the firm sales, account-ing for almost 60% of the company’s revenue. Therest of the turnover comes mainly from sales toappliances industry and process plants. Otherrelevant figures for 2001 are: 450 employees, 45millions Euro of aggregate turnover, about 35,000tons of steel are processed every year, while everyday 40 km of welded pipes of stainless and carbonsteel are shipped from warehouses directly to theassembly lines of the buyers.
The main customers of the firm are majormanufacturers which have recently set up programsto streamline the supply processes. Buyers havebeen requiring adequate logistics performancesfrom their suppliers to reduce inventory, avoidcontrol of orders accuracy and turn the supplyingprocess from a traditional approach to a JIT one.As a consequence, the firm has been asked not onlyfor remarkable products from a technical point ofview, but also for remarkable logistics perfor-mances, basically in terms of lead time, reliabilityand accuracy of shipments. Operating in a verycompetitive scenario from a logistics point of view,the firm needs to proactively manage customerservice to retain its customers and gain new marketshares. To this extent the QFD approach proposedin this paper has been recognized by the companytop management as a valid tool to control logisticsperformances and promptly tune service deliveredto match customer requirements.
Practical consequences of the application of theQFD tool were expected in the assessment of serviceprovided by the firm and in the evaluation of the
ARTICLE IN PRESS
Table 4
Service factors considered in the real case application
Service factors Description
Lead-time Time period elapsing from customer’s
order until receipt.
Flexibility Capability to modify orders in terms of
due date and quantity when required.
Accuracy Avoidance of mistakes and damages in
orders delivered.
Reliability Capability to deliver orders within the due
date.
Fill rate Common indicator of customer service
performance related to inventory. It can
be defined as the percentage of units
available when requested by customers.
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 593
most efficient actions that the management couldundertake to improve logistics performance andcustomer satisfaction.
A workgroup has been set up to apply themethodology. The team was headed by academi-cians from the Department of Industrial Engineer-ing of the University of Parma, and included firmexecutives reporting directly to logistics, IT, andproduction chief officers. The group has worked for4 months, mainly through roundtable discussions,where the main elements of the HOQ and theirmutual relationships have been defined. In thefollowing paragraph the main results obtained arepresented.
Frequency Number of deliveries accomplished in a
given time period.
Organization
accessibility
Customer’s opportunity to establish a
contact with firm’s staff. Usually,
customers need to contact firms because
they have some questions to be answered,
they need firm’s actions, they have some
problems with products, they want to
submit complaints.
Complaints
management
Process subsequent to the recognition of
some errors in service provided, that
allows service quality standards to be
reestablished.
3.2. Results and discussion
When applying the proposed HOQ to the realcase, appropriate ‘‘whats’’ have to be identified. Tothis extent, four firm main buyers were asked totake part in the application. In the following, theywill be indicated as C1, C2, C3, C4.
First of all, the importance of each customer hasbeen weighted through the percentage of profitmargin generated, as shown in Table 3.
The main service factors ‘‘whats’’ to be consid-ered in the real case application have emerged froma preliminary survey phase, which has beenperformed through direct interviews carried out byacademicians with the customers involved in theproject. A survey has been adopted because itemerged as one of the most efficient and effectiveways to ponder the performance perceived for eachfactor affecting customer satisfaction (see Kelleret al., 2002).
The relevant logistics ‘‘whats’’ are shown inTable 4, together with a brief description.
The second part of the application focused on theassessment of viable SAs ‘‘hows’’, their mutualcorrelations, as well as of the relationships judg-ments between SAs and customer SFs. The work-group agreed to adopt a linguistic approach,
Table 3
Importance ranking of the firm’s main customers
Customer Importance (%) Importance judgment
C1 0.35 Very high
C2 0.30 Very high
C3 0.25 High
C4 0.10 High
therefore a first instructive phase was required tointroduce the workgroup members to fuzzy settheory and fuzzy logic. In a similar manner,appropriate linguistic scales were set up for theevaluation of relative and weighted importance ofSFs, the relative and weighted importance of SAs,the costs for the implementation of SAs, togetherwith values in the relationships and correlationsmatrixes.
Strategic actions ‘‘hows’’ were identified basedboth on literature analysis and the firm character-istics, whose peculiarities have emerged from round-table discussions. Results are shown in Table 5 witha brief description for each point.
During the survey phase, the four customers havealso been asked about the importance of servicefactors, in order to determine the relative impor-tance of service factors, as well as to assess thedistance between the service delivered for eachfactor and the performance that is perceived assuperior. The four customers have been asked torank the relative importance of each SF on a 4-pointlinguistic rating scale, ranging from VL (Very Low)to VH (Very High). The fuzzy scale is shown inTable 6.
ARTICLE IN PRESS
Table 6
Linguistic judgments and corresponding fuzzy numbers
Judgment Fuzzy number
Very high (VH) (0.7; 1; 1)
High (H) (0.5; 0.7; 1)
Low (L) (0; 0.3; 0.5)
Very low (L) (0; 0; 0.3)
Table 5
Strategic actions considered in the real case application
Strategic actions Description
Just-In-Time
philosophy
JIT is a philosophy originated in the
manufacturing sector. It helps to smooth
the production process through the
efficient flow of materials, i.e. by
providing the right materials, in the right
quantities and quality, Just-In-Time for
production.
Order picking
optimization
Order picking is the activity by which a
number of goods are retrieved from a
warehousing system to satisfy a number
of customer orders.
Information
technology
Information technology is generic term
used to include hardware, software and
networking technologies, such as servers,
computer networks, expert systems,
software for communication, such as
Enterprise Requirement Planning (ERP),
Electronic Data Interchange (EDI), etc.
Demand forecasting
methods
Forecasting methods are tools that aim at
foreknow customers’ demand, in order to
reduce its uncertainty.
Customer
relationship
management
CRM is a generic term which
encompasses methodologies, software,
and Internet capabilities that help the firm
to manage customer relationships in an
organized way.
Warehouses lay-out
optimization
Warehouses lay-out embrace the optimal
assignment of items to storage locations,
the arrangement of the functional areas of
the warehouse, the number and location
of docks and input/output (I/O) points,
the number of aisles, etc.
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599594
wi,x is the fuzzy triangular number which isadopted to translate the linguistic importancejudgment given to the ith SF by the xth customer.wi,x fuzzy numbers have been pooled to determinean aggregate value to be used in the HOQ, that isthe relative importance Wi previously defined. Tothis extent, the relative importance Wi of servicefactor ith can be computed as a weighted average ofwi,x, being weight the importance of customers. The
weighted average takes into account the issue thatnot all customers are equal: being resources limited,the firm should tend to provide best class service forthose factors which are important for key custo-mers. In the specific case, the following equation isapplied.
W i ¼X4x¼1
Ix � wi;x; i ¼ 1; . . . ; n; (9)
being Ix the importance of xth customer surveyed(x ¼ 1,y,4).
Based on values shown in Table 3, the workgroup has expressed a fuzzy importance judgmentusing the same 4-point linguistic scale. The resultingfuzzy numbers have been used in the computationof Wi. Results are shown in Table 7. As can be seenfrom the table, the four customers consider deliveryreliability, accuracy and flexibility as the mostimportant factors.
Once Wi were calculated, the weighted impor-tance Wi* ½n� 1� of SFs was computed in accor-dance with Eq. (4). As regards to the crisp distancedi between the firm’s performance and the one thatis perceived by customer as superior, the parameterhas been computed as the average of crisp distancesdi,x the generic xth customer perceives against ithservice factor, as shown in the following equation:
di ¼
P4x¼1di;x
4; i ¼ 1; . . . ; n. (10)
Parameters di,x have been obtained basing on thesurvey results and by applying Eq. (3). To thisextent, a section of the survey was dedicated toperformance judgments about the service deliveredby the firm to its customers. The customers wereasked to judge the service level they were receivingfor each service factor, using the linguistic scaleshown in Table 6. Moreover, for each SF, thecustomers had to indicate the judgment which bestmatched their perception of a superior service. di,x
parameters as they result from the survey, di values,and the corresponding weighted importance Wi* areshown in Table 8.
From outcomes analysis, it emerges that custo-mers perceive a significant difference between thefirm’s service performance and optimum one interms of delivery accuracy. As can be seen compar-ing Tables 8 and 7, delivery accuracy is notconsidered as the most important service factorfrom customers’ point of view, but since theperformance delivered is far from meeting customer
ARTICLE IN PRESS
Table 7
Fuzzy importance wi,x assigned to service factors by each customer and the relative importance of service factors Wi
Service factors
Lead-time Flexibility Accuracy Reliability Fill rate Frequency Organization
accessibility
Complaints
management
Importance
judgment
C1 VH H H VH VL L L L
C2 L H H VH L L VL L
C3 L H VH H L L VL L
C4 H VH H H L VL L L
Relative importance
wi,x
C1 (0.7; 1; 1) (0.5; 0.7; 1) (0.5; 0.7; 1) (0.7; 1; 1) (0; 0; 0.3) (0; 0.3; 0.5) (0; 0.3; 0.5) (0; 0,3; 0,5)
C2 (0; 0.3; 0.5) (0.5; 0.7; 1) (0.5; 0.7; 1) (0.7; 1; 1) (0; 0.3; 0.5) (0; 0.3; 0.5) (0; 0; 0.3) (0; 0,3; 0,5)
C3 (0; 0.3; 0.5) (0.5; 0.7; 1) (0.7; 1; 1) (0.5; 0.7; 1) (0; 0.3; 0.5) (0; 0.3; 0.5) (0; 0; 0.3) (0; 0,3; 0,5)
C4 (0.5; 0.7; 1) (0.7; 1; 1) (0.5; 0.7; 1) (0.5; 0.7; 1) (0; 0.3; 0.5) (0; 0; 0.3) (0;0.3; 0.5) (0; 0,3; 0,5)
Relative importance
of service factors Wi
(0.740; 2; 3) (1.3; 2.59; 4) (1.3; 2.59; 4) (1.48; 2.98; 4) (0; 0.72; 1.8) (0; 0.81; 1.8) (0; 0.51; 1.6) (0; 1.02; 2)
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 595
requirements, it should be considered as one of thekey service factors to tune.
The next step in the construction of the HOQ wasthe assessment of the relationships matrix Rij
½n�m�. To this extent, strategic actions SAs forcustomer satisfaction have been listed in columns,while service factors SFs have been crossed over inrows. The degree of relationship (weak, medium,strong) between SAs and SFs has been expressed bythe work group using graphics symbols, which areusually adopted in crisp QFD approaches. Sincefuzzy logic is exploited to well cope with the ill-defined nature of linguistics judgements, graphicssymbols have been then translated into as manyfuzzy triangular numbers instead of crisp ones.Table 9 shows the correspondence between symbolsand fuzzy numbers.
During this phase, the work group benefited froma preliminary literature survey phase, which strivedto highlight the relationships between service factorsand strategic actions. The resulting relationshipsmatrix is shown in the centre of Table 11.
The roof of correlations was built up in a similarmanner. Again, traditional QFD symbols have beenused to express the correlations between strategicactions (strong negative, negative, positive, strongpositive); symbols have been thus translated intofuzzy triangular numbers, as shown in Table 10.
Once the relationships matrix and the roof ofcorrelations were compiled, the relative importanceRIj and the weighted importance RIj* of eachstrategic action were computed in accordance withEqs. (5) and (6) respectively. Results are shown inTable 11.
Then, the cost Cj for the implementation of eachstrategic action was determined to evaluate themarginal benefit Uj. To this extent, the work groupmembers were asked to express a linguistic judg-ment about the investment required for eachstrategic action, by using the same 4 value fuzzyscale previously shown in Table 6. Results areshown in Table 11. It should be remarked that fuzzylogic was found to be a very consistent and easy touse tool to handle such a vague, imprecise and ill-defined issue as costs estimation for strategicactions.
Then, the fuzzy resulting benefits Uj have beencomputed according to Eq. (7). Finally, fuzzy Uj
parameters were de-fuzzified applying Eq. (10).Crisp Uj obtained can be regarded as synthesisparameters, expressing the overall efficiency ofimplementing the jth strategic action. The finalranking of strategic actions together with the fuzzyand crisp Uj values are shown in the last two rows ofTable 11.
As a result, Information Technology emerged asthe strategic action with the highest implementationpriority, since, despite the very high cost forimplementation, it makes it possible to improvethe most important service factors, such as deliveryaccuracy and reliability. In addition, InformationTechnology has positive relationships against lead-time, fill-rate, delivery frequency and organizationaccessibility, and it has been proved to have positivecorrelations against other strategic actions. Inparticular, a strong positive relationship can befound between information technology and JITimplementation and there is a positive relationship
ARTICLE IN PRESS
Table
8
Distances
difrom
theoptimum
perform
ance
andweightedim
portance
Wi*
ofeach
servicefactor
Perform
ance
judgments
Optimum
perform
ance
Distance
di,
xDistance
di
Relative
importance
Wi
Weighted
importance
Wi*
C1
C2
C3
C4
C1
C2
C3
C4
C1
C2
C3
C4
Lead-tim
eVH
HH
HVH
VH
VH
VH
00.1
0.1
0.1
0.075
(0.740;2;3)
(0.056;0.150;0.225)
Flexibility
LVH
VH
HH
VH
VH
H0.5
00
00.125
(1.3;2.59;4)
(0.163;0.324;0.5)
Accuracy
LH
VH
LVH
VH
VH
VH
0.6
0.1
00.6
0.325
(1.3;2.59;4)
(0.423;0.842;1.3)
Reliability
HH
HH
VH
HVH
VH
0.1
00.1
0.1
0.075
(1.48;2.98;4)
(0.111;0.224;0.3)
Fillrate
HL
LH
HH
LH
00.5
00
0.125
(0;0.72;1.8)
(0;0.09;0.225)
Frequency
HL
LH
HH
HH
00.5
0.5
00.25
(0;0.81;1.8)
(0;0.203;0.45)
Organizationaccessibility
VH
LH
HVH
VH
VH
VH
00.6
0.1
0.1
0.2
(0;0.51;1.6)
(0;0.102;0.32)
Complaints
managem
ent
HL
HL
HH
HH
00.5
00.5
0.25
(0;1.02;2)
(0;0.225;0.5)
Table 9
Degree of relationship, graphic symbols and corresponding fuzzy
numbers
Degree of relationship Graphic symbol Fuzzy number
Strong K (0.7; 1; 1)
Medium J (0.3; 0.5; 0.7)
Weak m (0; 0; 0.3)
Table 10
Degree of correlation, graphic symbols and corresponding fuzzy
numbers
Degree of correlation Graphic symbol Fuzzy number
Strong positive K (0.7; 1; 1)
Positive J (0.5; 0.7; 1)
Negative & (0; 0.3; 0.5)
Strong negative ’ (0; 0; 0.3)
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599596
between information technology and demand fore-casting methods.
4. Conclusions
This study has addressed the applicability ofQFD in the logistics and supply chain managementcontext. More specifically, an original methodologyhas been proposed and adopted to rank viablestrategic actions a firm can undertake to improvelogistics performances.
The methodology developed could be rightlyconsidered as a useful tool for selecting the mostefficient and effective logistics leverages to reachcustomer satisfaction. In particular, the methodol-ogy allows the identification of the service factorsthat are perceived to affect logistics performancesfrom the customer’s point of view, enabling theassessment of possible gaps between customers’ andfirm’s perception of logistics service. As a matter offact, this is why firm’s perception should not beconsidered as the starting point in developingservice strategies, while direct interviews withcustomers are required. In our approach, such anissue is addressed through the computation of thedistance between firm performance in terms oflogistics service and that which is perceived bycustomers as superior.
In a similar manner, basing on the importance ofcustomers, the weighted importance of service
ARTICLE IN PRESS
Table
11
Thecustomer
servicehouse
ofquality
Stra
tegi
cac
tions
Rel
ativ
eim
port
ance
Wi
Rea
lim
port
ance
Wi*
Just
-In-
Tim
eph
iloso
phy
Ord
erpi
ckin
gop
timiz
atio
nIn
form
atio
nte
chno
logy
Dem
and
fore
cast
ing
met
hods
Cus
tom
erre
latio
nshi
pm
anag
emen
t
War
ehou
ses
lay-
out
optim
izat
ion
Serv
ice
fact
ors
Lea
d-tim
e(0
.7;
1;1)
(0;
0.3;
0.5)
(0.7
;1;
1)(0
;0.
3;0.
5)(0
.740
;2;
3)(0
.056
;0.
150;
0.22
5)Fl
exib
ility
(0.7
;1;
1)(0
.3;
0.5;
0.7)
(0;
0.3;
0.5)
(1.3
;2.
59;
4)(0
.163
;0.
324;
0.5)
Acc
urac
y(0
.3;
0.5;
0.7)
(0.3
;0.
5;0.
7)(1
.3;
2.59
;4)
(0.4
23;
0.84
2;1.
3)R
elia
bilit
y(0
.3;
0.5;
0.7)
(0.3
;0.
5;0.
7)(0
;0.
3;0.
5)(1
.48;
2.98
;4)
(0.1
11;
0.22
4;0.
3)Fi
llra
te(0
.3;
0.5;
0.7)
(0.3
;0.
5;0.
7)(0
;0.
72;
1.8)
(0;
0.09
;0.
225)
Freq
uenc
y(0
.3;
0.5;
0.7)
(0;
0.3;
0.5)
(0;
0.81
;1.
8)(0
;0.
203;
0.45
)O
rgan
izat
ion
acce
ssib
ility
(0.7
;1;
1)(0
.7;
1;1)
(0;
0.51
;1.
6)(0
;0.
102;
0.32
)C
ompl
aint
sm
anag
emen
t(0
.7;
1;1)
(0;
1.02
;2)
(0;
0.22
5;0.
5)
Rel
ativ
eim
port
ance
RI j
(0.3
125;
1.10
76;
2.16
)(0
;0;
0.06
75)
(0.2
48;
0.99
;2.
3075
)(0
;0.
045;
0.24
75)
(0;
0.35
7;0.
97)
(0;
0;0.
0675
)R
eal
impo
rtan
ceR
I j*
(0.4
86;
2.09
9;4.
5013
)(0
;0;
0.06
75)
(0.4
665;
2.13
1;4.
715)
(0.1
24;
0.73
91;
2.55
5)(0
;0.
357;
0.97
)(0
;0.
3323
;1.
1475
)C
ost
ofim
plem
enta
tion
Cj
(0.7
;1;
1)(0
.5;
0.7;
1)(0
.7;
1;1)
(0.5
;0.
7;1)
(0.5
;0.
7;1)
(0.7
;1;
1)U
tility
fact
orU
j(0
.486
;2.
099;
6.43
0)(0
;0;
0.13
5)(0
.466
5;2.
131;
6.73
6)(0
.124
;1.
056;
5.11
)(0
;0.
51;
1.94
)(0
;0.
3323
;1.
639)
Cri
spva
lues
2.77
90.
0337
52.
8659
1.83
60.
740.
5759
0.5;
0.7;
1
0.7;
1;1
0;0.
3;0.
5
E. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 597
ARTICLE IN PRESSE. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599598
factors allows the firm to identify the key factors ofintervention in order to improve the perceivedservice. As an example, delivery reliability emergesin Table 7 as the most important factor fromcustomers’ point of view, while, based on thejudgements the customers give on the performancedelivered, delivery accuracy should be considered asthe key service factors to tune (see Table 8).
In order to assess and rank viable strategic actions,in the approach proposed we have introduced a utilityfactor, which considers the costs of implementationfor each ‘‘how’’. The utility factor can be directlyadopted as a synthesis parameter to select the mostsuitable strategic action to implement.
The methodology has been found to be an effectiveand easy tool to adopt. Once a preliminary surveyphase concerning customer service perception isproperly set up, results returned can be easilyprocessed in the proposed HOQ, giving the mostefficient strategic action to improve customer service.In addition, the QFD approach proposed has made itpossible to appraise the beneficial impact of strategicleverages over service factors, as well as the positivecorrelations with other strategic actions.
Since personal judgements are required whenbuilding the customer service HOQ, fuzzy logic hasbeen adopted as a useful tool. Through fuzzy logiclinguistic judgments a DM gives to weights,relationships and correlations have been appropri-ately translated into triangular a fuzzy number.Moreover, fuzzy logic has allowed to cope well withuncertainties and incomplete understanding of therelationships between ‘‘hows’’ and between ‘‘hows’’and ‘‘whats’’. In addition, fuzzy logic becomesfundamental to dealing with several parametersthat seem difficult to express in a quantitativemeasure. As an example, detailed informationabout costs of implementation for strategic actionsare usually not available, while linguistic judge-ments on costs can be easily obtained.
The methodology proposed does not deal withthe practical implementation of strategic actions.Future work may be thus directed to extend asimilar QFD approach from a strategic level totactical and operational ones.
References
Akao, Y., 1990. Quality Function Deployment: Integrating
Customer Requirements into Product Design. The Produc-
tivity Press, Cambridge, MA.
Bailey, G., 1996. Customer care—Making it work. Managing
Service Quality 6 (3), 36–38.
Baines, A., 1996. Designing customer service programmes. Work
Study 45 (1), 20–23.
Behara, R.S., Chase, R.B., 1993. Service quality deployment:
Service quality by design. In: Sarin, R.V. (Ed.), Perspectives in
Operations Management: Essays in Honor of Elwood S.
Buffa. Kluwer Academic Publisher, Norwell, MA.
Bouchereau, V., Rowlands, H., 2000. Quality function deploy-
ment: The unused tool. Engineering Management Journal,
45–52 February.
Chien, C.J., Tsai, H.H., 2000. Using fuzzy numbers to evaluate
perceived service quality. Fuzzy Sets and Systems 116,
289–300.
Choen, L., 1995. Quality Function Deployment: How to Make
QFD Work for You. Addison-Wesley Publishing Company,
Reading, MA.
Christopher, M., 1998. Logistics and Supply Chain Management.
Pitman Publishing, London, UK.
Franceschini, F., Rafele, C., 2000. Quality evaluation in Logistics
services. International Journal of Agile Management Systems
2 (1), 49–53.
Franceschini, F., Rossetto, S., 1997. Design for quality: Selecting
product’s technical features. Quality Engineering 9 (4),
681–688.
Hauser, J.R., Clausing, D., 1988. The house of quality. Harvard
Business Review 66 (3), 63–73.
Keller, S.B., Savitskie, K., Stank, T.P., Lynch, D.F., Ellinger,
A.E., 2002. A summary and analysis of multi-item scales used
in logistics research. Journal of Business Logistics 23 (2),
83–281.
Lapidus, R.S., Schibrowsky, J.A., 1994. Aggregate complaint
analysis: A procedure for developing customer service
satisfaction. Journal of Services Marketing 8 (4), 50–60.
Lee-Kelley, L., Davies, S., Kangis, P., 2002. Service quality
for customer retention in the UK steel industry: Old
dogs and new tricks? European Business Review 14 (4),
276–286.
Morris, J., 1996. Leading to customer care. Industrial and
Commercial Training 28 (5), 7–10.
Philip, G., Hazlett, S.A., 1996. The measurement of service
quality: A new P-C-P attributes model. International Journal
of Quality & Reliability Management 14 (3), 260–286.
Robledo, M.A., 2001. Measuring and managing service quality:
Integrating customer expectations. Managing Service Quality
11 (1), 22–31.
Stuart, F.I., Tax, S.S., 1996. Planning for service quality: An
integrative approach. International Journal of Service In-
dustry Management 7 (4), 58–77.
Tang, J., Fung, R.Y.K., Baodong, X., Wang, D., 2002. A new
approach to quality function deployment planning with
financial consideration. Computers & Operations Research
29, 1447–1463.
Trappey, C.V., Trappey, A.J.C., Hwang, S.J., 1996. A com-
puterized quality function deployment approach for retail
services. Computers and Industrial Engineering 30 (4),
611–622.
Vandermerwe, S., Rada, J., 1988. Servitization of business:
Adding value by adding services. European Management
Journal 6 (4), 314–324.
Yager, R.R., 1981. A procedure for ordering fuzzy subsets of the
unit interval. Information Science 24, 143–161.
ARTICLE IN PRESSE. Bottani, A. Rizzi / Int. J. Production Economics 103 (2006) 585–599 599
Zadeh, L.A., 1965. Fuzzy sets. Information and Control 8,
338–353.
Zeithaml, V.A., 1988. Consumer perception of price, quality and
value: A means-end model and synthesis of evidence. Journal
of Marketing, 2–22.
Zeithaml, V.A., Parasuraman, A., Berry, L.L., 1990. Delivering
Quality Service: Balancing Customer Perception and Expec-
tations. The Free Press, New York, NY.
Zimmermann, H.J., 1991. Fuzzy Set Theory and its Applications,
second ed. Kluwer Academic Publishers, Boston.