You are on page 1of 8

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values

J. R. SHARMA, INSTITUTE OF MANAGEMENT TECHNOLOGY, NAGPUR A. M. RAWANI, NATIONAL INSTITUTE OF TECHNOLOGY, RAIPUR
2007, ASQ

Quality function deployment (QFD) is a methodology used to achieve higher customer satisfaction. The engineering characteristics (ECs) affecting product performance are designed to match the customer requirements (CRs). Computing raw weights of CRs and priority scores of ECs from various input variables is one of the most critical phases in QFD application. This study deals with the most neglected and often-omitted part of the QFD processthe interrelationship matrix among customer requirements. This article presents a mathematical solution to the correlation triangle problem that incorporates both CRs and ECs. An attempt is made to address the issue by considering the impact of these interrelationships on the overall calculations, using weighted average method for the CR interrelationship value operator in order to prioritize customers requirements in QFD. The outcome of the study is a mathematical solution to the perennial correlation triangle problem in the form of a framework that factors in the correlation triangle values in conjunction with computed raw weights for customer requirements. The model fine tunes and adds precision to an otherwise qualitative strategic decision process. The applicability of the authors proposed model is demonstrated with a real-life example. Key words: customer requirements, interrelationship values, prioritization, QFD

INTRODUCTION
In any kind of project, the design team needs to create, improvise, innovate, or improve a product by incorporating the customer requirements (CRs) and engineering characteristics (ECs) recognized in the quality function deployment (QFD) planning process into a product. A design team needs to make tradeoffs while selecting the CRs and ECs based on the order of their relative raw weights and priority scores and their ranking to achieve total customer satisfaction without violating the time, money, and effort constraints. The technical priority score and their rankings are the key results of QFD, since they guide the design team in decision making, resource allocation, and the subsequent product analyses. Therefore, deriving the final importance rankings of CRs and ECs from input variables is a crucial step toward successful QFD. The inherent abstraction and the inexactness of the result, however, presents a special challenge in effectively calculating the customer raw weights and technical priority scores and, hence, the importance rankings of CRs and ECs. The aim of this article is to advance the theory and practice of incorporating and factoring in the interrelationship values of the house of quality (HoQ). Implementation of this improvised approach to comprehensive matrix analysis is discussed from the perspective of a product by conducting a hypothetical analysis.

QFD: ORIGIN AND PROGRESSION


QFD is the systematic translation of the voice of the customer (VOC) to the actions of the manufacturer required to meet the customers desires, based on a

www.asq.org 53

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


matrix that compares what the customer wants and how the supplier plans to provide it. It is said that QFD was proposed in Japan in 1966 by Yoji Akao. However, it did not emerge as a viable and formalized approach to quality control in planning until 1972 when Akao developed a quality control chart previously introduced at the Kobe shipyard of Mitsubishi Heavy Industries and instituted the QFD quality tables (Griffin and Hauser 1996). QFD is an innovative approach to bringing qualityas demanded by the customers upstream into the product development process (Akao 1990). QFDs first industrial applications in the United States originated primarily in the automotive industry. With some modifications, QFD could be applied to the continuous improvement of any product, process, or technique (Sharma and Rawani 2006). Early users of QFD included Ford Motor Company, Procter & Gamble, and 3M Corp., but many other U.S. companies have also adopted it. In Europe, the first symposium on QFD took place in 1992 in the United Kingdom, but companies like Philips Corporation have reportedly worked with QFD since 1986 (Govers 1996). QFD has been in use in the manufacturing industry since 1987. Meanwhile, many authors have advocated it as a planning tool to help manage product and process development.
Figure 1 Basic building blocks of QFD (Sharma, Sharma, and Rawani 2006).
Roof Product Voice of customers Voice of engineer Strategic planning room
2007, ASQ

CR triangle

Relationship room

Technical priorities room

product or service. The most common approach is to ask: Who must be satisfied with the product for it to be considered successful? Identifying the CRs and their importance remains one of the most serious obstacles to a successful QFD application in product development.

CR Triangle
The CR correlation triangle is designed to show the interrelationship between the CRs of the matrix. Tradeoffs between the different CRs are identified and captured as strong positive, positive, negative, and strong negative. If two CRs help each other move in their desired direction, they are considered to be positively related. Similarly, if moving in the desired direction on one CR has a detrimental effect on another, they are considered to be negatively related. Since they are generally not used in calculation, the interrelationship data are defined with symbols for associated relationships.

QFD FRAMEWORK
In the QFD approach, the matrix to be built is the product planning matrix. It also is called the house of quality (HoQ) because of its house-like shape. Its purpose is to translate important customer requirements regarding product quality into key end-product control characteristics. The HoQ comprises several different parts or rooms, which are sequentially filled in order to achieve an actionable translation from requirements into characteristics (see Figure 1).

Strategic Planning Room


Once the CRs and their priority rankings are established, the next step is to understand where the company and its competitors stand in terms of satisfying customers in the marketplace. This is achieved through an index of the customer perceptions as to how well the product one uses meets his or her needs. This section provides a link between the QFD project and the companys strategic vision by uncovering market opportunities and

Voice of Customer
The HoQs first room concerns the product and deals with the definition of the selected product for which QFD is constructed. The second room concerns the VOC, but before this there is a need to identify the customers. There are different ways to identify the customers of a

54 QMJ VOL. 14, NO. 4/ 2007, ASQ

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


identifying priorities for product development or improvement. Moreover, it helps build competitive advantage by focusing on the requirements that customers would like to see better addressed by the market. Based on the information depicted in the HoQ, the QFD team then must select the end-product characteristics to be deployed through the remainder of the product development and market introduction activities. ECs showing a technical importance rating above a predetermined threshold, thus indicating a strong relative importance in achieving the CRs, are selected for further deployment. The same happens to ECs related with CRs that have strong sales points or poor competitive performance satisfaction rating.

Voice of the Engineer and Roof


At this stage, the QFD team should have a clear picture of what the customer requires from the concept or product and how this can be related to the companys strategy. The team now has to decide how these requirements can be incorporated into the final product, so that the customer is satisfied. This is achieved by building the Voice of Engineer Room. The goal is to develop a set of engineering specifications from the CRs.

RANKING METHODS IN QFD: A REVIEW


The study of the available literature on various models and frameworks for QFD suggests that most of the methodologies suffer from one major weaknessa lack of concrete, clear guidelines as to how one could adequately conceptualize, integrate, and implement its correlation triangle phase. Several methodologies have been worked out, but their validity and applicability to prioritize CRs remains inconclusive and the implementation results and findings are scarce. In QFD studies various CRs are always stated and included, but interrelationship with one another is hardly ever incorporated and, even if depicted, is never brought into the tabulations of the CR final importance ratings. One reason for this is the difficulty in obtaining the relevant data, although it could be expressed in a symbolic manner using appropriate synergistic or detrimental scales. Interrelationships among the CRs require the QFD personnel to make a lot of pair-wise comparisons about the degree of association and the direction to which these CRs are interrelated. Also, there are few, if any, acceptable methods to incorporate the interrelationship into the calculation of the final importance ratings of the CRs. Most of these methods, however, adopt calculation procedures using the weighted product of the relationship measures in the importance rating column, sales factor, and improvement ratio column, without considering the magnitude, direction, and degree of association among all CRs.

Relationship Room
The Relationship Room is the core of the HoQwhere the relationship between each CR (VOC) and the ECs (voice of the engineer), as well as their intensity, is depicted. The relationship between CRs and ECs might be achieved by asking, To what degree does this engineering characteristic predict the customers satisfaction with this requirement? Based on in-house expertise, surveys, data from statistical studies, and controlled experiments, the team seeks consensus as to how much each EC affects each CR. This task is widely recognized as highly complex and represents another critical stage in the HoQ building process.

Technical Priorities Room


The Technical Priorities Room starts with technical descriptors. These descriptors pertain to the ECs and the units of measurement in each column. They add the details necessary to bring the product definition from the abstraction of words to the concrete reality of product and process engineering. With technical competitive assessment of the ECs (voice of the company) in the currently marketed products, the QFD team can view their competitors and their own technical performance level regarding ECs directly affecting CRs. A row indicating the level of organizational difficulty related with realizing each EC can also be added.

www.asq.org 55

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


Figure 2 Flowchart for the proposed method.
START

i=1

IRWi = NPRWi j=1

Yes

If j i No

IRWi = NPRWi + [NPRWi * (1 + NPRWj) * Xij]


j=1

j=j+1

If j n No i=i+1

Yes

an approach based on the application of possibility theory and fuzzy arithmetic to address the vagueness in QFD operation. Fung, Popplewell, and Xie (1998) developed a hybrid system to incorporate the principles of QFD, analytical hierarchy progress, and fuzzy set theory for customer requirements analysis and product attribute targets determination. Shen, Tan, and Xie (2001) proposed a fuzzy procedure to examine the sensitivity of ranking different parameters to the defuzzification strategy and degree of fuzziness of fuzzy numbers. Vangeas and Labib (2001) proposed a model for deriving optimum targets of CRs and ECs through the implementation of the fuzzy weighted average. On the similar lines, new methods for rating CRs and ECs in fuzzy QFD is proposed by integrating fuzzy weighted average method (Kao and Liu 2001) and the fuzzy expected value operator (Liu and Liu 2002). Karsak, Sozer, and Alptekin (2002) combined the AHP and goal programming approach for determining the priority levels of CRs. Han, Kim, and Choi (2004) suggested a linear partial ordering approach for prioritizing ECs in QFD with incomplete information. Karsak (2004) used the fuzzy multiple objective programming framework for prioritization in QFD. Chen, Fung, and Tang (2005) proposed a ranking model using the weighted average method in the fuzzy expected value operator in fuzzy QFD.

If i n

OUTPUT: IRW VALUES


2007, ASQ

STOP

INTERRELATIONSHIP INTEGRATED MATHEMATICAL MODEL


To overcome the aforementioned problem, a new method for prioritizing CRs in QFD is proposed by integrating the CR correlation triangle values with preliminary (initial) raw weights. The proposed method can weigh the CRs more efficiently, as not only the relative importance ratings, sales points, and improvement ratios of CRs are considered, but the values and their degree of association in the correlation triangle are also factored in. Through the method proposed in this study, the CR correlation triangle symbols are translated into numerical values and with this a discreet but exact solution is obtained.

A few approaches have been introduced for determining CRs and ECs importance rating in the QFD process. The noticeable ones found in the QFD literature dealing with prioritization are mentioned here: The prioritization that is consistent with CRs importance ratings was advocated by Lyman (1990) through the concept of deployment normalization. Armacost et al. (1994) used the analytic hierarchical process to prioritize CRs and ECs in an industrialized housing application. Khoo and Ho (1996) proposed

56 QMJ VOL. 14, NO. 4/ 2007, ASQ

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


The prioritization weights are then computed by using the preliminary raw weights. To factor in the CR correlation triangle values which usually are in the form of symbols the model starts only after the prioritization of the CRs by rating them through their raw weights. This article considers QFD done on a healthcare product. It should be noted, however, that the whole QFD matrix preparation and its calculation for the selected product are beyond the scope of this article. The legends used are: PRWPreliminary raw weight SPRWSum of all preliminary raw weights NPRWNormalized preliminary raw weight IRWIntermediate raw weight SIRWSum of all intermediate raw weights NIRW Normalized intermediate raw weight FRWNew raw weight Xij Interrelationship values among CRs, that is, between row i and column j The interrelationship values are taken as 1.5 for + +, 1.25 for +, 0.5 for , 0.25 for and 0.00 for blanks. The value conversion has been done considering the synergistic or detrimental effect of these relationships on various CRs under consideration. The calculations for ith (i = 1 to n) CR is given next: NPRWi = PRWi / SPRW IRWi = NPRWi + [NPRWi * (1 + NPRWj) * X ij ], for j=1 all j i These intermediate raw weights are then normalized, and these normalized values are distributed according to preliminary raw weights to give final raw weights, as given next. NIRWi = IRWi / SIRW FRWi = NIRWi * SPRW
n

Figure 3 Customer requirements and their interrelationships.


Customer Requirements + + + + Cleanliness and purity Safe, reliable, and efficient Ease of handling and use Cost of the product PRW 69 72 40 86 59 63 42 47 67 96 NPRW 0.1076 0.1123 0.0624 0.1342 0.0920 0.0983 0.0655 0.0733 0.1045 0.1498 Rank IV III X II VII VI IX VIII V I
2007, ASQ

+ + Leakage factor (air/liquid) + ++ + + ++ Right size/Correct volume + ++ Proper visible markings Precise movement ++ Safe convenient packaging Tamper proof (no reuse)

nNumber of customer requirements

The rationale behind the devised formulae is the relative direction and the magnitude of change that one CR causes on to the other, depicted with the help of synergistic or detrimental symbols. If there is a synergistic (or detrimental) relationship between two CRs, the effect is positive (or negative). This causes CRs to move in the right (or opposite) direction, the intensity of which is decided by the nature of relationship, that is, strong or weak. The net effect is to increase (that is, multiplying factor > 1.00) or decrease (that is, multiplying factor < 1.00) the considered CRs weight by the related percentage value of other related CRs, thus taking into account the assigned degree and direction of association. A Visual Basic Code has been written for the proposed method. A flowchart for the same is shown in Figure 2.

AN ILLUSTRATIVE EXAMPLE
After going through all QFD steps depicted previously in the basic building blocks for the selected healthcare product syringe and needle, the calculated raw weights for each CR are shown next. The matrix ignores redundant steps and considers only the relevant ones, that is, CR, interrelationship among CRs, raw weights, normalized raw weights, and their ranks, as shown in Figure 3. The sample calculation (of CR2) shown next helps in deciphering how changes in the correlation triangle value get reflected in IRW and finally in FRW, thus affecting the resultant rankings of the CRs.

www.asq.org 57

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


Table 1 CR with correlation values, preliminary and final raw weights.
Tamper Proof (No reuse possibility) Safe & Convenient Packaging Ease of Handling & Use Right Size/ Correct Volume Safe, Reliable, & Efficient

No Leakages (Air/Fluid)

CR
Cleanliness & purity Safe, reliable, & efficient Ease of handling & use Cost of the product No leakages (Air/Fluid) Right size/ Correct vol. Proper markings Precise movements Safe packaging Tamper proof (No reuse)

Precise Movements

Cleanliness & Purity

Cost of the Product

Proper Markings

PRW NPRW
69 72 0.1076 0.1123

IRW
0.6217 1.1367

NIRW
0.1046 0.1912

FRW
67.03 122.54

1.25

1.25

0 0

0.50 0.25

0 1.25

0 1.25

0 1.25

0 0

1.25 1.50

1.25 1.50

0 0.50 0 0 0 0 1.25 1.25

0 0.25 1.25 1.25 1.25 0 1.50 1.50

0 0 0 1.25 1.25 0 0

0 0.50 0.50 0.50 0.25 0.50 0.25

0 0.50 1.25 0 1.25 0 0

0 0.50 1.25 1.50 0 0 0

1.25 0.50 0 1.50 0 0 0

1.25 0.25 1.25 0 0 0 0

0 0.50 0 0 0 0 1.50

0 0.25 0 0 0 0 1.50
SUM

40 86 59 63 42 47 67 96
641

0.0624 0.1342 0.0920 0.0983 0.0655 0.0733 0.1045 0.1498


1.000

0.2292 0.6128 0.5221 0.5819 0.3887 0.2916 0.6632 0.8976


5.9455

0.0386 0.1031 0.0878 0.0979 0.0654 0.0490 0.1115 0.1510


1.000

24.71 66.07 56.29 62.74 41.91 31.44 71.50 96.77


641.00
2007, ASQ

Considering row no. 2 (CR2: safe, reliable, and efficient) PRW2 = 72.00 (Old rank is 3) NPRW2 = 72.00 / 641.00 = 0.1123 X2j = {0.00, , 0.00, 0.25, 1.25, 1.25, 1.25, 0.00, 1.50, 1.50} SPRW = 641.00 IRW2 = NPRW2 + { [NPRW2 * (1 + NPRWj ) * X2j ]} =
j=1 n

Similarly, values of IRW for all other CRs are computed. SIRW = 5.9455 NIRW2 = IRW2 / SIRW = 1.1367 / 5.9455 = 0.1912 FRW2 = NIRW2 * SPRW = 0.1912 * 641.00 = 122.5449 (New rank is 1) Similarly, values of FRW for all other CRs are computed (see Table 1).

0.1123 + {0 + 0 + 0.1123 * (1 + 0.1342) * 0.25 + 0.1123 * (1 + 0.0920) * 1.25 + 0.1123 * (1 + 0.0983) * 1.25 + 0.1123 * (1 + 0.0655) * 1.25 + 0 + 0.1123 * (1 + 0.1045) * 1.50 + 0.1123 * (1 + 0.1498) * 1.50} = 1.1367

RESULTS
Results can be deciphered from the comparative analysis of CRs raw weights and their ranking (see Table 2). The old traditional method and the resulting

58 QMJ VOL. 14, NO. 4/ 2007, ASQ

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


Table 2 A comparative analysis of CRs preliminary and final raw weights and rankings.
C R (Conventional Method)
Tamper proof (no reuse) Cost of the product Safe, reliable, and efficient Cleanliness and purity Safe convenient packaging Right size/Correct volume No leakages (Air/Fluid) Precise movements Proper markings Ease of handling and use

PRW
96 86 72 69 67 63 59 47 42 40

NPRW
0.1498 0.1342 0.1123 0.1076 0.1045 0.0983 0.0920 0.0733 0.0655 0.0624

Rank
I II III IV V VI VII VIII IX X

C R (Proposed Method)
Safe, reliable, and efficient Tamper proof (no reuse) Safe & convenient packaging Cleanliness and purity Cost of the product Right size/Correct volume No leakages (Air/Fluid) Proper markings Precise movements Ease of handling and use

FRW
122.5449 96.7727 71.4992 67.0252 66.0727 62.7384 56.2857 41.9110 31.4355 24.7148

NIRW
0.1912 0.1510 0.1115 0.1046 0.1031 0.0979 0.0878 0.0654 0.0490 0.0386

Rank
I II III IV V VI VII VIII IX X
2007, ASQ

rankings leave a lot to be desired. However, this lacuna gets solved with the computation of the final raw weight obtained through CR correlation triangle values. The final raw weights, their rankings, and their order are much more precise and accurate, leading to better and informative decision making by product designers.

CONCLUSIONS
The prioritization and the judicious selection of CRs from the markets perspective is a complex task and needs special attention. Over the past few years, there has been an huge effort in establishing a decision framework that quantifies the imprecise and subjective customer information inherent in the product planning process. This research has proposed a novel QFD approach that solves some methodological problems encountered in the conventional QFD. While the conventional approach uses only symbols to express the strength or weakness of the relationships among CRs, albeit only in few cases, the new approach addresses relationships in a different and improved numerical method. The article shows the effectiveness and preciseness of the proposed model. Numbers are used to represent the imprecise nature of the judgments and to define

more appropriately the relationships among CRs. The model is shown to be computationally feasible for realistic problems and outperforms traditional approaches on the basis of being relatively straight forward and simplified. A systematic algorithm for calculating the critical targets of CRs is followed. The framework presented in this article for prioritizing CRs extends the single objective viewpoint of maximizing customer satisfaction by considering the companys other market-related objectives, and thus, precludes an unreasonable QFD planning in practice. The proposed methodology provides product developers, designers, and engineers with an objective method for making a judicious selection of critical CRs satisfying the overall CRs. The illustrated example clearly shows that the proposed approach can produce more informative results, adding credibility to the outcome and its analysis, for the benefit of the product designers and developers.
REFERENCES

Akao, Y. 1990. Quality function deployment: Integrating customer requirements into product design , Cambridge, Mass.: Productivity Press. Armacost, R., P. Componation, M. Mullens, and W. Swart. 1994. AHP framework for prioritizing customer requirement in QFD: An industrialized housing application. IIE Transactions 26, no. 4: 72-79.

www.asq.org 59

Ranking Customer Requirements in QFD by Factoring in Their Interrelationship Values


Chen Yizeng, R. Fung, and J. Tang. 2005. Rating technical attributes in fuzzy QFD by integrating fuzzy weighted average method and fuzzy expected value operator. European Journal of Operational Research (Production, Manufacturing and Logistics). Fung, R. Y. K., K. Popplewell, and J. Xie. 1998. An intelligent hybrid system for customer requirements analysis and product attribute targets determination. International Journal of Production Research 36, no. 1: 13-34. Govers, C. 1996. What and how about quality function deployment (QFD). International Journal of Production Economics : 575-585. Griffin, A., and J. Hauser. 1996. Integrating R & D and marketing: A review and analysis of the literature. Journal of Product Innovation Management: 191-215. Han Chang, Kim Jae, and Choi Sang Hyun. 2004. Prioritizing engineering characteristics in QFD with incomplete information: A linear partial ordering approach. International Journal Production Economics 91: 235-249. Kao, C., and S. Liu. 2001. Fractional programming approach to fuzzy weighted average. Fuzzy Sets and Systems 120: 435-444. Karsak, E. 2004. Fuzzy multiple objective programming framework to prioritize design requirements in QFD. Computers & Industrial Engineering 47: 149-163. Karsak, E., S. Sozer, and E. Alptekin. 2002. Product planning in QFD using a combined analytic network process and goal programming approach. Computers and Industrial Engineering 44: 171-190. Khoo, L., and N. Ho. 1996. Framework of a fuzzy quality function deployment system. International Journal of Production Research 34, no. 2: 299-311. Liu, B., and K. Liu. 2002. Expected value of fuzzy variable and fuzzy expected value models. IEEE Transactions on Fuzzy Systems 10, no. 4: 445-450. Lyman, D. 1990. Deployment normalization. In Transactions from the 2nd Symposium on Quality Function Deployment: 307-315. Sharma, J., and A. Rawani. 2006. Customer driven product development through quality function deployment (QFD). Asia Pacific Business Review A Journal of Asia-Pacific Institute of Management 2, no. 1: 45-51. Sharma, J., D. Sharma, and A. Rawani. 2006. Quality driven product development. Manufacturing Engineer (IET). (June-July): 38-41. Shen, X., K. Tan, and M. Xie. 2001. The implementation of quality function deployment based on linguistic data. Journal of Intelligent Manufacturing 12: 65-75. Vangeas, L., and A. Labib. 2001. A fuzzy quality function deployment (FQFD) model for deriving optimum targets. International Journal of Production Research 39, no. 1: 99-120.
BIOGRAPHIES J. R. Sharma is assistant professor at the Institute of Management Technology, Nagpur. Before joining this organization, he served as senior faculty in the Department of Production and Mechanical Engineering, Manoharbhai Patel Institute of Engineering and Technology (MIET), Gondia, India. He is pursuing his doctoral research at the National Institute of Technology, Raipur. His research interests are in the areas of quality management and product development. Sharma has more than 20 research papers published in various national and international conferences, technical periodicals, and journals. He has the honor of reviewing papers for several international and national journals including Elseviers European Journal of Operations Research. He can be reached by e-mail at sjiten1@gmail.com. A. M. Rawani is professor and head of the Department of Mechanical Engineering, National Institute of Technology, Raipur, India. He has more than 23 years of experience in academics. He acquired his doctorate from IIT-Delhi. He has pursued his research in the area of strategic management. Rawani has done commendable work in research and has more than 35 publications in national and international journals to his credit. He has been a nominated member of the Confederation of Indian Industry (CII) HR & Training Program Panel. He can be reached by e-mail at amrawani@rediffmail.com.

60 QMJ VOL. 14, NO. 4/ 2007, ASQ

You might also like