Application of the Analytical Network Process (ANP) to select a channel type for e-convenient chain stores.

ABSTRACT

For the convenient stores in Taiwan, three types of channels were used commonly [] including the real channel, the virtual channel, and the real-virtual channel The scope of this study is to analyze the suitable channel by advanced methods of the Analytical Network Process (ANP) for the convenient stores to meet the challenge of competitions. The patterns of management [] regular chain, franchise chain, and regular-franchise chain, were evaluated based on analysis of strategies of management [] innovation, growth, cost, differentiation, and alliance. The real-virtual channel was selected as the most suitable channel based on these studies, and the suitability of this channel to convenient stores was further evaluated in several cases to confirm its efficacy.

1. INTRODUCTION

According to convenient store chain 7-Eleven experienced explosive growth in Taiwan, the flourishing economy is mostly based on the contributions of the middle and small enterprises. A new definition may be required for the convenient stores approach the 21st century where many achievable goals. In order to stay ahead of competition, the convenient stores continuously establish the suitable channel system to enhance on the best utilization of technology for improving the ability to adapt to market demands.

There are three key components to a success of convenient store: the patterns of cooperation, the strategies of management, and the channel systems. Three patterns of cooperation include regular chain, franchise chain, and regular-franchise chain. The strategies of management include innovation, growth, cost, differentiation, and alliance. The channel systems include the real channel, the virtual channel, and the real-virtual channel. Channels create an important bridge between customers and stores; successful channels can make profit for business. Analytical network process (ANP) was applied in this study to select the suitable channel for convenient stores in order to guarantee gaining profit.

The study is organized into the following sections. Section III provides a review of convenient stores to consider and prepare in advance for effective implementation of the selected channel system. Section IV then examines the empirical evidence to date, in particular by examining survey research to understand the request of customers and the convenient stores. Finally, the study proposes a conceptual model of convenient stores was addressed in Section V.

2. SUITABLE CHANNEL AND THE CRITERIA FOR SELECTION

There are three patterns of management [] including regular chain, franchise chain, and regular-franchise chain. First, we explore the role of regular chain as independent companies. The second patterns of cooperation, companies provide the opportunity for small-sized stores to join their cooperation. The third pattern of cooperation combined both pattern of regular chain and franchise chain. The criteria used to evaluate the suitability for the selected channel are the five strategies of management used commonly in convenient stores [] including innovation, growth, cost, differentiation, and alliance.

According to the convenient stores in Taiwan, three types of channels were used commonly [] including the real channel, the virtual channel, and the real-virtual channel. Suitable channel for the convenient stores may ascertain their success in business, thus, selection of proper channel is very important.

Different patterns of cooperation and criteria were explored to evaluate the candidate channels. The relative weight for each of these criteria must be identified before it can be applied to the evaluation. Since the weighting process is influenced by the cooperation patterns. The traditional analytical hierarchical process that neglects mutual effect of different layers of the decision model is not adequate.

Patterns of management and the criteria interact with each other simultaneously, leading to the selection of a suitable channel for convenient stores as shown in Figure 1.

[FIGURE 1 OMITTED]

To deal with this dynamic problem, we applied the ANP approach to handle interdependence among criteria of different layers by obtaining the composite weights through the development of a "supermatrix". A major difference between AHP and ANP approach is that the AHP assumes that the system elements are not correlated and are undirectionally influenced by a hierarchical relation. However, the ANP approach eliminates these limitations and allows a feed back relationship among the criteria at different levels and interdependence between criteria at the same level through the development of a "supermatrix" (Saaty, 1996). For eliciting preferences of various components and attributes, decision-maker will compare two components at a time with respect to a "control" criterion.

In ANP, like AHP, pairwise comparisons of the elements at each level are conducted with respect to their relative importance towards their control criterion. The control criterion for these pairwise comparisons can be the criteria at the upper or lower levels. This is the fundamental for ANP to develop the supermatrix. The pairwise comparison of component at one level with respect to the control criterion will be expressed in the matrix form as shown in Table 1.

Once the pairwise comparisons are completed, the local priority vector w is computed as the unique solution to:

Rw = [[lambda].sub.max]w

Where [[lambda].sub.max] is the largest eigenvalue of R. Saaty (1986) provided several algorithms for approximating the vector w? In this paper, a two-stage algorithm proposed by Meade (1998) is referred to as the process of averaging over normalized columns is applied to approximate the vector w. This is represented as:

(1) [MATHEMATICAL EXPRESSION IS NOT REPRODUCIBLE IN ASCII]

In the assessment process the problem of transitivity or consistency of the pairwise comparisons must be considered. Saaty (1988) has provided an explanation on inconsistencies in the relationships and their calculations. In our example, it is assumed that the pairwise comparisons are consistent.

We have applied the ANP framework to select a suitable channel for convenient stores. The score of 1, 2, 3, 4, 5, 6, or 7 indicates absolute unimportant, very unimportant, unimportant, indifferent, important, very important, and absolute important, respectively. A pairwise comparison matrix of the criteria for the pattern (regular chain, franchise chain, and mix chain) of management is likely as Table 1. In this relationship, from the manager's point of view, the higher cost is considered as very important than the innovation in convenient stores. This is partially because that the major subject in this pattern considers the enterprise's intention for making profit. The attitude of the higher or top manager plays a critical role in this stage. The priority vector is obtained by using the equations (1) with the data in comparison matrix of the criteria for the pattern (regular chain, franchise chain, and mix chain) of management. The relative weight vector for each one of the patterns are normalized and summarized in Table 2, indicating the relative impact of each criterion on different patterns of management. Impact of various patterns of management on each one of the criteria was summarized such as Table 3.

Notice that regular-franchise pattern of management influences innovation more significant than the other two patterns. The matrix B was formed by all of these vectors that the three patterns has the analogous results of relative weights vectors as shown in Table 4.

The next step is to form the "supermatrix" that allows a solution of the effects of interdependence that exist between criteria at different levels of the system. We assume that the impacts of criteria in the same cluster are not significant. The corresponding area in the "supermatrix" will be assigned a value of zero. Hence, the two compiled matrices A and B are now combined to form the "supermatrix" M as shown in Table 5.

In the long term, the "supermatrix" converges to the stable values given in Table 5, which will be used for further analysis. At any one of the three management patterns, the innovation and the growth have earned relatively heavier weights from the decision-maker. In the meantime, the decision-maker emphasizes the innovation based on the evaluation of relative importance of the three patterns of management from the viewpoints of the five criteria.

The final step of the analysis is alternative evaluation. Each candidate enterprise needs to be evaluated based on each of the criteria. Making a pairwise comparison of the performance of each candidate enterprise on each criterion completes this. The score of 1, 2, 3, 4, 5, 6, 7 means bad, very poor, poor, indifferent, good, very good, and excellent, respectively. Since there are five criteria in this illustrative example, an additional five 4x4 pairwise comparison matrices will be needed for evaluation. As shown such as Table 6, each criteria is compared with the three candidate channels. The channel 3, real-and-virtual channel, is assumed to perform better than the other candidate enterprises. Equation (1) is applied again to calculate the priority vector in the matrix. Consequently, the adequate cooperated enterprise is selected by calculating the "desirability index" [D.sub.i] which is defined by:

(2) [MATHEMATICAL EXPRESSION IS NOT REPRODUCIBLE IN ASCII]

Where:

[P.sub.j] : The relative important weight for cooperated pattern j.

[q.sub.kj] : The relative important weight for criterion k of cooperated pattern j

[r.sub.ikj] : The relative impact of candidate enterprise i on criterion k of pattern j.

The channel 3 was found to obtain better scores than the other two channels in all three patterns of management, suggesting the real-and-virtual channel is the suitable selection.

3. CONCLUSION

As pointed out before, we found the real-virtual channel was the most potential channel of e-convenient stores in Taiwan though the Desirability Index.

1. Value chain of combining electronic real channel with virtual channel was shown as Figure 2.

[FIGURE 2 OMITTED]

2. The activities of value chain:

(1) Support System [] It will be very clear and saving lots of time using computer without checking account by the auto-stock control system in the future.

(2) Market Condition Investigation [] Analyzing the trend of market and the direction of consumers to control the newest information.

(3) Internet Human Resource(HR) Database [] It is an analytical software of personality characteristics for HR system of enterprise to enhance the inner management and use and improve quality of service and competitive. Recruiters whose characteristics are conformed to the capability of position in enterprise will be asked to interview by e-mail.

(4) Distant Learning Software [] It is a virtual learning style for sales in headquarters and retail sales to communicate with each other by interactive equipments such as mail, e-mail, internet, TV, radio, video ... etc. Employees could read and search data on the inner network library that are also included new books information and other books.

(5) Supplier and Storehouse [] Suppliers replenish the store to distribution center avoid running out of stock by computer.

(6) Advanced POS System [] It's used to handle data by SC system for external operations.

(7) Mail Affair [] It could save the search time of report and notification to improve the effect of the information delivery.

(8) Clean the concurrent message [] When the headquarters declare news temporary, the monitor on the cash register will show the company purpose in the retail sales. This function is used to clean the concurrent message in the SC system to indicate receiving notification.

(9) The new products could be marketed by dual weekly publication marketing that is the new marketing system by electronic data in Internet.

(10) Digital Image Delivery System [] In the future, it will be used to replace fax and photocopy machine by one machine that contain these functions of fax, photocopy, receive/send mail.

3. Virtual shopping on the Internet

Following strict planning, the on-line convenient shopping has created by a perfect synergy between real distribution and virtual networks. This service provides the safest and best shopping websites for consumers. It also constructs payment system with absolute security and 24-hour on-line services; consumers concern over goods distribution and payment are solved at once. Consumers can easily select the products of their own choices based on their needs.

E-Commerce is a current tendency in the 21st century. Convenient stores take the initiative to integrate three of the key flows [] the product flow in multi-layers, the cash flow between different banking cards, and the information flow into the business service. This integration provides an omni directional business service with high quality, safety, and stable.

Looking ahead into the future, convenient store has been enlisted as the concept of e-century life network into the business. Consider how Web services can enable to enhance on the best utilization of information technology for improving the quality and effect on service, and create the more convenient life in demand. With the Community Service Center comes a new living style; it makes our people truly enjoy a more convenient and simpler way of life.

Table 1 Comparison of matrix use regular chain as control criterion

Regular chain     Innovation   Growth   Differentiation   Cost

Innovation        4            6        5                 6
Growth            2            4        3                 4
Differentiation   3            5        4                 5
Cost              2            4        3                 4
Alliance          3            5        4                 5

Regular chain     Alliance     1

Innovation        5          0.262
Growth            3          0.158
Differentiation   4          0.210
Cost              3          0.158
Alliance          4          0.210

Table 2 Priority vectors of criteria for management of patterns matrix

Patterns          Regular   Franchise   Mix
Strategies

Innovation        0.262     0.168       0.210
Growth            0.158     0.274       0.210
Differentiation   0.210     0.168       0.160
Cost              0.158     0.221       0.210
Alliance          0.210     0.168       0.210

Table 3 Comparison of matrix uses innovation as control criterion

Innovation   Regular   Franchise   Mix   w

Regular      4         6           3     0.364
Franchise    2         4           1     0.181
Mix          5         7           4     0.455

Tahla 4 Vectors of matrix B

Patterns     Innovation   Growth   Differentiation   Cost    Alliance
Strategies

Regular      0.364        0.181    0.333             0.208   0.302
Franchise    0.181        0.364    0.246             0.302   0.208
Mix          0.455        0.455    0.420             0.490   0.490

Table 5 The long term converged "supermatrix" at M (21)

Super Matrix      Regular   Franchise   Mix     Innovation   Growth
M

Regular           0         0           0       0.276        0.276
Franchise         0         0           0       0.261        0.261
Mix               0         0           0       0.463        0.463
Innovation        0.214     0.214       0.214   0            0
Growth            0.212     0.212       0.212   0            0
Differentiation   0.176     0.176       0.176   0            0
Cost              0.199     0.199       0.199   0            0
Alliance          0.199     0.199       0.199   0            0

Super Matrix      Differentiation   Cost    Alliance
M

Regular           0.276             0.276   0.276
Franchise         0.261             0.261   0.261
Mix               0.463             0.463   0.463
Innovation        0                 0       0
Growth            0                 0       0
Differentiation   0                 0       0
Cost              0                 0       0
Alliance          0                 0       0

Table 6 Comparison matrix using innovation as control criterion

Innovation         Real      Virtual Channel   Real-and-Virtual   w
                   Channel                     Channel

Real Channel       4         2                 1                  0.181
Virtual Channel    6         4                 3                  0.364
Real-and Virtual   7         5                 4                  0.455
Channel

Table 7 Desirability index calculation for candidate channels

                                            Real
Cooperation                                 Channel
Patterns          Strategies                Weight
Weight            Weight                    Score

Real      0.276   Innovation        0.214   0.181   0.011
Channel           Growth            0.212   0.181   0.011
                  Differentiation   0.176   0.392   0.019
                  Cost              0.199   0.233   0.013
                  Alliance          0.199   0.246   0.014

Virtual   0.261   Innovation        0.214   0.181   0.010
Channel           Growth            0.212   0.181   0.010
                  Differentiation   0.176   0.392   0.018
                  Cost              0.199   0.233   0.012
                  Alliance          0.199   0.246   0.013

Real-             Innovation        0.214   0.181   0.018
and               Growth            0.212   0.181   0.018
Virtual   0.463   Differentiation   0.176   0.392   0.032
Channel           Cost              0.199   0.233   0.021
                  Alliance          0.199   0.246   0.023

Desirability Index                                  0.241

Cooperation                         Virtual         Real-and-Virtual
Patterns                            Channel         Channel   Weight
Weight                              Weight Score    Score

Real      0.276   Innovation        0.364   0.021   0.455     0.027
Channel           Growth            0.364   0.021   0.455     0.027
                  Differentiation   0.217   0.011   0.392     0.019
                  Cost              0.233   0.013   0.533     0.029
                  Alliance          0.333   0.018   0.420     0.023

Virtual   0.261   Innovation        0.364   0.020   0.455     0.025
Channel           Growth            0.364   0.020   0.455     0.025
                  Differentiation   0.217   0.010   0.392     0.018
                  Cost              0.233   0.012   0.533     0.028
                  Alliance          0.333   0.017   0.420     0.022

Real-             Innovation        0.364   0.036   0.455     0.045
and               Growth            0.364   0.036   0.455     0.045
Virtual   0.463   Differentiation   0.217   0.018   0.392     0.032
Channel           Cost              0.233   0.021   0.533     0.049
                  Alliance          0.333   0.031   0.420     0.039

Desirability Index                          0.306            0.453

REFERENCES

[1] Daniel, Elizabeth, and George Klimis.(1999) " The Impact of Electronic Commerce on Market Structure: An Evaluation of the Electronic Market Hypothesis." European Management Journal, (June), pp. 318-26.

[2] Meade, L., and Sarkis, J., "A Strategic Analysis of Logistics and Supply Chain Management Systems Using Analytical Network Process". Transportation Research, Vol.34, No. 3, 1998, 201-215.

[3] Saaty, T. L., Decision Making with Dependence and Feedback: The Analytic Network Process, RWS Publications, and Pittsburgh, PA, 1996.

[4] Sattinger, M. (1993), "Assignment Models of the Distribution of Earning," Journal of Economic Literature, 31, pp. 831-880.

[5] Woodside, A. G., Trappey III, R. J. (2001), "Learning Why Some Customers Shop at Less Convenient Stores," Journal of Business Research, Vol. 54, pp. 151-159.

Ben-Jeng Wang, Tunghai University, Taichung, Taiwan

Maw-Yang, National CHIN-YI Institute of Technology, Taichung, Taiwan

AUTHOR PROFILES:

Dr. Ben-Jeng Wang earned his doctoral degree from Graduate School of Industrial Engineering, Lamar University, Beaumont, TX, USA. Currently, he is teaching operations and information management courses in the Department of Business Administration, and also serves as chairman of Institute of Continuing Education both in Tunghai University, Taiwan, ROC.

Dr. Maw-Yang Hsu earned his master degree from School of Industrial Engineering and Management, Cheng Kung University, ROC. He is teaching distribution management and information management course in the department of distribution management in National Chinyi Institute Of Technology, ROC.