Analysis of Cross-border E-commerce Operating Performance Based on FAHP

Cross-border E-commerce has been an important engine to push bilateral economic prosperity. Evaluating its performance involves multiple criteria in legal, economics, and techniques. These criteria have the characteristics of fuzziness, complexity and diversity. The paper sets up a criteria hierarchy with four primary criteria and twenty secondary criteria to evaluate operating performance on Sino-Australian cross-border E-commerce using the fuzzy analytic hierarchy process (FAHP) method. The results show that the overall operating performance in SinoAustralian cross-border E-commerce is good; however, the most influence of the performance is from government policies rather than from techniques.


I. INTRODUCTION
With the rapid development of globalisation and integration of world economic, cross-border E-commerce has become a burgeoning model of E-commerce.By the prediction of Chinese Ministry of Commerce, the total value of import and output from E-commerce will approach to 6500 billion Chinese Yuan in 2016.In the nextfew years, the contribution of cross-border Ecommerceto total value of import and export will up to 20%, andthe annual averaged growth rate will over 30%.
As a new trade model, cross-border E-commerce has been the important engine to push economy development.So the importance of measuring the operating performance cannot be overemphasized.In view of its fuzziness, complexity and diversity, the paper will use FAHP to evaluate the operating performance on Sino-Australian cross-border E-commerce.
The reminder of the paper is organized as follows.Section 2 overviews the hierarchy of evaluation criteria.Section 3 discusses FAHP.Section 4 focuses on the application on a cross-border E-commerce.Section 5 provides analysis of the evaluation outcome.Finally, Manuscript received October 15, 2016; revised November 10, 2016.Section 6 presents the conclusion and directions for further steps of this study.There have been many studies on multi-criteria analysis on operating performance.Cho and Park regard waiting time and cost of delivery as most important criteria [1].Devaraj emphasizes practicality and reliability [2].Ulusoy states that logistics cost, production capacity, and information sharing are essential criteria [3].Studies such as Sukati et al. [4], Jabbour et al. [5], Ghobakhloo et al. [6], Gilaninia et al. [7] specifically focus on information technology.Furthermore, many business have proven that sharing information is most essential (Richey Jr et al. [8];Fawcett et al. [9]).Zulkiffli [10], Sinem [8] and Battistoni et al. [11] all argueoperation capabilities is the key criteria.Chinomona [12] argues logistic integration, IT and strategic purchasing are very important for E-commerce.Zuo [13] mainly concentrates on the cost and the delivery of the convenience and speed.

II. EVALUATION CRITERIA
In order to select objective evaluation criteria for Sino-Australian cross-border E-commerce, this study invited four experts who are two managers of logistic operations (each from a company in farming product imports and exports in NSW, Australia and a cross-border Ecommerce business in Chongqing, China) and two researchers on cross-border E-commerce (each from a university in Australia and in Chongqing, China).This study chooses four primary evaluation criteria which are ``Operating Capacity'', ``Rules and Policies'', ``Service Quality'', and ``Logistics Cost''.Each of these has several secondary criteria.The hierarchy of the primary criteria and secondary criteria is listed in Table I.

III. FAHP EVALUATION MODEL
FAHP method tries to solve objective selection and justification problems by using fuzzy sets which describe uncertainties in those problems and process hierarchical evaluation criteria or attributes.Many variants of the FAHP method have been presented, studied and applied.In this paper, we used the method presented in [14].

A. Fuzzy Number
Suppose is a fuzzy number on a universe of discourse .If its membership function is Then is called a triangular fuzzy number (TFN) and denoted by .Let and be two TFNs, then we define .

TFN
is called the reciprocal TFN of in the sense that .The comparison scales used in this paper are TFNs shown in Table II.

B. FAHP
The FAHP method contains two main modules: 1) calculate weight distribution on criteria, and 2) aggregate evaluation from bottom to top.

1) Calculate weight distribution
Weight distribution describes the relative importance degree of each criterion compared with other criteria under the same parent criterion.Considering the given evaluation criteria hierarchy in Table I, we need to identify five weight distributions in total which are: one for four primary criteria and four for their secondary criteria.Weight distribution is calculated through pairwise comparison matrices.
Suppose is the index of a criterion in TableI.

2) Aggregate evaluation
Aggregating evaluation is an iterative process from bottom to top along the hierarchy of criteria.The bottom level input includes an evaluation matrix for objectives with respect to bottom level criteria.The aggregated output from the bottom level is taken as the input of the criteria at successive level along the hierarchy of criteria.This procedure stops until the top level criterion, which is always the ultimate decision or evaluation goal.
Suppose is a set of criteria with corresponding weights and is a set of objectives to be evaluated.Let be a fuzzy evaluation matrix where is the evaluation score on objective regarding to criterion , ; .Then the aggregation output is calculated by where means the transpose of vector, and is a pair of operators to fuse information in and.
Commonly-used operator pairs for are , , and and they are defined as The output indicates an evaluation on objectives regarding the parent criterion of .We can repeat this process until the ultimate decision or evaluation goal.
In some situations, an evaluation involves many experts and each expert provides individual evaluation ; therefore, we treat experts as criteria in a special twolevel hierarchy and use the process aforementioned to get the output of the top level of it, which is the R required.

IV. CASE STUDY
In this study, we invited four experts to evaluate sino-Australian cross-border E-commerce progress based on the criteria given in TableI.These experts compared those criteria and presented the fuzzy comparison result as shown in Table III.

A. Identify Criteria Weights
Because there are four experts, we will integrate the four fuzzy comparison matrices to one, i.e., conducting the Step 0 before obtaining the weight distribution on the primary criteria.By setting and using Equation (3), we get the integrated fuzzy comparison matrix as shown in Table IV.

B. Evaluation on Cross-border E-commerce
After identifying the weights of all criteria, we can build evaluation matrix against all bottom criteria which are the secondary criteria in Table I.In this case study, we use a five-grade evaluation for cross-border Ecommerce' performance.The five grades are treated as five objectives to select, i.e, = {excellent ( ), good ( ), medium ( ), less ( ), bad ( )}.With regard to the five grades, a evaluation matrix is obtained as shown in TableV .Suppose the weights of those secondary criteria are calculated based on the method in Section 3.B.1; and they are Then based on Equation (11), the aggregated evaluation for , , is calculated.Taking the secondary criteria of ``logistics cost ( )'' as an example, the calculation is Similarly, we obtain the evaluation results of other primary criteria.Now denoting by and using weights of , , and in Eq. ( 14), we obtain the overall evaluation following Equation (11).We test the four operators in Equation ( 12) and summary the outcomes in Table VII.By the maximum membership degree method, all four overall evaluations are the same, i.e. ``good''.

V. ANALYSIS
With the FAHP evaluation, the result indicates:  All four primary criteria passed the consistency verification and the total ranking has satisfiable consistency.The overall operating performance of Sino-Australian cross-border E-commerce is good. Criterion ``Rules and Policies'' has the biggest impact among the four primary criteria.This is particularly conspicuous in China.Recent years, providing preferential policies to cross-border Ecommerce is always an attenuation point of all kinds of governments. Criterion ``Service Quality'' has the most significant impact.The impact is bigger than the granted ``Logistics Cost''.The secondary criterion ``customer satisfaction'' gains the highest mark, which clearly indicates that Sino-Australian crossborder E-commerce can provide sufficient goods and guaranteethe supplying in both countries.However, some ``Service Quality" requirements such as timely delivery, reliability, and valueadded capacity still need to be improved. Criterion ``Operating Capacity'' claims about 28% impact on performance.It shows that limitation exists in enterprise internal operation and there is improvement space.The ``subject quality" is not good enough.The achievement of informationalisition is satisfiable.But the information network is not sound enough and the information transfer is not efficient. The evaluation score of ``Logistics Cost" has remarkable difference from expected.Although the logistics cost of cross-border trading is higher than that of domestic, the real increased cost is acceptable compared with expectation.The secondary criteria ``transportation cost'' and `` warehousing cost'' gain high marks, which indicates the choice of transportation routes is reasonableand the infrastructure facilities and warehouse monitoring are good.However, the marks for ``trading cost'' and ``cost of empty return'' are lower, which shows the drawback of the domestic delivery in both countries and further cooperation needs to be planned more rational to meet the demand of both sides.

VI. CONCLUSIONS AND FUTURE WORKS
In this study, we adopted FAHP model to analyse the operating performance on Sino-Australian cross-border E-commerce with four primary and 20 secondary criteria.
The study provides evident reference to Chinese crossborder E-commerce development.Due to these criteria are artefacts, these criteria are inevitably subjective and incomplete.Regarding to the evaluation result, we claim that the overall operating performance on Sino-Australian cross-border E-commerce is sound.The impact of ``Rules and Policies'' are remarkable.Some ways need to be upgraded further such as quality of services, internal operating procedure, infrastructure facility development, enterprise service platform, etc.

TABLE III .
FUZZY COMPARISON MATRIX FOR PRIMARY CRITERIA

TABLE V .
EVALUATION MATRIX FOR SECONDARY CRITERIA Table VI lists the outcomes.

TABLE VI .
AGGREGATED EVALUATION FOR PRIMARY CRITERIA

TABLE VII .
OUTCOMES OF DIFFERENT OPERATORS.