SELECTION OF THE BEST RIDE-HAILING BY USING ANALYTIC HIERARCHY PROCESS

The online-based or ride-hailing service business in Indonesia continues to develop and metamorphose into multifunctional transportation services, this development makes it difficult for customers to determine alternatives to the top two ride hailing services, namely Goride and Grabbike. For this reason, the purpose of this study is to make decisions based on performance, service, security, price, promo and branding criteria to determine the best and make it easier to choose the transportation. The method used is the Analytic Hierarchy Process (AHP) model with the help of Super Decisions software. The data sampling technique was carried out by distributing online questionnaires of 100 respondents to the public using transportation services. The results of this research indicate criteria and alternatives. Shows the Safety Criteria 36.02%, Service Criteria 23.73%, Price Criteria 16.36%, Promo Criteria 11.29%, Performance Criteria 7.79%, Branding Criteria 4.81%. It seems that security is the most important factor for potential customers because it is related to customer safety, while the two most popular alternatives to Goride and Grabbike are Grabbike. Where GrabBike is superior in terms of cheap prices per kilometer and good branding and everywhere, while Gojek excels in terms of services provided to customers.


Introduction
The increasing number of features on smartphones that are embedded with GPS, coupled with the availability of a digital road map via an API (application programming interface) provides the technology support needed for ordering ride-hailing services. A common feature of ridehailing services is the ability of customers to request drivers and vehicles via a smartphone application that provides the customer's location to the driver via GPS. With the support of GPS technology, digital maps, and routing algorithms, users are provided with real-time waiting time information. (Clewlow, Regina R.Mishra, 2017) As time went on, the competition between GoJek and Grab became increasingly fierce. Both of them salute each other in developing services and improving their quality. The hope, of course, can get more users than competitors. However, in the end it is the consumer who decides. This is because the Indonesian market is very broad. You can imagine if the total population of Indonesia will reach 270 million by 2020. based on the calculations of the Central Bureau of Statistics, of course it will be a special attraction for business people, especially in the transportation industry. This is because a very large population is clearly accompanied by a high figure for consumption and mobility needs. With a large population of Indonesia, of course it is a special attraction for business people, especially in the transportation industry. This is because a very large population is clearly accompanied by a high figure for consumption and mobility needs. That is why the online transportation ride-hailing industry is flourishing these days. Many players have sprung up, although currently there are only two, namely Go-Jek, a unicorn start-up from Indonesia and Grab, the largest ride-hailing service provider in Southeast Asia. PT GoJek Indonesia and Grab Bike have taken advantage of this opportunity to advance, which can be a This strong competition causes difficulties and confusion for consumers in determining alternatives to the criteria for choosing the two best ride-hailing options. A decision support system (DSS) is a system that can help someone make decisions accurately and in accordance with the desired goals, for that in this study it helps customers determine alternatives to Goride or Grabbika with the criteria of Security, Service, Price, Promo, Performance, The method used is the AHP (Analytical Hierarchy Process) method with the help of software super decisions. . (Saaty, 2005).

2.
Research Method Decision making is a choice based on certain criteria on two or more possible alternatives (Syamsi, 2000). States that in their life, human beings frequently face conditions in making decision under some alternative criteria. The AHP method can provide optimum solution in transparent way through: (a) Analysis of the decision in qualitative and quantitative way, (b) Evaluation and representation of the solution in simple way through hierarchical model, (c) Logical argumentation, (d) Test of quality of the decision, and (e) Shorter time reqired (Saaty, 1994).
The AHP method consists of three levels of hierarchy. The first hierarchy level is the goal of the decision making, the second level of hierarchy is how each of the existing criteria contributes to the goal achievement, and the last level of hierarchy is to find out how each of the alternatives contributes to each of the criteria (Exhibit 1). (Weinwurm 1961)Taylor (2002 explains that the steps in decision making by using the AHP method are as follows: 1) Establishing Pairwise Comparison Matrix for each decision alternative to each criteria 2) Synthesization 3) Establishing Pairwise Comparison Matrix for each criteria 4) Establishing The Normalized Matrix 5) Establishing The Preference Vector 6) Calculating overall value for each decision alternative 7) Determining the rank of alternatives according to the value acquired in the previous step  Saaty (1994: 202) states that there are three basic principles in the AHP method, which are as follows: 1) Decomposition After the problem has been defined, decomposition is necessary to to be done, which is dividing a problem into some smaller parts. The division process will result some levels of a problem. That's why this process of analysis is named hierarchy.

2)
Comparative Judgment This principle assesses the relative importance of two elements in a certain level related to those at higher level. This assessment is the main point of the AHP method because it influences the priority of the elements. This assessment result can be observed better if displayed in the form of Pairwise Comparison Matrix.

3)
Synthesis of Priority From each of Pairwise Comparison Matrix, the eigenvector value can be determined to acquire local priority. Because the Pairwise Comparison Matrix is available in each level, the global priority can be acquired by synthesizing between those local priorities. The procedure of ynthesizing is different according to each hierarchy. To rank the elements according to its relative importance through synthesizing procedure is called priority setting.
According to Saaty (1994:203), this AHP method is appropriate to be used in making decision that involves decision element comparison, which is difficult to be assessed quantitatively. This matter is based on the assumption that human beings' natural reaction when facing a complex decision making, is by grouping the decision elements according to its common characteristics. This grouping process includes rank the decision elements, and then comparing between each pair in each group in a form of matrix. Afterward, inconistency ratio and weight for each element will be acquired. Thus, it will provide ease in testing the data consistency.
The ratio-scale form is used as an input in the AHP method, which states one's perception when facing the decision-making situation. The values in the ratio are then organized in a matrix, which is called the pairwise comparison matrix. Due to the limitation of human beings' brain capability, the ratio-scale is limited as well. In the AHP method, the scale range 1-9 is assumed sufficiently representing human beings' perception. The reason why te AHP method limits the ratio-scale 1-9, is acording to the research conducted by a psychologist (Miller, 1956: 256), which shows that human beings cannot simultantly compare more than seven objects, either it increases or decreases two objects. In such condition, human beings will lose their consistency in making the comparison. The Standard Preference Scale used in the AHP method is provided in Table 1 as follows:

Types of research
This research is a descriptive quantitative study, the model used in decision support in this study is the Analytical Hierarchy Process method. While the data collection technique used in this study used an instrument in the form of a questionnaire. A sample that can represent the population is called a representative sample. Representative samples have the same or relatively the same characteristics as the population characteristics. Representative level of the sample taken from a particular population depends on the type of sample used, the size of the sample taken, and the method of collection (Taherdoost 2016). Sample is a set or part of the number and characteristics possessed by the population. Sampling technique is a sampling technique, where the researcher uses purposive sampling, a technique of determining the sample based on chance, that is, anyone who happens to meet the researcher can be used as a sample, if it is considered that the person met is suitable as a data source. (Etikan 2016)

Time and Place of Research
This research was conducted in Jakarta while the questionnaires were distributed online in the Jakarta, Bekasi, Depok, Bogor areas. The time of this research is November 2019-February 2020. From the input process the value obtained from the questionnaire shows the comparison between the criteria, where the eigen value for branding is 0.04812, the eigen value for performance is 0.07786, the eigen value for price is 0.16366, the eigen value for promotion is 0.11288, the eigen value for security is 0.36021, the eigen value for service is 0.23726. It can be seen that the inconsistency value of 0.09530 is still below 0.1 based on the consistency ratio (CR).  Priorities for criteria, and alternatives. Shows that the Security Criteria are 36.02%, Service criteria 23.73%, Price Criteria 16.36%, Promo Criteria 11.29%, Performance Criteria 7.79%, Branding criteria 4.81%. For an alternative, it looks like Grabbike is worth 0.59 and Gojek is 0.41.

Figure 11.
Overal priorities for the "Ride Hailing" the highest is Grabbike.

Conclusion
Based on the research that has been done, the following conclusions can be drawn, criteria for online transportation voters are: 36.02% security criteria, 23.73% service criteria, 16.36% price criteria, 11.29% promo criteria, 7.79% performance criteria, branding 4.81%. Furthermore, the order of priority for online ojek transportation voters, namely the most important alternative is: GrabBike online transportation received 50.33%, Gojek online ojek transportation gained 49.67%.

Suggestion
First, further research can expand the object of research, for example, for business people, school children or others. Second, use the TOPSIS method to strengthen research.