CAUCHY –Jurnal Matematika Murni dan Aplikasi Volume 7(2) (2022), Pages 220-230 p-ISSN: 2086-0382; e-ISSN: 2477-3344 Submitted: September 19, 2021 Reviewed: December 10, 2021 Accepted: January 07, 2022 DOI: http://dx.doi.org/10.18860/ca.v7i1.13356 Analysis of Insurance Customer Factors to Renewal Using Hybrid AHP-FTOPSIS Kwardiniya Andawaningtyas, Evi Ardiyani*, Corina Karim Departement of Mathemathics Faculty of Mathematics and Natural Science Brawijaya University, 65145, Malang, Indonesia *Corresponding Author Email: eviardiyani98@gmail.com* dina_math@ub.ac.id, corinaub@gmail.com ABSTRACT Human life is full of uncertainties that have enormous risks. Insurance is one way that can help humans reduce this risk. The human need for insurance causes competition among insurance companies in Indonesia to be very competitive. Competition between insurance companies is influenced by several factors, one of the factors is having customers who do insurance renewals. This study aims to determine the factors that influence customers to renew using the Analytical Hierarchy Process (AHP) method and to rank customers' favorite insurance using the Fuzzy Technique for Order Preference by Similarity to Ideal Solution (FTOPSIS) method. The results of the analysis using this method concluded that the main factors that influence customers in making renewals are features with sub-criteria for health protection needs. Meanwhile, the customer's favorite insurance ratings for extending are Takafullink Salam Cendikia with a closeness coefficient of 0.645, Takaful Al-Khairat with a value of 0.563, Takaful Dana Pendidikan with a value of 0.552, and Takafullink Salam with a value of 0.341. Keywords: Insurance; Renewal; AHP; FTOPSIS INTRODUCTION Human life is full of elements of uncertainty that have enormous risks, such as accidents and death. Humans need a guarantee or a method to reduce this risk which we usually call insurance. The human need for insurance causes the competition of insurance companies in Indonesia to be very competitive. The biggest factor for an insurance company to be competitive is a customer who carries out a renewal. Each customer has its own criteria which are the determining factors for a customer to renew. The Decision Support System (DSS) is specific information that is intended to assist management in making decisions related to semi-structured issues. DSS aims to assist decision makers in establishing an unstructured decision. Unstructured decisions have vague problems, and it's difficult to find solutions. Decision support systems are basically designed to support every stage of decision making, namely identifying problems, selecting relevant data, determining approaches, and evaluating alternative choices. In 2018, [1] conducted research on how to improve consumer satisfaction http://dx.doi.org/10.18860/ca.v7i1.13356 mailto:eviardiyani98@gmail.com* mailto:dina_math@ub.ac.id mailto:corinaub@gmail.com Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 221 learning (LBB) in Malang using the DANP-TOPSIS method. Identifying important human error factors in emergency departments in Taiwan using HFACS, AHP, and FTOPSIS by [2]. [3] conducted research on the selection of favorite banks using the AHP and TOPSIS methods. [4] discusses the selection of the best health applications and features that affect the AHP and FTOPSIS methods. Comparasion on of ANP and AHP methods studied by [5]. [6] conducted reseacrh comparison between TOPSIS and SAW. [7]discusses decision making using hybrid AHP-TOPSIS. Hybrid Fuzzy AHP-TOPSIS researched by [8]. [9] researched decision making using hybrid AHP-TOPSIS. Comparison beetwen SAW, AHP, and TOPSIS researched by [10]. [11] conducted research on the comparison between SAW method and AHP method. Integrated ANP and TOPSIS Method for suplier performace assesment researched by [12]. [13] reasearched evaluation of smart and suistainable cities with ANP and TOPSIS method. [14] conducted research hybrid AHP- TOPSIS method under spherical fuzzy sets for system selection. Hybrid AHP-TOPSIS for selecting supplier in construction supply chain researched by [15]. This study aims to determine the factors that most influence insurance customers to renew and obtain favorite insurance alternatives by combining the AHP and FTOPSIS methods. The combination of the AHP and FTOPSIS methods is to obtain the criteria weights using the AHP method, then the FTOPSIS method uses the criteria weights that have been obtained by the AHP method to obtain the best alternative. METHODS Analytical Hierarchy Process (AHP) Method. AHP is a decision-making process with compilation of functional hierarchies with the main input being human [16]. AHP requires ideas from individuals and groups by obtaining their respective assumptions and obtaining the desired solutions. These ideas are used to determine criteria that can solve a problem. In this research, AHP method is used to determine criterion weight to be used in the FTOPSIS method. According [16] there are general measures of AHP Method consists of seven steps. 1. Defining the problem and determining the desired solution then arranging hierarchy of the problems by setting goals which are the overall system goals at the top level. 2. Determine the priority of the elements. a. Making pair comparasons by comparing elements in pairs according to given criteria. b. The pairwaise comparison matrix is filled using numbers to represent the relative importance of one element to another. The pairwise comparison matrix entry is the result of a questionnaire converted using Table 1. 3. Synthesis Considerations for pairwise comparisons are synthesized to obtain overall priority. a. Sum each column on the matrix. b. Divide each value from the column by the total column obtain a normalized matrix. c. Sum each row and divide by the number of elements to get the average value. 4. Measure consistency. a. Multiplies each value in the first column by the relative priority of the first element, the value in the second column by the relative priority if the second element, and so on. b. Adding each row, the result divided by the corresponding relative priority element. Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 222 c. Adding the results above for the elements that exist, called λmaks 5. Calculating the Consistency Index (CI) (1) = number of elements 6. Calculating the Consistency Ratio (CR) (2) Where is Index Random Consistency contained in Table 2 7. Check hierarchy consistency, the consistency ratio must be less or equal to 0.1. The calculation result can be declared correct. Table 1. AHP rating scale Difference -8 -7 -6 -5 -4 -3 -2 -1 0 AHP Scale 9 8 7 6 5 4 3 2 1 Source : [16] Table 2. Index Random Consistency Matrix Size IR Value 1,2 0.00 3 0.58 4 0.90 5 1.12 6 1.24 7 1.32 8 1.41 9 1.45 10 1.49 11 1.51 Source : [16] Fuzzy Technique for Order Preferences by Similarity to Ideal Solution (FTOPSIS) Method. The FTOPSIS Method is a development of the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution). TOPSIS Method first introduced by Yoon and Hwang in 1981. The TOPSIS Method has a weakness, when the decision maker has difficulty determining a value. Therefore, it is necessary to provide an assessment in the form of intervals such as applying fuzzy logic. Fuzzy numbers, linguistic values and membership function shown in the Figure 1 and Table 3. In this research, FTOPSIS method is used to rank the alternatives. Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 223 Figure 1. Fuzzy numbers and linguistic Table 3. The membership function of linguistic value Linguistic Value Fuzzy Number Very low (VL) ( 0 , 0 , 0.2 ) Low (L) ( 0, 0.2, 0.4 ) Medium (M) (0.2, 0.4, 0.6) High (H) (0.4, 0.6, 0.8) Very High (VH) ( 0.6, 0.8, 1 ) Excellent (E) ( 0.8, 1, 1 ) Source : [3] General measurer of FTOPSIS Method consists of 9 steps. 1. Assesing criteria and alternatives Assumed that there are alternatives that will be evaluated against criteria . The weight of each criterion is denoted by . The ranking of the fuzzy criteria value of each decision for each alternative against the criterion denoted by with the membership function . 2. Calculate the comparison value of each criterion and alternatives The fuzzy values for each decision maker are presented as fuzzy triangle The value of the fuzzy ratio is given by , with (3) Fuzzy Number Linguistic Value Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 224 Fuzzy weight ratio with : (4) 3. Make a decision matrix Creating a decision matrix (Dk) that is appropriate for the alternatives to be evaluated based on the following defined criteria : With states the performance of the calculation for i alternatives against the j criterion. 4. Normalize the fuzzy decision matrix Normalize the data using a linear scale transformation, the normalized matrix is defined by (5) with (6) (7) (6) benefit criteria, (7) cost criteria 5. Calculate the normalized matrix weights The normalized matrix weight is calculated by multiplying the weight of the evaluated criterion by the normalized decision matrix (8) with 6. Calculate the value of Fuzzy Positive Ideal Solution (FPIS) and Fuzzy Negative Ideal Solution (FNIS) (9) (10) With and is the set of benefit criteria. and is the set of cost criteria. 7. Calculate the distance for each alternatives from FPIS and FNIS If there is and is two fuzzy triangular numbers, defined as and then the distance between and can be calculated by Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 225 (11) Distance of each weighted alternative (i = 1, 2, 3, … , m) from FPIS and FNIS can be calculated by (12) (13) 8. Calculate the closeness coefficient value The Closeness Coefficient ( ) represents the distance between FPIS(A+) and FNIS (A- ) simultaneously for each alternative, the Closeness Coefficient ( ) can be calculated by (14) With 9. Sort alternative Each alternative is sorted according to the decreasing Closeness Coefficient ( ) value. The best alternatives is the Closeness Coefficient ( ) value is close to FPIS and far from FNIS. RESULTS AND DISCUSSION Analytical Hierarchy Process (AHP) Method. The first step in the AHP Method is arrangement the hierarchichal structure. The hierarchical structure in this study consists of 4 levels, the first level is the goal, namely to determine the favorite type of insurance. The second level is the elaboration of the main aspects that influence the objectives, namely the criteria. The third level is the aspects that influence the criteria, namely sub-criteria. The fourth level or the lowest level is the level that consists of alternatives. The structure of the hierarchical system in this study can be seen in Figure 2. Then, we determine the priority of the elements by create formation of Pairwise Comparison Matrix between sub-criteria and create weight matrix between sub-criteria based on the results of the questionnaire. Next step is calculate value. The five criteria have a , it can be concluded that the pairwise comparison matrix between these subcriteria is consistent. The most influential criterion in choosing the customer’s favorite insurance for renewal in company A is the insurance feature with the subcriteria for the need for health protection having a weight value of 0.800. Table 5 shows the evaluation result and final ranking of criterion. To determine the level of data consistency, we calculate the value. First, calculate the value of then calculate the using equation (1). is obtained by adding the results for the elements that exist and each number of subcriteria is the value of n used. Table 5 shows the results. Using the value that has been obtained, calculate the value using equation (2). If the the research can be continued. Table 4 Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 226 shows the value. It is shown that the five criteria have a value of less than 0,1 which means that the data for the five criteria are consistent. So, research can be continued. Table 4. value Criteria value Company Image 0,068 Agent 0,034 Insurance Features 0,064 Claim 0,020 Income 0,055 Figure 2. Hierarchical System Table 5. AHP Method Result Subcriteria (Priority Vector) Honesty 0.639 Achievment 0.087 TDP TLS TLSC TAK Customer Income (Rp. 2,5 - 4,9 million /month) Customer Income (Rp. 5,0 - 7,5 million /month) Customer Income (>Rp. 7,5 million /month) Goal Favorite type of insurance and factors that influence customers to renew Criterion Sub-Criterion Alternative Agent Product Mastery Communication Ease of Contact Honesty Achievment Track Record Insurance Features Health Education Investation Claim Ease of Taking Claims Great Claim Company Image Time Period for Claiming Income Customer Income (Rp. 7,5 million/month) 0.244 Afterward, we analyze the best alternative in FTOPSIS Methods. The weights of criteria to be used in evalution process are calculated by using AHP Method combined with the scores from the expert questionnaire. Table 6 shows the data from the expert questionnaire. Table 6. Data from the expert questionnaire TDP TLS TLSC TAK Honesty H VH M H Achievment VH VH H M Track Record M H VH L Product Mastery L VH M M Communication H L VH H Ease of Contact M L L M Health H H VH E Education E VH H H Investation H VH E H Ease of Taking Claims VH H H VH Great Claim M L VH H Time Period for Claiming VH H VH H Customer Income (Rp. 7,5 million/month) H H M H Then the next step is calculating the weight of the alternative matrix. Table 7 shown the multiplication results of the expert questionnaire values that have been converted based on table 3 with the priority vector value For example, criteria Honesty on Alternative TDP is H then convert the value to fuzzy number based on Table 3 which is (0.4, 0.6, 0.8). Then do fuzzy multiplication with the value of the priority vector which is 0.639. After we get the multipclication of the priority vectors and the expert quiestionner, we calculate the FPIS and FNIS values, then we use these values to calculate the FPIS and FNIS distances using equation (12) and (13). Table 8 shows the value of the FPIS and FNIS distance. After calculating the distance between FPIS and FNIS, we calculate value using equation (14). Table 9 shows the results of calculating the value. Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 228 Depends of the value in Table 9 the alternatives ranking in FTOPSIS method, The first order is the TLSC alternative, the second is the TAK alternative, the third is the TDP alternative, and the last order is the TLS alternative. Table 7. Multiplication of the priority vectors by the results of the expert questionnaire TDP TLS TLSC TAK Honesty (0.256,0.384,0.511) (0.384,0.511,0.639) (0.128,0.256,0.384) (0.256,0.384,0.511) Achievment (0.052,0.070,0.087) (0.052,0.070,0.087) (0.035,0.052,0.070) (0.017,0.035,0.052) Track Record (0.055,0.109,0.164) (0.109,0.164,0.219) (0.164,0.219,0.274) (0,0.055,0.109) Product Mastery (0,0.069,0.137) (0.206,0.274,0.343) (0.069,0.137,0.206) (0.069,0.137,0.206) Communication (0.230,0.345,0.460) (0,0.115,0.230) (0.345,0.460,0.575) (0.230,0.345,0.460) Ease of Contact (0.016,0.033,0.049) (0,0.016,0.033) (0,0.016,0.033) (0.016,0.033,0.049) Health (0.320,0.480,0.640) (0.320,0.480,0.640) (0.480,0.640,0.800) (0.640,0.800,0.800) Education (0.099,0.124,0.124) (0.075,0.099,0.124) (0.050,0.075,0.099) (0.050,0.075,0.099) Investation (0.030,0.045,0.060) (0.045,0.060,0.075) (0.060,0.075,0.075) (0.030,0.045,0.060) Ease of Taking Claims (0.074,0.098,0.123) (0.049,0.074,0.098) (0.049,0.074,0.098) (0.074,0.098,0.123) Great Claim (0.446,0.557,0.557) (0,0.111,0.223) (0.334,0.446,0.557) (0.223,0.334,0.446) Time Perios for Claiming (0.192,0.256,0.320) (0.128,0.192,0.256) (0.192,0.256,0.320) (0.128,0.192,0.256) Customer Income (Rp. 7,5 million/month) (0.098,0.146,0.195) (0.098,0.146,0.195) (0.049,0.098,0.146) (0.098,0.146,0.195) Table 8. FPIS and FNIS distances. TDP TLS TLSC TAK 0.982 1.445 0.787 0.958 1.207 0.748 1.429 1.234 Table 9. value Alternative TDP 0.552 TLS 0.341 TLSC 0.645 TAK 0.563 CONCLUSION The results of data analysis that has been carried out from the combination of the two methods indicate that the most influencing factor for insurance customers to renew at PT Asuransi Takaful Keluarga is the insurance feature, namely the customer's need for health protection with weight value of 0.800. All health insurance companies must have health protection features. So, we see the next order of sub-criteria, honesty with a weighted value of 0.639, agent communication with a weighted value of 0.575, and good claims with a weighted value of 0.557. The sub-criteria that have the highest value Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 229 weight are the criteria for company image, agent, and claims. The five criteria have a priority value that is quite close, indicating that the five criteria are mutually sustainable.The order of alternative choices for the customer's favorite insurance who renews at PT. Family Takaful Insurance is Takafulink Salam Scholar, Takaful Al-Khairat, Takaful Fund Education, and Takafulink Salam. REFERENCES [1] K. Andawaningtyas, E. W. Handamari, and C. Karim, “Improving the Guidance Learning (LBB) Consumer Satisfaction in Malang using DANP - TOPSIS method,” Cauchy, vol. 5, no. 3, p. 117, 2018, doi: 10.18860/ca.v5i3.5541. [2] M. chih Hsieh et al., “Application of HFACS, fuzzy TOPSIS, and AHP for identifying important human error factors in emergency departments in Taiwan,” Int. J. Ind. Ergon., vol. 67, pp. 171–179, 2018, doi: 10.1016/j.ergon.2018.05.004. [3] H. I, “Analisis Keputusan Pemilihan Bank Favorit Menggunakan Kombinasi Metode AHP dan Metode TOPSIS,” 2019. [4] M. Rajak and K. Shaw, “Evaluation and selection of mobile health (mHealth) applications using AHP and fuzzy TOPSIS,” Technol. Soc., vol. 59, 2019, doi: 10.1016/j.techsoc.2019.101186. [5] A. J. Olanta, M. E. Sianto, and I. Gunawan, “Perbandingan Metode ANP Dan AHP Dalam Pemilihan Jasa Kurir Logistik Oleh Penjual Gadget Online,” Widya Tek., vol. 18, no. 2, pp. 96–101, 2019, doi: 10.33508/wt.v18i2.2275. [6] Sunarti, “Comparison Beetwen TOPSIS and SAW Method in The Selection Of Tourist Destinations in West Java,” Techno COM, vol. 18, 2019. [7] R. Agusli, M. I. Dzulhaq, and F. C. Irawan, “Sistem Pendukung Keputusan Penerimaan Karyawan Menggunakan Metode Ahp-Topsis,” Acad. J. Comput. Sci. Res., vol. 2, no. 2, 2020, doi: 10.38101/ajcsr.v2i2.286. [8] V. Julianto, H. S. Utomo, and H. Herpendi, “Analisis dan Penerapan Metode Fuzzy AHP-TOPSIS dalam Penentuan Mitra Industri Sebagai Tempat Praktek Kerja Lapangan,” J. Ilm. Inform., vol. 5, no. 2, pp. 108–121, 2020, doi: 10.35316/jimi.v5i2.942. [9] G. Surya Mahendra and I. P. Y. Indrawan, “Metode AHP-Topsis Pada Sistem Pendukung Keputusan Penentuan Penempatan Automated Teller Machine,” J. Sains dan Teknol., vol. 9, no. 2, pp. 130–142, 2020. [10] Wawan Firgiawan, Sugiarto Cokrowibowo, and Nuralamsah Zulkarnaim, “Komparasi Algoritma Saw, Ahp, Dan Topsis Dalam Penentuan Uang Kuliah Tunggal (Ukt),” J. Comput. Inf. Syst. ( J-CIS ), vol. 1, no. 2, pp. 1–11, 2019, doi: 10.31605/jcis.v1i2.426. [11] Qiyamullaily Arista, Nandasari Silvia, and Amrozi Yusuf, “Perbandingan Penggunaan Metode Saw Dan Ahp UntukSistem Pendukung Keputusan Penerimaan Karyawan Baru,” Tek. Eng. Sains J., vol. 4, 2020. [12] C. Natalia, I. P. Surbakti, and C. W. Oktavia, “Integrated ANP and TOPSIS Method for Supplier Performance Assessment,” J. Tek. Ind., vol. 21, no. 1, pp. 34–45, 2020, doi: 10.22219/jtiumm.vol21.no1.34-45. [13] G. Ozkaya and C. Erdin, “Evaluation of smart and sustainable cities through a hybrid MCDM approach based on ANP and TOPSIS technique,” Heliyon, vol. 6, no. 10, 2020, doi: 10.1016/j.heliyon.2020.e05052. [14] M. Mathew, R. K. Chakrabortty, and M. J. Ryan, “A novel approach integrating AHP and TOPSIS under spherical fuzzy sets for advanced manufacturing system selection,” Eng. Appl. Artif. Intell., vol. 96, 2020, doi: Analysis of Insurance Customer Factors to Renewal using Hybrid AHP-FTOPSIS 230 10.1016/j.engappai.2020.103988. [15] M. Marzouk and M. Sabbah, “AHP-TOPSIS social sustainability approach for selecting supplier in construction supply chain,” Clean. Environ. Syst., vol. 2, p. 100034, 2021, doi: 10.1016/j.cesys.2021.100034. [16] Kusrini, Konsep Dan Aplikasi Sistem Pendukung Keputusan. 2007.