從大數據創造價值 : 金融產業的多個案研究 - 政大學術集成

全文

(1)國立政治大學資訊管理學系. 碩士學位論文指導教授: 尚孝純博士. 立. 政治大. ‧ 國. 學. 從大數據創造價值:金融產業的多個案研究. ‧. y. Nat. Generating Value from Big Data:. n. sit. er. io. A Multiple Case Study in Financial Industry al v i n Ch engchi U. 研究生：蔡佑晟中華民國一○七年七月 DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(2) Acknowledgement After finishing my first complete research, I really gained a lot of knowledge and broadened my vision. However, I have never thought that I could finish a complete study so successfully, and in fact, I had really gotten many help from many people during this research. First, I would like to express my sincerest appreciate to my instructor, Dr. Shari S. C. Shang, who treated me as her own child, taught me a lot of knowledge in these years, and guided me how to do research. Without her suggestions during the whole research process, I would not be able to overcome many difficulties while researching. I feel so lucky to have the chance to be her student.. 政治大. Second, I would like to thank for the useful advices and encouragement from my thesis committees, Dr. Ya-Ling Wu and Dr. Yu-Ju Tu, which significantly improved my research process and my manuscript.. 立. ‧ 國. 學. ‧. Third, it is no doubt that this study could not be finished without the help of the senior IT managers from the selected case companies. I really appreciate the generous support of selected case companies and the sharing willingness of interviewee.. y. Nat. sit. n. al. er. io. Fourth, I am also very grateful for all lovely people who had helped me to get in touch with the case companies and interviewees, including my mother, the famous lawyer Jack J.T. Huang and his wife, my friends Li-Cheng Li, Chi-Ten Huang, Lucy. Ch. i n U. v. Chang, Kao-Chin Chen, Erh-Chin Ni, Andre Liu, and so forth.. engchi. Fifth, thanks for my parents for raising me up and supporting me to study. I cannot have today's achievement without their love, and their love is always what I most cherished. Finally, special thanks to my girlfriend Yong-Jyun Jhu, my family, Dr. Howard H. C. Chuang, Dr. Yen-Chun Chou, and all of my friends who had accompanied me, encouraged me, supported me, and gave me suggestions throughout these years. All of you created many wonderful memories for me, and it is my honor to share this achievement with all of you, thank you.. I. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(3) Abstract With rapid advances in social and analytics technology, Big Data has become a popular subject in many industries. Numerous well-known multinational companies, such as Google, Walmart, and Amazon, reported deriving enormous value from Big Data applications. However, except for these emblematic examples, there is no promise that large investments in Big Data can result in material benefits. Business decisionmakers remain doubtful as to returns from Big Data technologies. Performing a systematic literature review of data-driven decision-making cases and Big Data applications, this study identified several possible benefits that may be generated from Big Data, and identified several Big Data-related key factors that may affect value creation. Then, this study conducted in-depth interviews with senior IT managers from selected financial companies. Finally, by a cross-sectional analysis of financial industry, this study intends to provide insights into Big Data implementation.. 立. 政治大. ‧. ‧ 國. 學. Key Words: Big Data, Business Value, Data-Driven Decision-making. n. al. er. io. sit. y. Nat. This paper has been accepted as “Generating Value from Big Data: A Multiple Case Study in Financial Industry,” Proceedings of International Conference on Business and Information (BAI 2018), July 6-8, 2018, COEX Convention Center, Seoul, Korea. Ch. engchi. i n U. v. II. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(4) 摘要隨著社群和分析技術的快速發展，大數據已經成為許多產業中的熱門議題。眾多知名跨國公司都從大數據應用中獲得了巨大的價值，如: Google、Walmart、和 Amazon 等。然而除了這些具有代表性的成功例子之外，在大數據上進行大量的投入並不一定能帶來實質的收益，商業決策者們仍然對大數據科技的回報持懷疑態度。本研究對數據導向決策的案例和大數據應用進行了系統化的文獻回顧，並歸納出大數據可能創造的效益，以及一些可能影響大數據創造價值的相關關鍵因素。之後，此研究將與金融公司的高階資訊主管進行深度的訪談與了解。最後，本研究透過對金融產業的跨個案橫向分析，期望能為大數據的應用提供相關的發現與見解。. 政治大關鍵字: 大數據, 商業價值, 數據導向決策立. ‧ 國. 學 ‧. 本文已被接受為 “Generating Value from Big Data: A Multiple Case Study in Financial Industry”，國際商管與資訊研討會 (BAI 2018), 2018 年 7 月 6 日至 8 日，韓國首爾 COEX 會議中心. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. III. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(5) Contents Acknowledgement ......................................................................................................... I Abstract ........................................................................................................................ II 摘要.............................................................................................................................. III Contents ...................................................................................................................... IV Figures......................................................................................................................... VI Tables .......................................................................................................................... VI Chapter 1: Introduction .............................................................................................. 1 1.1 Research Background ........................................................................................... 1. 政治大 1.1.2 Big Data ......................................................................................................... 1 立 1.1.1 Data-Driven Decision-Making ...................................................................... 1. ‧ 國. 學. 1.1.3 Big Data Application in Industries ................................................................ 3 1.2 Research Motivation and Research Purpose ........................................................ 5. ‧. Chapter 2: Literature Review ..................................................................................... 7. y. Nat. 2.1 Measuring Possible Benefits of Big Data ............................................................ 7. io. sit. 2.1.1 Three Categories of IT-enabled Business Values ......................................... 7. n. al. er. 2.1.2 Possible Benefits of Big Data ........................................................................ 8. i n U. v. 2.2 Factors Affecting Big Data................................................................................. 10. Ch. engchi. 2.2.1 Accessibility, Timeliness, and Quality of Data ........................................... 10 2.2.2 Data Policy .................................................................................................. 11 2.2.3 Staff Capacity and External Support ........................................................... 12 2.2.4 Tools and Technology ................................................................................. 13 2.2.5 Organizational Culture and Leadership ....................................................... 15 Chapter 3: Research Methodology ........................................................................... 18 3.1 Research Design and Process............................................................................. 18 3.2 Data Collection .................................................................................................. 19 3.2.1 Select Criteria.............................................................................................. 19 3.2.2 Case Description ......................................................................................... 20 IV. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(6) Chapter 4: Research Results ..................................................................................... 23 4.1 Case Study ......................................................................................................... 23 4.1.1 Case Study of Company A .......................................................................... 23 4.1.2 Case Study of Company B .......................................................................... 24 4.1.3 Case Study of Company C .......................................................................... 25 4.1.4 Case Study of Company D .......................................................................... 26 4.2 Cross-case Analysis - Benefits Created from Big Data ..................................... 29 4.2.1 Informational Benefits ................................................................................ 29 4.2.2 Transactional Benefits................................................................................. 30 4.2.3 Strategic Benefits ........................................................................................ 31. 政治大. 4.3 Cross-case Analysis - Key Factors that Enable Big Data to Create Business Value ........................................................................................................................ 34. 立. 4.3.1 Accessibility, Timeliness, and Quality of Data ........................................... 34. ‧ 國. 學. 4.3.2 Data Policy .................................................................................................. 35 4.3.3 Staff Capacity and External Support........................................................... 36. ‧. 4.3.4 Tools and Technology ................................................................................. 37. sit. y. Nat. 4.3.5 Organizational Culture and Leadership ...................................................... 38. io. er. Chapter 5: Conclusion ............................................................................................... 41 5.1 Summary ............................................................................................................ 41. al. n. v i n 5.2 Contribution ....................................................................................................... 41 Ch engchi U 5.3 Limitation and Future Research ......................................................................... 42 Reference .................................................................................................................... 43 Appendix A: Semi-Structured Questionnaire ......................................................... 48. V. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(7) Figures Figure 1-1 Properties of Big Data .................................................................................. 3. Tables Table 2-1 Three Categories of IT-enabled Business Values ......................................... 8 Table 2-2 Possible Benefits of Big Data ...................................................................... 10 Table 2-3 Factors Affecting Big Data .......................................................................... 16 Table 3-1 Description of the research process ............................................................. 19. 政治大 Table 4-1 Summary of Big Data implementation in selected cases ............................ 28 立 Table 4-2 Summary of Benefits Created from Big Data ............................................. 33 Table 3-2 Description of selected cases ....................................................................... 22. ‧ 國. 學. Table 4-3 Summary of Key Factors that Enable Big Data to Create Business Value . 40. ‧. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. VI. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(8) Chapter 1: Introduction 1.1 Research Background 1.1.1 Data-Driven Decision-Making Data-driven decision-making is a process of transforming raw data into meaningful information and information into knowledge that can be applied to make decisions (Mandinach et al., 2006). The primary concept of data-driven decision-making involves making decisions based on data analysis rather than experience or intuition (Provost and Fawcett, 2013). The world will undoubtedly become more data-driven, as data are very useful and valuable (Lohr, 2012). Rather than guessing, decision-making can become more datadriven in many areas, as data availability is greatly improved compared to the past (Jagadish et al., 2014). Organizations generate and gather large quantities of information in the course of everyday business that can be beneficial and valuable (Kabir and Carayannis, 2013). Due to the large scale of available data, companies in many industries are interested in using data to improve decisions and gain competitive advantages (Provost and Fawcett, 2013). From today’s perspective, decisions made. 立. 政治大. ‧ 國. 學. ‧. without consulting data seem likely to fail (Shim et al., 2015), i.e., data-driven decisions are usually better decisions (McAfee et al., 2012).. n. al. er. io. sit. y. Nat. 1.1.2 Big Data Due to technology advances, data is much more available than in the past (Jagadish et al., 2014). Data are becoming not only easier to access but also more amenable to computer processing as mathematical models and computer capabilities improve (Lohr,. Ch. i n U. v. 2012). Advances in computer capabilities enable more sophisticated data analysis (Provost and Fawcett, 2013), resulting in a tendency towards data-driven discovery and Big Data-enabled decision-making (Lohr, 2012). As many prior studies have reported, Big Data can be defined by the following properties in figure 1-1 (Ding et al., 2014; Gandomi and Haider, 2015; Kitchin and McArdle, 2016; Sivarajah et al., 2017). The volume relates to a large scale of datasets,. engchi. a common challenge posed by Big Data (Sivarajah et al., 2017). Velocity relates to fast or real-time processing needed for quick decision-making, which is associated with the data collection speed, data transferring reliability, data storage efficiency, and excavation speed of discovering knowledge (Chen et al., 2015; Zhong et al., 2016). Variety relates to the diversity of Big Data sources, which includes structured, semistructured, and unstructured data (Gandomi and Haider, 2015; Kitchin and McArdle, 2016). Veracity relates to accuracy, truthfulness, and precision of data (Shim et al., 2015). Value relates to economic benefits of Big Data (Wamba et al., 2015). Viability 1. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(9) relates to making forecasts based on all relevant information (Fouad et al., 2015). Visualization relates to reader-friendly presentation of data (Wamba et al., 2015). Variability relates to inconsistencies in data flow (Katal et al., 2013). Many studies defined that Big Data means having data on a scale that is too large and difficult to process efficiently by conventional analysis tools (Katal et al., 2013; Madden, 2012; Provost and Fawcett, 2013). A survey made by IBM Institute for Business Value and Saïd Business School at the University of Oxford found out that more than half of respondents considered that Big Data is datasets that bigger than one terabyte (Schroeck et al., 2012). Another survey made by Intel IT center defined that storing more than ten terabyte of dataset as Big Data (Intel IT center, 2012). Besides, the volume of Big Data can also be measured by the number of records, transactions, tables, or files; an interviewee of TDWI survey indicated that their company never counted how many terabytes their data is, but they processed billions of data record (Russom, 2014). In addition to volume, the variety of data makes Big Data really big (Russom, 2014). The Method for an Integrated Knowledge Environment (MIKE2.0) project introduced a definition that Big Data relates to big complexity, which means that even. 立. 政治大. ‧ 國. 學. ‧. though the volume is not very large, data can be considered as Big Data as long as the data is complex and unstructured, such as videos (Rindler et al., 2013). The recent interest in data-driven insights has contributed to the popularity of Big Data (Madden, 2012). While it might be a marketing term, it also represents a new approach to exploring the world and making decisions (Lohr, 2012). The present is a new age of data-driven decisions and Big Data, with increasingly more decisions of many businesses relying on data analysis to improve efficiency and performance. n. er. io. sit. y. Nat. al. Ch. i n U. v. (Jagadish et al., 2014). Big Data opens great opportunities for real-time decisionmaking, operational and customer intelligence, business innovation, and gaining competitive advantages (Chen et al., 2015). However, it is essential to ensure that humans can fully comprehend the analysis results and avoid getting lost in the data sea (Jagadish et al., 2014).. engchi. 2. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(10) Figure 1-1 Properties of Big Data. 政治大 1.1.3 Big Data Application in Industries 立 Financial Industry. ‧ 國. 學. . ‧. In financial industry, by not only analyzing structured data from exchanges, banks, and data vendors, but also analyzing unstructured data from news, twitters, and social media, Big Data enables financial firms to discover innovative and strategic ways to manage portfolio, analyze risk exposure, perform enterprise-level analytics, and comply with regulations (Fang and Zhang, 2016). For example, Big Data technology enabled Bank of America to analyze all transactional data, customer data across multiple channels and relationships, and unstructured data at once, which successfully. n. er. io. sit. y. Nat. al. Ch. i n U. v. improve the understanding about customers (Davenport and Dyché, 2013). Give another example of a European company that provides credit-scoring solution for banks and consumer lending companies, the company used Big Data to analyze loan payment data and Facebook user data, and improved their scoring models, saved money on credit losses, expanded potential client base, and even increased revenue (Fang and Zhang, 2016). According to the study of Davenport and Dyché (2013), banks such as Wells. engchi. Fargo, Bank of America, and Discover used Big Data to analyze billions records of unstructured and semi-structured data including data from financial industry and other industries, such as website clicks, transaction records, bankers’ notes, and voice recordings from call centers. Big Data analytics help them improve their customer relationship, enhance the quality of customer interaction, and segment their customer better (Davenport and Dyché, 2013).. 3. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(11) . Healthcare Industry In Healthcare industry, the amount of available health data is increasing because of medical applications, remote monitoring technologies, and biological sensors; and these data usually need to be processed in real-time to not delay treating (Ong et al., 2016). By analyzing various data such as health record data, sensors' data, patient arrangement data, visualizing record data, Big Data analytics is helpful for disease interpreting, diseases avoiding, and diseases curing in healthcare industry (Das et al., 2018). For example, Big Data analytics enables United Healthcare to analyze unstructured data such as the recorded voice files from customer calls to call center, and understanding its customer satisfaction better (Davenport and Dyché, 2013). In another example, the study of Chen et al. (2017) analyzed more than 20 million records of inpatient department data. The inpatient department data includes structured data such as laboratory data and patient’s basic information, and it includes unstructured data such as patient’s narration of illness, doctor’s interrogation record and diagnosis. For complex diseases, analyzing both structured data and unstructured data improved the feature descriptions of diseases and the accuracy rate of disease risk prediction, which is helpful for disease preventing and early treatment (Chen et al., 2017).. 立. ‧ 國. 學. . 政治大. ‧. Retail Industry In retail industry, Big Data may help companies to evaluate customers’ behavior, segment and target customers more accurately, and forecast demand trends of products (Srivastava, 2018). For example, Walmart and Kohl’s improved product selection and price markdown times by analyzing sales, pricing, economic, demographic, and weather data (Lohr, 2012). Another example is that Big Data enabled Sears to process. n. er. io. sit. y. Nat. al. Ch. i n U. v. petabytes of customer, product, sales, and campaign data in real-time; and resulted in understanding increase marketing return, increasing customer return rate, and decreasing the release time of a set of complex marketing campaigns (Davenport and Dyché, 2013). Moreover, in the case study of Evans and Kitchin (2018), a large retail store operating in Ireland analyzed large amount of data in its ERP system, SCM system, CRM system, stock control system, online order fulﬁlment system, and so on. Big Data. engchi. is found out to be helpful for decision- making, increasing efﬁciencies, improving customer service, reducing risk and costs, and managing the work of staff (Evans and Kitchin, 2018). . Manufacturing Industry In manufacturing industry, innovative applications, such as active preventive. maintenance, production line optimization, and energy consumption optimization, can be motivated by analyzing device data, product data, and command data (Wan et al., 4. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(12) 2017). For example, a sensor of GE can generate 500 gigabytes of data per day, and analyzing the data from more than ten thousands of sensors helped GE optimize inspection, maintenance and repair processes of machines, which improved the efficiency and increased the revenue (Davenport and Dyché, 2013). Another example is that Intel used Big Data to decrease its quality checking time of chips by analyzing their historical data from pre-release chips, and successfully launching its chips in the market faster and decreasing its quality ensuring cost (“How big data analytics yields big gains,” 2017). As a final illustration, according to the study of Zhang et al. (2017), by analyzing real-time data such as flow rate, pressure, temperature, vibration signal, as well as non-real-time data such as maintenance history and failure record, Big Data enabled a cleaner production manufacturing company to provide better products, customized products, better service, and more innovative service. With these improvements, the company successfully decreased cost, improved manufacturing efficiency, increased revenue, increased customer satisfaction, increased customer loyalty, identified more potential customers, innovated its business model, and improved strategic cooperation with other companies (Zhang et al., 2017).. 立. ‧ 國. 學. . 政治大. ‧. Service Industry In service industry, Caesars Entertainment implemented Big Data tools to improve marketing and service in real time by analyzing data about its customers from its Total Rewards loyalty program, web clickstreams, and from real-time play in slot machines (Davenport and Dyché, 2013). Furthermore, Starwood Hotels and Resorts used Big Data to improve its pricing strategy and demand forecasting and it successfully increased their revenue by analyzing their structured data (such as bookings data and. n. er. io. sit. y. Nat. al. Ch. i n U. v. cancellations data) and external semi-structured data (such as weather reports) (“How big data analytics yields big gains,” 2017).. engchi. 1.2 Research Motivation and Research Purpose Big Data is causing revolutionary changes in many industries (Shim et al., 2015). The potential to improve decisions and generate insights attracts many organizations interested in creating value with Big Data (Sivarajah et al., 2017). However, a CIO survey in 2013 reported that more than half of Big Data projects failed (Chen et al., 2015). The return on investment in Big Data is not guaranteed. Big Data is an emerging trend that may greatly benefit businesses (Katal et al., 2013). The predictive power of Big Data is beyond doubt, as demonstrated by many wellknown successes, such as Walmart’s use of Big Data to make better decisions on pricing and product selection, Google measuring user behaviors more precisely (Lohr, 2012), and Amazon and Netflix successfully predicting future customer preferences (Bughin, 5. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(13) 2016; Ding et al., 2014; Ren et al., 2017). However, the return on investment in Big Data remains a big question for many firms (Shim et al., 2015). This study intends to assist financial companies interested in investing in Big Data technologies by providing insights into value creation via such investments. The main objectives of this study is: 1. To determine whether Big Data benefits financial companies in real-world cases 2. Identify the primary factors affecting the outcome.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 6. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(14) Chapter 2: Literature Review 2.1 Measuring Possible Benefits of Big Data Business performance can be improved with data-driven decision-making and Big Data (Provost and Fawcett, 2013). A recent study observed that performance is positively related to the extent the firm is data-driven (Brynjolfsson et al., 2011). Generally, Big Data is regarded as a highly valuable knowledge resource for organizations and a source of new knowledge, valuable insights, and innovation (Kabir and Carayannis, 2013). The massive quantity of data makes it easier to enhance the understanding of customer needs, improve service quality, and predict and prevent risks (Cai and Zhu, 2015). For example, in addition to millions records of customers data, packages data, and tracking requests data, UPS analyzed online map data and sensor data on vehicles, including speed , direction, braking, and drive train performance, to monitor daily performance and optimize route structures, and successfully decrease its fuel cost (Davenport and Dyché, 2013). Another example is that Netflix used Big Data to improve its content recommendation engine by analyzing its show features data, social data of users, box-office data, and successfully increased its movie and TV series. 立. 政治大. ‧ 國. 學. ‧. consumptions (Bughin, 2016). Besides, according to a study by Chen et al. (2015), each of ten big data outsourcing projects of Softserve Inc. significantly contributed to solving business challenges, achieved business goals, and increased company revenues. However, it is not easy to create business value with Big Data, despite a companies’ beliefs (Chen et al., 2015). Most companies fail to break even on Big Data projects (Shim et al., 2015). Success is challenging, as it is easy to misinterpret data (McAfee et al., 2012). Moreover, the large scale of datasets and the requirements for complex. n. er. io. sit. y. Nat. al. Ch. i n U. v. analysis techniques raise the risk of "false discoveries" with Big Data (Lohr, 2012). Unless data are appropriately introduced into a complex decision-making process, breaking even on investments in Big Data is not guaranteed, i.e., such investments may be worthless (Shah et al., 2012).. engchi. 2.1.1 Three Categories of IT-enabled Business Values As presented in table 2-1, the prior study of Mirani and Lederer (1998) and Gregor et al. (2006) proposed the three categories of business values generated from information technologies, which includes informational benefits, transactional benefits, and strategic benefits. Although scholars have proposed many conceptualizations of business values generated from information technology, the prior research of Ren et al. (2017) founds out that the three dimensions of business value, which were proposed by Mirani and Lederer (1998) and Gregor et al. (2006), significantly enhanced firms’ performance in the Big Data analysis environment. Thus, referring to these prior studies 7. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(15) by Mirani and Lederer (1998), Gregor et al. (2006), and Ren et al. (2017), the business value created from Big Data can be categorized as informational benefits, transactional benefits, and strategic benefits. Informational value relates to effectiveness and efficiency of decision-making and information access, such as real time decision making. Transactional value relates to efficiency improvements and cost savings. Strategic value relates to increasing competitive advantages (Gregor et al., 2006; Mirani and Lederer, 1998; Ren et al., 2017). Table 2-1 Three Categories of IT-enabled Business Values (Gregor et al., 2006; Mirani and Lederer, 1998; Ren et al., 2017) Category. Possible benefits. Informational Enabling a faster or easier access to information for decisionmaking. 政治大 Providing立 information in more useable formats. Improving information quality or accuracy for decision-making Reducing costs. 學. ‧ 國. Transactional. Increasing financial returns Enhancing productivity or efficiency. Strategic. ‧. Growing the business. Improving partnerships or relationships with other companies. sit. y. Nat. Enabling a faster response to changes. Improving customer relations and segmentation. er. io. Providing better or innovative products, services, or business. al. n. models. i n C Aligning analyticshwith business strategy U i e h n c g Creating competitive advantages. v. 2.1.2 Possible Benefits of Big Data Many related literatures mentioned many kinds of business value that can be generated by Big Data, and most of them belong to the three categories of IT-enabled Business Values. For example, Big Data may create informational values, e.g., leading to higher information quality, reducing the needed time for data gathering, and decreasing the data processing time (Wamba et al., 2015). With Big Data, Walmart and Kohl’s improved its decision quality on pricing and product selection (Lohr, 2012), Caesars Entertainment decreased its data processing time to make real-time decisions on marketing (Davenport and Dyché, 2013), and Netflix also improved its content recommendation engine (Bughin, 2016). 8. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(16) Moreover, Big Data may create transactional value, e.g., by improving the speed of business execution, growing the business, and decreasing costs (Jablonski, 2014). With Big Data, UPS reduced its fuel cost, GE improved its efficiency and increased its revenue (Davenport and Dyché, 2013), Netflix grew its business (Bughin, 2016), and Intel launched its products in the market faster and decreased its cost (“How big data analytics yields big gains,” 2017). Big Data may also create strategic value by identifying customer segments for customized promotions, or innovating new business models, products and services (Wamba et al., 2015). With Big Data, Google understood users better (Lohr, 2012), Amazon and Netflix understood their customer better (Bughin, 2016; Ding et al., 2014; Ren et al., 2017), and Wells Fargo, Bank of America, Discover Financial Services, and United Healthcare improved their customer relationship (Davenport and Dyché, 2013). Moreover, a cleaner production manufacturing company provided better and customized products and service, and successfully increased customer relationship, innovated its business model, and improved strategic cooperation with other companies (Zhang et al., 2017). A summary of Big Data’s possible benefits is presented in Table 2-2. We can measure. 立. 政治大. ‧ 國. 學. ‧. the return on investment in Big Data by ascertaining whether companies create any of the identified business values by implementing Big Data technologies.. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 9. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(17) Table 2-2 Possible Benefits of Big Data (Gregor et al., 2006; Mirani and Lederer, 1998; Ren et al., 2017) Category. Possible benefits. References. Informational Enabling a faster or easier access Lohr, 2012; Davenport and to information for decisionDyché, 2013; Wamba et al., making. 2015; Bughin, 2016. Improving information quality or accuracy for decision-making Providing information in more useable formats Transactional. Davenport and Dyché, 2013;. Increasing financial returns Enhancing productivity or efficiency. Jablonski, 2014; Bughin, 2016 ; “How big data analytics yields big gains,” 2017. Improving partnerships or relationships with other. Lohr, 2012; Davenport and Dyché, 2013; Wamba et al.,. companies. 2015; Zhang et al., 2017. 政治大 Growing the 立business. 學. ‧ 國. Strategic. Reducing costs. ‧. Enabling a faster response to changes. n. al. products, services, or business models. Ch. engchi. Aligning analytics with business strategy. sit er. io. Providing better or innovative. y. Nat. Improving customer relations and segmentation. i n U. v. Creating competitive advantages. 2.2 Factors Affecting the Business Value of Big Data Although Big Data applications may appear to present an attractive business opportunity, numerous challenges remain (Shim et al., 2015). Reviewing prior studies related to factors influencing the value-creation process of Big Data and data-driven decision-making, this study summarizes the following five primary factor types that are strongly associated with the business value of Big Data. 2.2.1 Accessibility, Timeliness, and Quality of Data Accessibility refers to the degree of data openness, with high accessibility meaning 10. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(18) that a user can obtain many different types of data (Cai and Zhu, 2015). Accessibility of multiple data types enables complex analyses (Ikemoto and Marsh, 2007). However, acquiring data from diverse sources constitutes a significant challenge (Sivarajah et al., 2017). A lack of data accessibility poses an important obstacle to data-driven decisionmaking (Marsh et al., 2006). Indeed, numerous companies have experienced challenges while gathering all available data (Bughin et al., 2010). According to a survey conducted by MIT Sloan Management Review and the IBM Institute for Business Value, more than 20% of respondents cited data accessibility as the primary obstacle to becoming more data-driven (LaValle et al., 2011). The study of Wamba and colleagues (2015) reported that 34% of reviewed articles identified accessibility as the primary challenge of the Big Data-driven value-creation process. Timeliness relates to the velocity of Big Data. To make both timely and correct decisions, it is also necessary to access data accurately and completely (Katal et al., 2013). The "shelf life" of data is very short due to the rapid data changes; without collecting the right data on time and processing it in real time, companies might reach outdated or invalid conclusions (Cai and Zhu, 2015). Moreover, the bigger the data volume is, the more important timely data collection and real-time processing will. 立. 政治大. ‧ 國. 學. ‧. become (Jagadish et al., 2014). Several studies identified data quality as an important factor affecting data use (Ikemoto and Marsh, 2007; Marsh et al., 2006). Recent research determined that data quality was associated with project results (Katal et al., 2013). Another recent study also observed data quality to be the primary challenge in several Big Data projects of Softserve Inc, a global software outsourcing company (Chen et al., 2015). Poor data quality leads to creating less business value and making wrong decisions in Big Data. n. er. io. sit. y. Nat. al. Ch. i n U. v. projects (Janssen et al., 2017). Therefore, the high quality of data is the foundation of motivating data-driven actions (Black et al., 2005). As poor data quality may lead to serious errors in decision-making, while ensuring Big Data quality remains challenging, the use and analysis of Big Data must be based on accurate and high-quality data to create value (Cai and Zhu, 2015). Besides, data quality issues are obstacles for realtime and actionable use of Big Data (Grover et al., 2018).. engchi. 2.2.2 Data Policy Because of the new collecting sources and new processing ways of data, data security concerns and data privacy concerns are increasing, and they need to be emphasized and solved (Bertino and Ferrari, 2018). A survey conducted by MIT Sloan Management Review and the IBM Institute for Business Value reported that more than 20% of respondents identified data policy as the primary challenge to becoming more datadriven (LaValle et al., 2011). Research also determined that data policies were an 11. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(19) important aspect of business value enabled by Big Data (Wamba et al., 2015). For instance, many studies indicated that privacy of personal data was the primary topic of concern with Big Data applications (Emani et al., 2015; Jagadish et al., 2014; Shim et al., 2015). A recent study also regarded privacy and security as the most important issues of Big Data (Katal et al., 2013). Due to the growing security threats, Big Data must be protected to keep from negative impacts (Grover et al., 2018). The ethical concerns arise regardless of whether data are used for marketing (Shim et al., 2015). The larger amount of data that companies can collect from customers, the more individual privacy concerns will arise (Aloysius et al., 2016). That is, the security and privacy challenges result from the large scale should be solved in order to achieve the expected benefits of Big Data (Li and Gao, 2016). Addressing the privacy concerns in the digital era represents the primary challenge of Big Data (Sivarajah et al., 2017), and it is suggested to combine different techniques to address the security concerns (Xu and Shi, 2016).. 立. 政治大. ‧ 國. 學. 2.2.3 Staff Capacity and External Support All forms of data require skills to translate it into information and actionable. ‧. knowledge (Ikemoto and Marsh, 2007). However, for some organizations, the biggest challenge in implementing Big Data is the human capital issues, in details, the lack of Big Data experts may lead to failure on Big Data investment (Grover et al., 2018). In the survey conducted by MIT Sloan Management Review and the IBM Institute for Business Value, approximately 30% of respondents encountered a shortage of skills within their organizations (LaValle et al., 2011). A recent review observed many studies identifying the role of talent in the Big Data-driven value-creation process (Wamba et. n. er. io. sit. y. Nat. al. Ch. i n U. v. al., 2015). The demand for Big Data skills continues to rise due to the increasing speed, variety, and volume of information (Shim et al., 2015). However, many companies still lack the talent needed to create business value with Big Data (Bughin et al., 2010). For instance, care providers in the healthcare industry typically do not have the ability to perform complex analyses on Big Data (Neff, 2013). The data lifecycle of Big Data includes planning, data collection, data preprocessing,. engchi. analysis, interpretation, execution, and so forth (Blazquez and Domenech, 2017; Jagadish et al., 2014). Technical skills are very important in the data lifecycle of Big Data, especially in data collecting, data preprocessing, and analysis phase. However, while dealing with Big Data, not only technical but also research, analytic, interpretive and creative skills are important (Katal et al., 2013). While numerous individuals possess the mathematical knowledge underlying the analytical models, few have the skills to apply such models correctly to Big Data (Madden, 2012). In conclusion, analytical and information technology skills are important to turn raw data into 12. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(20) meaningful and actionable knowledge, but in order to provide useful insight and support decision-making, the ability of understanding business problems, developing solutions, and communicating are also necessary in Big Data process (Chen et al., 2012). Qualified staff are the key success factor of Big Data, as numerous varied skills and extensive knowledge are required in Big Data projects, necessitating recruitment of talented employees, training current workers, or organizing appropriate teams according to business goals (Provost and Fawcett, 2013). For instance, GE recruited approximately 400 talented data scientists and developed a special training program for them (Davenport and Dyché, 2013). A study of Wamba et al. (2017) found out that expertise capability and management capability are important pillars of Big Data, organizations must improve their staff's technical knowledge, technological management knowledge, business knowledge, and other relational knowledge of Big Data (Wamba et al., 2017). Finding enough skilled professionals to help companies solve problems by using Big Data is important yet very difficult (McAfee et al., 2012), and it is necessary for firms to keep training their staff on analytics techniques and communication ability (Fang and Zhang, 2016). Planning skills, technical skills, managing skills, interpreting skill, and communicating skills are enablers to transform. 立. 政治大. ‧ 國. 學. ‧. data into valuable actions. The research of Janssen et al. (2017) indicated the importance of staff having appropriate skills and capabilities and highlighted the difficulty of identifying the suitable individuals with sufficient knowledge and communication skills to work with Big Data and interpret results. The difficulty arises from Big Data involving analysis of a large variety of parameters and variables, with knowing how to identify and apply the correct technique being a significant challenge. Overcoming this challenge may. n. er. io. sit. y. Nat. al. Ch. i n U. v. necessitate skills training of internal staff, hiring external personnel, and potentially collaborating with external organizations (Janssen et al., 2017). It is unsurprising that outsourcing is the primary choice of many enterprises, as it helps them deploy Big Data systems quickly and return to focusing primarily on core competencies (Chen et al., 2015). Numerous large companies outsource a portion of Big Data processing to external organizations (Jagadish et al., 2014). External support. engchi. is undoubtedly helpful to organizations transforming raw data into actionable knowledge (Ikemoto and Marsh, 2007). 2.2.4 Tools and Technology Tools, especially complex ones, are an important factor in data-driven decisionmaking (Ikemoto and Marsh, 2007). Although the potential value of Big Data is wellrecognized, numerous technical challenges remain, e.g., the task of loading a large quantity of data (Jagadish et al., 2014). It is difficult to reduce the time required to store 13. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(21) and process a large amount of data (Katal et al., 2013). The large size and related storage aspects of Big Data, especially involving unstructured data that is very difficult for traditional tools to work with, are common challenges for many companies (Tankard, 2012). Most traditional analytic tools, e.g., R, SAS and MATLAB, are not designed to process a quantity of data exceeding a single computer's amount of RAM (Madden, 2012). Developing Big Data systems is very different from developing those oriented towards traditional structured data (Chen et al., 2015). Such unstructured data, as text, images, videos, and sensor data streams, are difficult to work with using traditional databases (Lohr, 2012). Furthermore, methods of querying and mining Big Data differ greatly from those of traditional data analysis, as the data are noisier and more dynamic, heterogeneous and correlated, while at the same time being less trustworthy (Jagadish et al., 2014). Identifying the appropriate tools and techniques for analysis and visualization of Big Data is challenging due to its complexity (Janssen et al., 2017). A recent study observed that technology was the most important challenge to Big Data-enabled business value (Wamba et al., 2015). However, many companies do not have the key technologies to create information and value from data (Bughin et al., 2010). In particular, many. 立. 政治大. ‧ 國. 學. ‧. companies encountered technology-related difficulties; e.g., Walmart experienced difficulties transmitting a million transactions per hour to a database (Swan, 2013). Technology selection affects many aspects of system performance, consistency, availability, latency, scalability and modifiability (Chen et al., 2015). Big Data involves numerous innovative technologies due to its complexity (Provost and Fawcett, 2013). It is important to use scalable technologies to store data, such as Hadoop, NoSQL, HBase, MongoDB, Cassandra, Microsoft Azure, or Amazon Web Service ( Mikalef et. n. er. io. sit. y. Nat. al. Ch. i n U. v. al., 2017). For example, non-traditional data, such as large-scale log data, sensor data, or social media data, needs new technology such as Hadoop to handle, because nontraditional data is hard to handle by traditional data warehouse (Katal et al., 2013). Many organizations, such as NASA and Quantcast, succeeded in using Big Data following installations of new technology and tools (Simon, 2013). Moreover, United Healthcare was able to analyze voice data after adapting a variety of tools and. engchi. techniques, such as “natural language processing” software, Hadoop and NoSQL storage (Davenport and Dyché, 2013). New technology and methods need to be adopted to create value from Big Data due to its scale (Katal et al., 2013), and the growth of the data scale appears to outpace advances in computer technologies (Jagadish et al., 2014; Sivarajah et al., 2017). Therefore, technology will remain a key component of Big Data (McAfee et al., 2012).. 14. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(22) 2.2.5 Organizational Culture and Leadership Sometimes organizational barriers, such as lacking a culture of collaboration, may pose a challenge to companies trying to benefit from Big Data (Beath et al., 2012). A trusting and data-driven culture enables complex data-driven decision-making (Ikemoto and Marsh, 2007). The head of analytics at Macys.com, a well-known online retailer, mentioned that the company’s ROI-driven culture is the primary reason for failing to prioritize the Big Data technologies (Davenport and Dyché, 2013). The survey conducted by MIT Sloan Management Review and the IBM Institute for Business Value reported that more than 20% of respondents cited the organizational culture as an obstacle to becoming more data-driven (LaValle et al., 2011). It is important for organizations to have a data-driven culture, meaning that all employees are aware of data’s importance to decision-making and understand that data are not merely the responsibility of management or the IT department, but rather everyone’s business (Provost and Fawcett, 2013). That is, organizations should avoid intuition-driven decisions and avoid pretending to be more data-driven than is the case (McAfee et al., 2012). In addition to the organizational culture, leadership might be a key factor affecting. 立. 政治大. ‧ 國. 學. ‧. the value-creation process with Big Data. According to the survey conducted by MIT Sloan Management Review and the IBM Institute for Business Value, the primary challenge for leaders is not knowing how analytics can help business (LaValle et al., 2011). Leaders with data-driven visions increase the openness and collaboration within an organization (Ikemoto and Marsh, 2007; Marsh et al., 2006). To succeed in the era of Big Data, companies should be led by those who can set clear goals, define the path taken to reach them, and ask the right questions (McAfee et al., 2012). For instance,. n. er. io. sit. y. Nat. al. Ch. i n U. v. medical doctors often refuse to invest in Big Data over concerns with the expense of time, resources and personnel, with the perceived cost being much higher than the expected benefits, hence posing an obstacle to creating value from Big Data (Neff, 2013). In conclusion, the leaders must take the responsibility for the organization becoming aware of the importance and benefits of Big Data, motivate and provide clear goals to staff, monitor all processes and resources, and understand the data that can. engchi. provide the needed insights (Provost and Fawcett, 2013). In conclusion, company culture that is data-driven and evidence-driven, and leadership teams that can ask the right questions and have clear goals, are both helpful for generating value from Big Data (Grover et al., 2018).. 15. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(23) Table 2-3 Factors Affecting Big Data Factors. Description. References. Accessibility, Timeliness, and Quality of. Accessibility refers to the degree of data openness. Timeliness refers to. Black et al., 2005; Marsh et al., 2006; Ikemoto and Marsh, 2007; Bughin et al., 2010; LaValle et al., 2011; Katal et. Data. timely access to and processing of valid data. High data quality enables creation of business value and prevents incorrect decisions.. al., 2013; Jagadish et al., 2014; Cai and Zhu, 2015; Chen et al., 2015; Wamba et al., 2015; Janssen et al., 2017; Sivarajah et al., 2017; Grover et al., 2018. Data Policy. The privacy, security, and ethical concerns of data.. LaValle et al., 2011; Katal et al., 2013; Jagadish et al., 2014; Shim et al., 2015; Wamba et al., 2015; Aloysius et al., 2016; Li and Gao, 2016; Xu and Shi, 2016; Sivarajah et al., 2017; Bertino and Ferrari, 2018; Grover et Marsh et al., 2006; Ikemoto and Marsh, 2007; Bughin et al., 2010; LaValle et al., 2011; Madden, 2012; McAfee et al., 2012; Davenport and Dyché, 2013; Katal et al., 2013; Neff, 2013; Provost and Fawcett, 2013;. deploy Big Data systems quickly, to be able to restore the focus on core competencies.. Jagadish et al., 2014; Chen et al., 2015; Shim et al., 2015; Wamba et al., 2015; Fang and Zhang, 2016; Janssen et al., 2017; Grover et al., 2018. Using the appropriate tools and techniques to. Ikemoto and Marsh, 2007; Bughin et al., 2010; Lohr, 2012; Madden, 2012;. process data that cannot be readily analyzed by traditional tools and techniques.. McAfee et al., 2012; Tankard, 2012; Davenport and Dyché, 2013; Katal et al., 2013; Provost and Fawcett, 2013; Simon, 2013; Swan, 2013; Jagadish et al., 2014; Chen et al., 2015; Janssen et al., 2017; Mikilef et al., 2017;. ‧. Staff should have technical, research, analytical, interpretive and creative skills. External organizations can help enterprises. y. sit. n. al. er. io. Tools and Technology. 學. al., 2018. Nat. Staff Capacity and External Support. ‧ 國. 立. 政治大. Ch. engchi. i n U. v. Sivarajah et al., 2017; 16. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(24) Organizational Employees should know Culture and that data is important to Leadership decision-making and fully engaged in the project. Leaders should make staff aware of importance and benefits of Big Data, set clear goals, monitor processes and resources, and understand the data that can provide the. Marsh et al., 2006; Ikemoto and Marsh, 2007; LaValle et al., 2011; Beath et al., 2012; McAfee et al., 2012; Davenport and Dyché, 2013; Neff, 2013; Provost and Fawcett, 2013; Grover et al., 2018. needed insights.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 17. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(25) Chapter 3: Research Methodology 3.1 Research Design and Process The research process of this study is shown in Table 3-1. The research objective is to determine whether Big Data provides sizable benefits to companies and identify the key factors affecting the Big Data-driven value-creation process. A qualitative case study is a useful tool for analyzing complex phenomena; to understand and compare similarities and differences of various cases, multiple case studies will be conducted (Baxter and Jack, 2008). This study will choose four enterprises that invested in Big Data as study targets, and use a semi-structured questionnaire (Appendix A) which is designed based on the reviewed literature to conduct an in-depth interview. However, prior to interviewing the selected firms, a contextual analysis was completed. To accomplish the first objective, this study summarized the possible benefits of Big Data according to multiple related stories that can help measure the business value created by Big Data in practical cases. To accomplish the second objective, this study summarized five primary key factors, enabling beneficial applications of Big Data according to multiple related studies.. 立. 政治大. ‧ 國. 學. ‧. Subsequent to the interviews, this study will try to validate the benefits Big Data has created for the respective companies and will attempt to analyze each key factor hypothesis, proposed based on the literature review, in four practical cases. Finally, this study will systematically compare the findings across the four cases and draw a conclusion based on the research results of the multi-case study and the literature review. The conclusion of this study is expected to be a guideline that can provide insights for Big Data implementations.. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 18. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(26) Table 3-1 Description of the research process Activities. Multi-case Study. Results. To understand how to measure the benefits of Big Data and the. Review the existing related research to summarize the. The possible return on investment in Big. possible key factors that enable Big Data to create business value.. possible benefits created by Big Data and identify several primary key factors.. Data and hypotheses for key factors.. To determine whether Big Data are beneficial. This study will conduct interviews. Raw findings of a multi-case study.. in practical cases and validate our hypotheses regarding the on key. with selected firms and perform an indepth analysis of their. factors.. Big Data projects.. 政治大. 立. To summarize the findings.. Analyze, compare, and summarize the. 學. Conclusion. How. ‧ 國. Literature Review. Why. case study findings.. The finalized conclusion and. ‧. further observations.. y. Nat. sit. 3.2 Data Collection. n. al. er. io. 3.2.1 Select Criteria Due to the exploding volume, variety, and velocity of data in financial industry, Big. i n U. v. Data becomes an important role to improve financial companies’ competitiveness (Prabhakar and Maves, 2017). Although some financial service firms already used Big Data to support decisions (Barr et al., 2018), they are still interesting that what kind of benefit does Big Data bring to financial companies? What factors affects the result? In. Ch. engchi. conclusion, financial industry is a suitable research target for this research. For case selection, this study selected four companies from financial industry in Taiwan to interview. First, the selected study targets are all representative companies in financial industry with high competitiveness. Second, they must have implemented Big Data for at least two years because according to our preliminary survey, the impacts of Big Data may be insignificant if the implementing time is too short. Third, for confirming the difference between the traditional data analysis and the Big Data analysis, the data of selected cases must be difficult to be handled by traditional tools and technique. That is, referring to the selection criteria of a prior survey conducted by Intel IT center (Intel IT center, 2012) and a prior survey conducted by TDWI (Russom, 19. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(27) 2014), the data scale of our selected cases must be larger than ten terabytes or more than one billion of records. Alternatively, the data source of selected cases must include semi-structured data or unstructured data that is hard to be handled by traditional tools and technique. Last, for data collection, the senior IT managers of the cases must be willing to attend our interview and share their experience. In summary, the selection of the cases are based on the following criteria: 1. Representative firms with high competitiveness in financial industry. 2. At least two years of Big Data implementation. 3. The volume of data is bigger than ten terabytes or more than one billion of records, or the data source includes complex unstructured data. 4. Senior IT managers are willing to share information and experience about Big Data implementation.. 立. 政治大. ‧ 國. 學. 3.2.2 Case Description This study collected data by interviewing experienced IT managers who has involved in their companies’ Big Data implementation of the selected cases. Each manager is. sit. y. Nat. n. al. er. Company A Company A is a financial holding company with some key subsidiaries. The major. io. . ‧. interviewed and discussed the process of implementing Big Data for their firm, the impact after implementing Big Data, and the critical factors of implementing Big Data. The details of four studied companies are as follows, and the description of selected cases is summarized in Table 3-2.. Ch. i n U. v. subsidiary of the company is the bank, which is one of the highest growth banks in Taiwan with more than 100 domestic branches, and more than 20 overseas branches in 8 countries. Besides, the bank is currently the top three credit-card issuing bank in Taiwan. The contributed capital of company A is more than 3 billion US dollars, with total assets of more than 60 billion US dollars. The company is a pioneer of financial companies in Taiwan that strives to use. engchi. technologies, including Big Data technology, to launch many financial innovations. Unfortunately, the top IT managers of the company were too busy to attend our interview, so we interviewed a senior IT manager who is familiar with the Big Data application of the company. . Company B. Company B is a financial holding company with key subsidiaries of a capital group company, a securities company, a bank, and so forth. Many subsidiaries controlled by 20. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(28) the company leads the market in Taiwan, for example, the capital group company has more than 30% market share in Taiwan, and the market share of the security company’s underwriting business is about the top place in Taiwan’s market. The contributed capital of company B is more than 5 billion US dollars with more than 100 home branches, and more than 30 oversea branches in 4 countries. For years, the company dedicates in financial technologies, and keeps trying to strengthen its competitive advantage by financial technologies. Fortunately, the Chief Technology Officer, a senior manager from Fintech and New Financial Service Department, and a senior IT manager from Operation and Technology Department were willing to attend our interview. . Company C Company C is one of the largest insurance company in Taiwan with more than 400 home branches, more than 60 overseas branches, and more than 20% market share in Taiwan. The company has been in operation for over 50 years, and has more than 8 million customers now. The contributed capital of the company is about 2 billion US dollars, and the company has won many awards for many years.. 立. 政治大. ‧ 國. 學. ‧. For years, the company has invested many resources to develop new technology applications in the insurance field. Besides, the company also dedicated to training technology talents, including Big Data talents, in order to improve their services by technologies. Fortunately, the Chief Information Officer was willing to attend our interview and share his experiences.. sit. n. al. er. io. Company D. y. Nat. . Ch. i n U. v. Company D is a leading financial holding company with many subsidiaries such as bank, insurance company, and so forth. The company integrated the resources of its subsidiaries and improved overall operating efficiency. The bank is one of the largest bank in Taiwan and it is also the top 50 banks in Asia, and it is currently the top three credit-card issuing bank in Taiwan. The contributed capital of the company is more than 6 billion US dollars, and it has more than 150 home branches and more than 100. engchi. overseas branches in 14 countries. Moreover, the company has been in operation for over 50 years and has accumulated more than 7 million customers over the years. The company keeps innovating and improving its services for decades, and it is a pioneer of financial technology in Taiwan. Fortunately, the Chief Technology Officer and a senior IT manager from Information Service Division were willing to attend our interview and share their opinions.. 21. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(29) Table 3-2 Description of selected cases Main Business Service Contributed Capital Number of Branches. Company A. Company B. Company C. Company D. Financial holding company. Financial holding company. Insurance company. Financial holding company. More than 3 billion US dollars. More than 5 billion US dollars. About 2 billion US dollars. More than 6 billion US dollars. More than 150. More than 150. More than 450. More than 250. Chief. io. n. Time. al. 60 minutes. 150 minutes. Ch. engchi. Officer. y. Information Service Division. sit. Nat Interview. Department, Senior IT Manager / Operation and Technology Department. Chief Technology Officer, Senior IT Manager /. er. ‧ 國. CTO Office. Chief Information. ‧. of Interviewee. 立. Senior IT Manager /. 政治大. 學. Job Title of Interviewee / Department. Technology Officer, Senior IT Manager / Fintech and New Financial Service. i n U. v. 90 minutes. 120 minutes. 22. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(30) Chapter 4: Research Results 4.1 Case Study 4.1.1 Case Study of Company A More than 10 years ago, when the term of Big Data was not popular at all, the company began to invest in Big Data analysis due to the rapidly expanding data volume and the more complex data types. The company analyzed both structured data and unstructured data. They analyzed structured data such as credit card transaction data, counter transaction data, customer data, and so on. They also analyzed semistructured data and unstructured data such as online news, social media data, etc. Furthermore, the compressed data in their data warehouse is bigger than 16TB, which does not include their log data. In the past, the company used package software to analyze data because of the user-friendly user interface, but now, because of the rapid increase in data volume and data complexity, it is difficult to handle Big Data by traditional tools. Therefore, the company gradually switched their tools, such as some open source software, to support Big Data analysis. For instance, the company used Hadoop for data storage, Python, SAS for analysis, Tableau for visualization.. 立. 政治大. ‧ 國. 學. sit. y. Nat. n. al. er. Digital Intensive Intelligent System (DIIS) For marketing, company A integrated artificial intelligence technologies to develop. io. . ‧. The company has used Big Data in the field of customer relationship management (CRM), decision supporting, resource allocating, business process improving, and precision marketing, and so on. Moreover, they are also trying to apply Big Data in more fields.. Ch. i n U. v. a digital intensive intelligent system that analyzed 460 million records of data from 6000 web pages. The system helped the company to understand customers’ intends and customers’ demands, and let the company can conduct real-time recommendation. For instance, if a customer search for exchange rate, the system will automatically recommend related service such as currency exchange service, travel insurance, and credit cards. As a result, the system increased the open rate of e-mail direct marketing. engchi. (EDM) for three times, increased the open rate of short message service (SMS) for six times, increased the loan volume for six times, and save 75% cost. . Automatic Teller Machine (ATM) Set up Decision-making For decision supporting, the company applied Big Data for deciding where to set up automatic teller machine (ATM). They collected many kinds of data such as the automatic teller machine (ATM) location of competitors and the data of the flow of crowd. Then, they visualized all related information on a dynamic map platform to 23. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(31) support decision, which clearly provided all useful information such as information of competitors, information of flow of crowd, and cash flow information. . Real-time Online Loan Platform For business process improving, the online loan platform of company A applied Big. Data technology to provide suitable loan quotes intermediately. In the past, customers should wait for many days to get their loan quotes, but now they only have to wait for three minutes for getting loan quotes. Besides, the improvement of business process also saved a lots of labor costs. 4.1.2 Case Study of Company B Company B was experienced in data analysis for more than 10 years, and applied Big Data analysis for about two years. They analyzed internal structured data such as customer profile data, transaction data, and product data. They also analyzed external data such as social media data, open-data, web crawler data, and data from other industries’ business partners, which includes structured data, semi-structured data, and unstructured data. In the past, the company used to purchase solutions such as. 立. 政治大. ‧ 國. 學. ‧. package software from external companies, but sometimes the solutions did not really solve the problems of the company. However, by applying Big Data technology, the problems of company B were solved well than before. For Big Data implementation, the company applied many new technologies such as open source tools, cloud services, cloud storage, graphics processing unit (GPU), the cleaning tools of IBM, Google, and Microsoft, etc. In addition, the company also cooperated with external specialists, such as start-up teams and professional companies, to get technical. n. er. io. sit. y. Nat. al. Ch. i n U. v. supports and consultancy services. The company has used Big Data in the field of customer relationship management (CRM), decision supporting, pricing, risk controlling, precision marketing, financial management, and so on. . engchi. Understanding Market Changes and Marketing to Specific Segmentation. For example, Big Data helped the managers to decide which credit card to market, helped the company to understand market changes better, helped each subsidiaries to cross-sell, improved the loan approval rate, and helped the company marketed to specific segmentation by learning more about customers, such as marketing to young people. . Obtain Information More Efficiently In the past, the company spent few months to obtain some analysis results such as 24. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(32) customer sales list. Nevertheless, after applying Big Data technologies and machine learning technologies, the company can obtain the information they need more efficiently. . Combing AI technologies to Develop Innovative Service Model. By combing artificial intelligence (AI) technologies and Big Data technologies, the company were able to develop many innovative service model such as financial roboadvisor, intelligent customer service robot, intelligent human resource robot, and so on. Besides, it is worth mentioning that the company develops an innovative financial technology platform. The platform combines Big Data and other technologies to provide application programming interface (API) for other partners to use their financial service, and it becomes a leading financial technology platform in Taiwan.. 政治大. 4.1.3 Case Study of Company C Company C established their data warehouse in 2001 that is almost the largest and most complete one in insurance industry. The company has developed many applications on business intelligence, and started to analyze unstructured data for. 立. ‧ 國. 學. ‧. more than two years. The company has a big volume of data due to the large number of customers, and the active structured data, which does not include the unstructured data, is more than five terabytes. The company analyzed structured data such as transaction data, customer data, government open-data, and so forth. Recently, they also analyzed semi-structured data and unstructured data such as social media data, telephone-recording data, scan file, fax file, etc. Furthermore, the company continued researching how to apply technology to improve their Big Data implementation, they. n. er. io. sit. y. Nat. al. Ch. i n U. v. applied tools such as Python, R, Hadoop, Tableau, and so forth. The company has used Big Data in the field of customer relationship management (CRM), decision supporting, risk controlling, business process improving, and automated decision-making, and so on. . engchi. Customer Segmentation and Precision Marketing. For instance, the company has applied Big Data analysis for years to identify customers’ needs, improve customer segmentation, and conduct precision marketing. They segmented their customer into nine main groups according to customers’ purchasing power and customers’ career stage, and provided the most appropriate product or service at the right time and with the right channel in the right way. Rather than marketing as shooting in the dark, Big Data helped the company knowing their customers well, and conduct precision marketing according to customers’ needs. Through these improvements, the contact yield has increased by two times than 25. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(33) before. However, the chief information officer of company C noted that they thought that potential customers searching was still not good enough, and they still needed to improve the way they used Big Data to find new customers. . Improved Business Process by Integrating Machine Learning Technologies. Another example is that company C integrated Big Data and machine learning technologies to improve their business process, such as automated underwriting and automated insurance claim verifying. In fact, the underwriting time was shortened from three days to one day after applying automated underwriting. . Customer Risk Score and Investment. Moreover, Big Data had helped the company calculate their customer risk score, which strongly supported the automated decision-making of insurance claim verifying. However, the chief information officer of company C mentioned that they hoped the power of Big Data could improve the decision-making on investment more significantly in the future, because investment is a main revenue for insurance companies.. 立. 政治大. ‧ 國. 學. ‧. 4.1.4 Case Study of Company D Company D involved in data analysis for more than 20 years, and invested in Big Data for more than 3 years. The company analyzed structured data such as customer profile data, transaction data, government open-data, and so on. They also analyzed semi-structured and unstructured data such as customer’s web behavior, social media data, geographic information system data (GIS), scan files, fax files, telephone. n. er. io. sit. y. Nat. al. Ch. i n U. v. recording data, etc. The company has an enormous volume of data due to the huge customer base that is more than 7 million and the annual transaction that of more than 300 million records. The active data in their data warehouse is about 50 terabytes to 100 terabytes. However, the chief technology officer mentioned that the data volume is not the main concern, the quality and variety of data is more important. Because the company take Big Data as one of the most important competitive. engchi. weapon, they set up a data research and development center to integrate the resources of all subsidiaries, and study the integration of Big Data and artificial intelligence (AI). The chief technology officer of company D indicated that the improvement of technology is the key enabler of Big Data application. The company applied many tools and technologies in Big Data analysis process, such as Hadoop, SAS, Python, cloud storage, and so forth. The company also cooperated with many external partners such as IBM, SAS, Google, and McKinsey & Company, who provided assistance and consultancy service for the importing of tools, techniques, and processes. 26. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(34) For many years, the company is well known for applying Big Data technology to find customers’ needs. In fact, the company has not only used Big Data in the field of customer relationship management (CRM) and precision marketing, but also the fields of credit scoring, decision supporting, risk controlling, business process improving, and so on. . Customer Segmentation and Customer Labeling For marketing, Big Data improved the customer segmentation and customer labeling. Rather than marketing as shooting in the dark such as traditional telephone marketing, Big Data helped the company to decide that how to sell what product to which customer in what way in what time, which increased the efficiency and effectiveness of marketing and also improved the customers’ experience. The improvement significantly influenced the enormous marketing cost of millions of US dollars.. 立. 學. Obtaining Accurate Credit Score For risk controlling, the company obtained accurate credit score by applying Big. ‧ 國. . 政治大. sit. y. Nat. n. al. er. Business Process Optimizing For business process improving, Big Data helped the company optimize their. io. . ‧. Data technology to analyze many sources of data, including geographic information system (GIS) data. Thus, they could determine in a few minutes that whether a customer is suitable for microfinance or making a loan to. The improvement in risk controlling saved enormous cost that is much bigger than the cost saved in marketing.. Ch. i n U. v. business processes such as the processes of application accepting, information verification, application approving, and so on. These improvements successfully saved much time and costs.. engchi. 27. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.

(35) Table 4-1 Summary of Big Data implementation in selected cases Company B. Company C. Company D. More than 10 years. About 2 years. More than 2 years. More than 3 years.  Business partners’ data  Customer  Customer data data  Social media  Open-data data  Transaction data  Web news. 立. 政治大. ‧ 國. 學. data. io.  Business process. n. al. Ch. improving  CRM Field of Big Data  DecisionApplication support.  Marketing  Resource allocating. Technology and Tools. Hadoop, Python, SAS, D3.js, Tableau. Inconvenient to disclose. More than 5 TB. y. Nat. More than 16 TB. ‧. Data Volume.  Product data  Scan file  Social media  Telephonedata recording  Transaction data data  Transaction  Web crawler data.  Automated decisionmaking. sit. Data Scope.  Customer data  Fax file  Government open-data.  CRM  Decisionsupport  Financial management  Marketing. engchi  Pricing.  Risk control. Open source tools, cloud services, cloud storage, GPU, cleaning tools. er. Years of Big Data Implementation. Company A. v i n U Business. process improving  CRM  Decisionsupport  Marketing  Risk control. SAS, Python, Hadoop, R, Tableau.  Customer data  Fax file  GIS data  Government open-data  Scan file  Social media data  Telephonerecording data  Transaction data More than 50 TB  Business process improving  Credit scoring  CRM  Decisionsupport  Marketing  Risk control. Hadoop, SAS, python, cloud storage, GPU, R. 28. DOI:10.6814/THE.NCCU.MIS.006.2018.A05.