Business Analytics - Term Paper Example

Executive summary This paper presents the basics of Business analytics (BA), also known as the business intelligence (BI). The paper presents a basic understanding of the BA/BI through two cases where BI/BA applies. The cases that demonstrate the understanding of BI include: case 1: understanding BI with a keen attention to the insurance industry. The second case is simply a demonstration of how a BA/BI is formulated to the stage of information collection using a case of students applying for universities in the USA. Conclusively speaking, this paper finds out that the insurance industry embraces business analytics. BI/BA plays a very crucial role in identifying and attempting to find solutions to the business problems that would otherwise undermine the operationalization of the business in the insurance industry. In insurance, BI/BA is used for customer relationship management, channel management, actuarial decisions, underwriting and policy management, claims management, and finance and asset management. BA is done in four phases (analytics ecosystem) which include the descriptive/diagnostic, predictive, prescriptive, and most recently, the exploratory. Business analytics is done using either of the two strategies; strategy-based/top-down and data-based/down-up strategy. Implementation of BA is done in five steps starting with understanding the business problem, understanding the required data, collecting data, analyzing data and finally reporting on the findings. The second case is simply divided into three parts;, the first part seeks to explore the details of what is being researched about and the second part details the formulation of the study by designing the questions that will be used to capture the desired information. Alongside formulating the study, the sampling design is discussed just to form an extension of the part as a third section. Contents Executive summary 1 Introduction 1 Customer relationship management (CRM) 1 Channel management 2 Actuarial 2 Underwriting and policy management 2 Claims management 2 Finance and asset management 3 Analytics ecosystem 3 Descriptive analytics 4 Predictive analytics 5 Prescriptive analytics 6 Exploratory analytics 7 Implementation of analytics and challenges of achieving/cultivating analytic leadership and culture in practice 8 Understanding business problem 9 Determining the metrics and analyses techniques appropriate for the study/understanding data 9 Collecting the data 9 Analyze the Data 9 Reporting on the findings 10 Challenges in implementing Business analytics 10 Changes of human behavior 10 Data mining process 10 Job-related human factors 11 Conclusion 11 Case 2: Observational study for student segments 12 Student segment expectations 12 Explorers 12 Highfliers 12 Strivers 13 Strugglers 13 Observational study design 13 Appropriate sampling method 16 Appendix 17 References 18 Introduction Business analytics(BA), also called business intelligence (BI), can be described as the technology-driven process for analyzing data and presenting actionable information that can be used by the end users in making more informed business decisions (Laursen and Thorlund, 2010). Cloud Data Integration Software | Matillion, (2017) web page says that insurance has been slow in embracing BI, but upon realizing the importance of BI, they have greatly embraced it in the recent past. Like any other business, BI in the insurance industry is very helpful creating the strategic value and competitive advantage of any player. As indicated earlier therefore, the main purpose of BI in the insurance industry is to present actionable information based used by different end users in making more informed business decisions. As far as strategic value and competitive advantage is concerned, the end users that find BI very important include the insurance underwriters, the claim handlers and the sales managers. The most important role that BI is embraced for is its role in customer relationship management. Mainly, the insights of the BI are used in addressing the ever-changing needs of customers. For the underwriters, sales managers and claim handlers, the BI plays an important role in strategic value and competitive advantage in the insurance industry as described below by Boobier, (2016) in supplementation of the lecture notes chapter 3 and other sources; Customer relationship management (CRM) CRM is important for strategic value and competitive advantage of an industry player. In CRM, BI plays an important role in increasing customer profitability, modeling customer lifetime value, customer segmentation, attrition analysis (very important since losing a customer in insurance is very expensive), affinity analysis, target marketing, campaign analysis, and cross selling analysis. These roles are important for the sales manager in determining the effective marketing decisions as well handling the existing customers. Channel management Until recently, insurers had not embraced the various effective distribution channels. One of such channels is the online channels. They mainly relied on methods such as agent marketers. BI has played a very important role in channel management including: optimal deployment of agents and salesforce at the right locations, agent development and relationship management, and E business development. The latter is particularly one of the greatest achievements that BI has played role in the insurance industry due to the ever-increasing market opportunities in the online. E business development has helped businesses in expanding client network as well as improving business communication. Actuarial BI is very helpful to an actuary for calculating future premiums and to allocate portions of a book of business for reinsurance. Underwriting and policy management BI helps in improving the underwriting and policy management. BI helps in tracking of premium performance by product or product line; by geographical region, agency/agencies, and by branch or office. BI also enables an underwriter to track underwriting losses so as to enable the underwriters salvage their books of business. Claims management BI plays an important role in claims analysis. One example would be behavioral analysis in the healthcare insurance. Through Improved claim management, BI also helps in fraud detection as well as claims estimation. Finance and asset management In order to compete effectively, insurers need to increase their return on investments and reduce their underwriting costs. This particularly helps in budget analysis, asset liability management, profitability analysis, and many other financial ratio analyses. In summary, BI is useful in the insurance industry for strategic value and competitive advantage if the insights are applied by the respective users well. The users include the underwriters, sales managers and the claim managers. BI enables insurance underwriters in underwriting policy management, setting up premiums and handling loss ratios effectively. The sales managers use BI in managing the customer relations well as well as working out the marketing channels effectively. It, for example, helps in managing churning. For claim handlers, BI insights help optimize the claim process, minimize losses and increase customer satisfaction by speeding up the entire process. With all these, the insurance industry benefits from BI in respects that impart strategic value and competitive advantage for the players that adopt the BI. It enables the players cope with the dynamics in the industry including, but not limited to changes in government policy and changes in behavioral trends of the clients. The decision process in the analytics is arrived at through some process. This therefore means that the BI/BA involves a process. Business analytics takes place in phases towards achieving the best outcome in decision-making process. This process comprises what is referred to as analytics ecosystem. The basics of the analytics ecosystem and how they are adopted in the insurance industry are illustrated below. Analytics ecosystem The analytics ecosystem can be defined as the phases that comprise the analytics process up to making sensible business decisions with respect to the needed outcomes of the process. There are four phases in the analytics ecosystem. The first phase is the descriptive, also known as the diagnostic, then the predictive, the prescriptive, and, most recently, the exploratory phase. However, as will be seen later in this chapter, the phases are only for clarity. The analytical process does not always follow the given phase design/order in practice. The description of these phases and a basic illustration of how they are adopted in the insurance industry is as shown below by guide of the lecture notes and Tony Boobier (2016) among other sources. Descriptive analytics This phase consists of organizing, summarizing, and presenting data in an informative manner, essentially using graphs and/or numbers. It also comprises what is referred to as the inferential statistics, which essentially draws conclusions about populations based on the samples. With this understanding, It can be concluded that the diagnostic/descriptive analytics/statistics seek to give an understanding of “what happened?” and “why?” as far as conclusions made from the informative presentation of the data is involved. In insurance, descriptive/diagnostic analytics seek to answer the question “why is it happening?” and so, it strives to identify the causes, unseen patterns and the key factors behind a given situation. The analytics examine the a given company’s data and past performance to find reasons behind the past success and/or failure. In insurance, the descriptive analytics reviews the standard descriptive reports that illustrate premium growth, loss ratios and hit ratios as well (Boobier, 2016). From it, trends and quantification of how many, how often and the time certain findings occurred can be identified. In some cases, descriptive analytics have been programmed to make alerts when certain trends emerge. One example is when the hit ratio begins to deteriorate, an alert is issued. This informs the underwriters that they should be able to investigate their pricing competitiveness. Similarly, through descriptive analytics and proper programming of an alert, such emerging loss patterns as sewer related problems can be used to trigger tightened underwriting rules. In this context of descriptive analytics, there’s the proactive logic that brings to the forefront the potential issues and thus alleviating the risk that they may remain buried within standard reports until the problem worsens. According to Liebowitz, (2013), insurance carriers start their analytics journey with the diagnostic/descriptive analytics. As an illustrative example of how some of the insurance companies use descriptive analytics, this paper presents the case of Progressive’s Snapshot tool which is used by the insurance company to help it’s client in understanding the determination of the car insurance pricing based on the driving of the customers. In this case, the application compiles the driving behavior and categorizes it and then eventually computes the insurance pricing. Among the driving behavior is the braking, sudden changes in speed and many other aspects that categorize risky driving. Of course, the end result of the analytics is to check the drivers’ behavior and eventually reduce the claim rates. Back to the basics of descriptive analytics, the analytics enable the clients understand what might have happened to warrant the pricing, similarly answering the question of why the pricing is the way it shows. Having looked at what happened, it’s always good to have a look at what is likely to happen (Vertafore.com, 2017; Boobier). This leads to the next subtopic; predictive analytics. Predictive analytics Predictive analytics essentially seek to address what is likely to occur given the current data. Benefits of predictive analytics include, but not limited to identifying trends, understanding customers, and driving strategic decision making as well as predicting behavior. Predictive analytics consist of the likely outcomes and the respective probabilities. This leads to the “what if ” scenarios and risk assessment. As Boobier (2016) describes, it mainly uses statistical and quantitative analysis in identifying meaningful relationships in datasets. It involves use of both the internal and external datasets. It involves modeling of correlations which are then used for prognosticating the future from the current conditions of the modeled correlation. In the simplest of terms, predictive analytics involve using a model to predict an unknown future outcome. In the insurance industry, for example, predictive analytics has been in use for quite some time. Predicting a policyholder mortality based on age is a common practice implicit of the predictive analytics. There are many areas in the insurance where predictive analytics are of much use. By studying the relationship between the weather and loss patterns helps set pricing/marketing/assessing risk in the agricultural sector. Predictive analysis is widely applied in detecting fraud in the healthcare system. Apart from the forecasting the future, predictive analytics are also used to determine the current situation of a particular variable based on its relationship with other variables from a predefined model. One such example is the Aetna Company in which the company monitors claims so as to identify spikes that may indicate fraudulent behavior. It aims to predict the prevalence of individual fraudulent claims and stop payments from being issued. In so doing, the predictive analytics is used to identify, and therefore, maintain a better control of fraud (Boobier, 2016). Prescriptive analytics Prescriptive analytics mainly addresses the issue of what actions to be taken. It lies beyond the descriptive and predictive analytics in that it anticipates what will happen, why it will happen, and when it will happen. It goes further to suggest the decision options for potential. It goes ahead to show potential conclusions/results for each of the decision options. By this basic description, it means that prescriptive analytics involve the descriptive and the predictive analytics only that after the predictive phase, conclusive solutions are made about the outcomes. Of course it can be a bit confusing when it comes to understanding the difference between where descriptive analytics leaves off and prescriptive analytics begin. The basic distinction is that prescriptive analytics translates the forecast, probably by the predictive statistics, into a more feasible plan for going ahead with company growth. Essentially, the prescriptive analytics uses such tactics as the A/B testing in order to gain clear insight into one that works best under certain circumstances and scenario planning. Thus leading to a better understanding of how changing individual aspects of a given product affects the perception of the individuals and use a particular offering. Consequently important, it can be used to identify the changes that are infeasible and could result in bad performance. One example of prescriptive analytics in the insurance industry is the property and casualty insurance providers use catastrophe modeling pricing to recommend ways of setting up prices and establishing policy conditions as well as optimizing portfolios to keep accumulation in check. One other case is when providers choose to increase their services in relation to the such threats as cyber-attacks and natural disasters based on climate change patters. In this view, the prescriptive analytics are used to shape the future of the insurance industry (Saporito, 2014). Recently, there has a change in technology and a lot of data of all dimensions come are available for the insurance industry. This is referred to as the big data and requires a lot of exploration to make the data useful. This leads us to the most recent analytics phase, exploratory. Exploratory analytics Exploratory analytics is a more recent way of analytics. It aims to address the big data and social media challenges. Just have a clue of what big data, hence exploratory analytics is, it involves making useful of the information found on all the social networks (twitter, facebook, Instagram, whatsapp and many others ) and trying some good sense in redress of a particular issue. This is massive information coming in so fast (big data). It very complex yet very important since it captures the original data. It aims at exposing the “unknown unknowns”. In the case of insurance, the exploratory analytics tries to find the information from the public and then trying to analyze the unknown data dimensions. In insurance, unknown data from unknown people can be explored so as to understand the basic needs and likely trends of individuals. The insurers can then make decisions based on this explorations of the data.. Based on explored data, anticipations can be made about given trends and thus new products introduced in the insurance industry. One such example where the insurance industry can benefit is the driving behavior during festive seasons, or even the number of accidents as well as the nature of the accidents. From the public reactions, as read from the social media can be used to figure out the likely response to certain insurance products in this line (Saporito, 2014). Now that the analytics ecosystem is known, it’s always important to know how the business analytics is implemented. In the next section, an illustration of how the business analytics can be implemented. Alongside the implementation, challenges of achieving the analytic leadership and culture are discussed. Implementation of analytics and challenges of achieving/cultivating analytic leadership and culture in practice From Boobier(2016), lecture notes, and Saporito, (2014) there are two strategies that are used in implementing BA; These very material that have been referred to in this section. 1. Strategy driven analytics, also known as Top Down implementation 2. Data-driven strategy, also known as the bottom up strategy The order of strategy driven implementation BA is as follows; First of all, analytical model(s) are used to evaluate an existing strategy. After identifying the strategy, follows data exploration from which problem/decision analysis is done. Model development and implementation is then done and finally, the model is deployed to implement business strategy in an optimal way. The order of data-driven implementation of BA is as follows; It involves data analysis models to change or formulate new strategy, data is first explored, then analyzed and model developed and implemented. Threats and opportunities are then identified to guide business strategy and eventually leads to new/changed business strategy. Regardless of the strategy, there are 5 steps towards implementation of business analytics in the insurance industry as discussed below. Understanding business problem The first step is identifying the business problem. Companies defined by data driven strategies apply quantitative results to future business decision. BA are used for the monitoring, capturing and analyzing operations performance in reporting. Identifying the business problem is the first step of this process. Determining the metrics and analyses techniques appropriate for the study/understanding data The metrics should be aligned with the study. From the analysis, proper review of the statistical reporting methods to inform the study is done. Some of the statistical reporting that can be reviewed include the spss, market, SAS, Salesforce, and many more reporting tools. Collecting the data Data can be collected using multiple ways. They include the historical data to collecting data by way of interviews. Data collection involves internal enterprise systems as well as some of the metric tools for reporting such as ADP electronic invoicing. With the collected data, the next step is transform it to usable information. Most recently, data important to the insurance industry is availed by way of data mining from social media sources. Analyze the Data The insurance industry conducts data analysis using the analytic tools such as the SPSS, SAS or any as indicated in step 2. There are many aspects that a statistician would be looking for. They include co-efficient analysis, graphical presentation for pricing, risk deviations, regression frequencies and ay other particular presentation of the results that would give proper view of the identified business problem. Reporting on the findings This is the last step of implementing Business analytics. The analyzed data is then presented and reported in forms easy to understand such as graphical presentations or by comparative reporting. BA is usually effectively through a team. In the insurance industry, the leadership is likely to face challenges in the implementation of the Business analytics as described below. Challenges in implementing Business analytics Changes of human behavior The insurance industry is particularly faced with this problem. As a challenge, this comes with new information that needs to be analyzed and thereby posses challenge to the analytics leadership and the development of the analytics culture. Additionally, changes in human behavior may lead to the former information about the clients generally changed and most likely uncertain. This posses a great challenge to the analytics leadership and change of the analytics culture already established (Capgemini Capgemini Worldwide, 2017). Data mining process With big data and social network information, there is a lot of information available. However, not all the information is relevant to the insurance industry. This results to complex data mining processes that are a challenge to the leadership in the sense that they are expensive and generally not very easy to manage (Capgemini Capgemini Worldwide, 2017). Job-related human factors Agility Whereas analytics is a very serious endeavor, the analytics results are not easily communicated to the relevant business users that would find the worth value of the analytics results (Capgemini Capgemini Worldwide, 2017). Strategic alignment Although the insurance industry has some element of business analytics, it’s unfortunate that analytics are often viewed by top management as esoteric research at best and at worst as fringe experiments. This is implicit of alignment, availability and trust by the management (Capgemini Worldwide, 2017). As explained by Boobier (2016), the other factors that would pose challenge to analytics leadership in the insurance industry include domain/job protection, job politics as well as fears of undermining. Conclusion Conclusively speaking, BA is an essential part of business success. There are several ways of doing BA/BI in the insurance industry. The choice of how to conduct BA is dependent on the desired outcome of the analytics process. If applied, BA/BI is of great value to different players in the insurance industry. However, regardless of the importance it serves to the business, analytics leadership faces some challenges in dispensation and cultivation of the analytics culture. The problems range from technical to human related factors. When the problems are solved, the analytics leadership will have an easier time in promoting the analytics culture. Case 2: Observational study for student segments There are four student segments namely the explorers, highfliers, strivers, and strugglers (Choudaha, Chang and Schulmann, 2017). This categorization of student segments is based on the financial resources and academic preparedness of the students. With this categorization in mind, each segment of students has different expectations of quality of service as discussed below Student segment expectations Explorers This segment of students has high financial resources and comparatively low academic preparedness. This implies that the students will need more academic support. For international students, this would include language support. The students are more interested in experiential aspects of studying abroad and thus want to gain valuable life experience at their schools. They need support when during application. The institutions should therefore put in place a range of support services, including language, academic, and non-academic support services. The students have the financial means to pay for the services thus provision of the services is compensated for by the pay (Choudaha, Chang and Schulmann, 2017). Highfliers Highfliers have high financial resources as well high academic preparedness. This segment does not require a lot of institutional support. Family support is key to their financial independence and do not need institutional financial support. The students hold the institutions reputation at heart and seek prestigious institutions with better communication devices and high academic quality alike (Choudaha, Chang and Schulmann, 2017). Strivers This segment comprises of students that are academically prepared but financially pressed. Most of the students in this segment rely on institutional financial aid while others rely on loans thus financial aid information is a valuable service to then. This segment requires high academic quality. The students feel that they are academically prepared and therefore such extra services as English lessons would not be necessary. Strivers prefer to seek information from college platforms and would like interaction with admission officers (Choudaha, Chang and Schulmann, 2017). Strugglers This segment constitutes of low financial resources and low academic preparedness. Reputation of the school does not really matter thus have lower expectations for quality of education. This segment of students needs financial aid information. In order to accommodate these segment of students, the institution should ensure to have enough student support services, including, but not limited to, English language support structures and the availability of scholarships (Choudaha, Chang and Schulmann, 2017). Observational study design In order to carry out this study effectively and obtain high rate of response that would appropriately captured the desired information, multiple choice answers are found necessary. This, as the World education services(WES) report by Choudaha, Chang and Schulmann (2017) describes is referred to as the structured research/closed-ended questions. The most appropriate way of carrying out this study would be by targeting international and focusing the questions that would ensure answers are in line with desired information about the students. The students would then be categorized depending on their answers. These questions would be in such a way as to capture information including the financial capabilities of the students and the academic preparedness since this is the criterion of the segmentation. This will be inferred from the services the students are likely to need from their learning institution. The mode of data collection would be an online survey targeting over 1000 students visiting the US universities. This would help capture the services being sought after and thereby give more information about the students. The selection of the US Universities is tactical in the sense that the country is a preferred destination for international students (Choudaha, Chang and Schulmann, 2017). Since the students attend the universities as either undergraduate, masters or doctoral students, these categories will be considered as groups within the study sample. The country of the applicants is also good information to find about the international students. Given that the students can be sorted into the three levels, then a large number of respondents is desired so as to capture reliable information. To capture this, students in Universities across the USA and targets 1000 students randomly selected. Studies show that most international students come from India, China, Saudi Arabia and South Korea (Iie.org, 2017). Formulation of the survey would therefore follow the design as given below: To capture nationality, the question would be, “Please select one of the following concerning your where you come from”. The multiple choices would be a) India, b) China, c)Saudi Arabia, d) South Korea, e) Other Having known the country of origin, it is better to understand the academic level of the student. This would help in identifying which services would be most sought after by the particular academic level. The question to capture this information is “Choose your academic level of pursuit”. The answers include a) undergraduate, b) masters, c) Doctoral, d) other To capture financial resources, two possible questions are posed. The first one would be; “Do you earn monthly income? How much?” The possible answers would be income brackets starting with 0-500 dollars and then last at above 100, 000 dollars. However, only 4 options would be given so as to easily capture financial resources as either capable or otherwise. The second question would be to understand the possible sources of income under the question “please choose where most of your income comes from” with possible answers as a) parents, b) other family members, c) other In order to capture academic preparedness, questions that determine academic capability include those that capture language ability as well as other academic achievements. Given English as the language of communication, the first step towards preparedness would be to know if the student is cable to communicate well in English. The first question would therefore be “Which statement applies to your English language experience”. The possible answers would be a) I have completed training program, b) I have not completed but plan to undertake some training, c) Have not planned to attend any English training program. In order to know if the student is academically prepared by knowledge, a question to know how they performed to get to USA is good to ask. “Your last academic performance was very high and that’s why you can be accepted in USA”. The possible answers would be a) strongly agree, b) agree, c) neither agree nor disagree, d) disagree, e) strongly disagree. Having created the identity questions that can be used to categorize the students into the four student segments in the above section, then questions that would help identify their needs are formulated. The questions are as shown in the appendix section, which capture information including; 1. Action and feeling of preparedness to study in USA including how they rate their level of English versus the native USA students 2. Main sources of funding for the student’s education 3. Information sought when applying for college as well as the sources of information 4. willingness to pay for consultations The compilation of the questions is found in the appendix section. In order to achieve optimal results an appropriate as shown in the next section. Appropriate sampling method The most appropriate sampling technique would be simple random sampling. This helps eliminate a lot of errors especially the sampling bias. Given that it is an online survey, quite a number of applicants can be easily accessed thus the random sampling would very well give the population representation by the sample. Given that it would be very easy to quantify the number of total applicants, random sampling is the most appropriate method. This technique also beats all other methods since it gives the population representation of the composition of the students by academic levels of students as either undergraduate, masters, and doctoral. For example, if stratification is used, chances are that there would be biased in the samples due to biased sample sizes. The one advantage that this study enjoys from this survey design is that a large sample size can be used and information easily captured due to the online tool. Such sampling methods as stratified, cluster, systematic would be used but they can easily lead to sampling bias/error (Ardilly and Tillé, 2006). Appendix 1. Please select one of the following concerning where you come from South Korea China India Saudi Arabia Other 2. Choose the level you currently wish to apply for Bachelor’s program Master’s program Doctoral program 3. Which of the following captures your experience in English Already completed an English training program I haven’t attended any English language training program but plan to attend one No plans to attend any English language training program 4. Which statement applies to you regarding test preparation courses or tutoring I have taken course preparation courses or tutoring I have not taken any test preparation courses or tutoring, but I want to take one in the future I haven’t planned to take any test preparation courses or tutoring 5. Please specify how much you paid or would pay for English lessons Less than $200 $201-$500 $501-$1000 More than $1000 Won’t consider studying English 6. I feel academically prepared to enter USA universities Strongly agree Disagree Neither agree nor disagree Agree Strongly agree 7. I expect that my English will be as good as native American students Strongly agree Disagree Neither agree nor disagree Agree Strongly agree 8. What do you expect to be the main source of funding? Personal savings Support from family and/or friends Loan Financial support from the university through merit-based scholarships/teaching assistance Grants from external sources 9. Please rank up to 4 areas you seek information when applying for university in the USA Application requirements Tuition and cost of living Career products after school Reputation of the school Course offerings Program structure including credit transfer Faculty research and expertise Financial aid opportunities Cultural accommodations Student services (safety, student life) 10. Please select the sources of the information you found valuable about studying abroad Family and friends University websites Social media Web search Education fairs Educational consultants Online forums 11. If education consultant was involved, how much did you pay? Less than $500 $501-$2000 $2001-$5000 $5001-$10000 More than $10000 Didn’t pay Didn’t consult an education consultant 12. Which of the following sources of information do you trust the most? Family and friends Alumni from USA Social media Faculty from the USA Education consultant Current students in the USA Admissions office 13. Which social media would you like your university to post information? Please rank by order you prefer to have them Facebook Twitter Linked in Instagram Whatsapp Google plus Youtube pinterest References Ardilly, P. and Tillé, Y. (2006). Sampling methods. 1st ed. New York: Springer. Boobier, T. (2016). Analytics for insurance. 1st ed. Chichester, West Succex: John Wiley & Sons, Inc. Capgemini Capgemini Worldwide. (2017). Four key challenges for Business Analytics | Blog post. [online] Available at: https://www.capgemini.com/blog/capping-it-off/2012/04/four-key-challenges-for-business-analytics [Accessed 21 Jan. 2017]. Choudaha, R., Chang, L. and Schulmann, P. (2017). Student Segmentation for an Effective International Enrollment Strategy. [online] NY. Available at: http://knowledge.wes.org/rs/worldeducationservice/images/RAS-Paper-05-Student-Segmentation-for-an-Effective-International-Enrollment-Strategy.pdf [Accessed 21 Jan. 2017]. Cloud Data Integration Software | Matillion. (2017). How can Business Intelligence transform the insurance industry?. [online] Available at: https://www.matillion.com/insights/how-can-business-intelligence-transform-the-insurance-industry/ [Accessed 21 Jan. 2017]. Iie.org. (2017). International Students in the United States. [online] Available at: http://www.iie.org/Services/Project-Atlas/United-States/International-Students-In-US#.WIOQKRt97IU [Accessed 21 Jan. 2017]. Laursen, G. and Thorlund, J. (2010). Business analytics for managers. 1st ed. Hoboken, N.J.: Wiley, p.190. Liebowitz, J. (2013). Business analytics. 1st ed. florida: CRC press. Saporito, P. (2014). Applied insurance analytics. 1st ed. New jersy: FT press. Vertafore.com. (2017). 4 Types of Analytics Defining the Future of the Insurance Industry | Vertafore. [online] Available at: http://www.vertafore.com/Resources/Blog/4-Types-of-Analytics-Defining-the-Future-of-the-Insurance-Industry [Accessed 21 Jan. 2017]. Read More