Course: Business Analytics Week 1 “Introduction to Business Analytics” - Learning Packet 1 Learning Packet Introduction: In this week we’ll try to understand the basic concepts in business analytics. In doing so we’ll clearly understand the difference between terms such as analytics, data-science, business analytics and business intelligence. We will also understand different types of analytics, i.e., descriptive, prescriptive and causal. This will lay the groundwork for upcoming topics in this course. Further we delve deeper into the steps of the business analytics process. The students will also understand how to draw connections between a typical organizational decision-making process and business analytics process. Lastly, in this learning packet we explain different degrees of coordination between the strategies in an organization and the role of business analytics. Module1 : Business analytics and the relationship with intelligence 1 Chapter: Introduction to Business Analytics: Some Basic Concepts 1.1 Business Analytics Several statistical techniques can be applied to discover, visualize and identify trends or patterns in data. Analytics involves the use of statistical techniques such as measures of central tendency, graphical representations, information system software, and operations research methodologies (for example, inventory models, linear programming). In simple words, analytics is used to obtain useful information from data. Analytics applies to all disciplines and not only businesses. A typical example of the modern-day use of analytics is the assessing the data related to the demographics, views and beliefs of voters. For example, a communication strategy for the political party will be developed by using the insights from the analysis of this data. Business analytics can be defined as the process of applying quantitative methods in order to derive meaning from data and making informed business decisions. Thus, business analytics This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 goes one step further by utilizing analytics to generate a measurable improvement in “business performance.” 1.2 Relationship with data science The main focus of analytics lies in understanding and analyzing datasets to obtain answers to specific issues. Business analysts participate in tasks such as budgeting, forecasting, and product development. They work closely with users and other stakeholders. Data science encompasses developing, cleaning, and systematizing datasets. Data scientists collect raw data and convert it into something understandable by leveraging various tools such as algorithms, statistical models and their own analyses. The focus data scientists lie in data cleaning, combining complex datasets, programming, and statistical modeling. Data Science focuses on uncovering new and unique questions that can help the businesses. In doing so concepts such as data mining, predictive modeling, and machine learning algorithms are used to extract patterns from complex datasets. Business analytics, on the other hand, aims at finding solutions to important questions and support data-driven decision making. In simple terms business analytics falls under the broad umbrella of data science and focuses on finding answers to specific questions to help businesses. 1.3 Business Intelligence It can be often noticed that business intelligence and business analytics are both used interchangeably by professionals. It is a matter of debate whether business analytics is a subset of business intelligence, or is it other way round or if there is an overlap between two concepts. Many experts consider business intelligence as umbrella term and business analytics falls under its realm, to describe a set of predictive tools. For ease of understanding, we define business intelligence as a set of processes and tools that transform data into meaningful and valuable information to help the businesses. It uses past and present data to make ongoing decisions. On the other hand, business analytics can predict the future by utilizing historical data. For example, while business intelligence might help the managers to understand what is current demand for their product, business analytics provides insights about the future demand of the product. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Table below provides an overview of the characteristics of analytics, business analytics and business intelligence Characterstics of analytics, business analytics and business intelligence. Source: Schniederjans et al. 1.4 Statistical analysis Statistical analysis is at the heart of business analytics. Several statistical techniques and tools are applied to obtain useful insights from data. Statistical analysis can be broadly categorized as descriptive statistics and inferential statistics. In descriptive statistics, we summarize data using charts and graphs. A sizeable number of data points are reduced to certain meaningful summary values and graphs using descriptive analytics. Variance, range, mean, median etc. are measures of descriptive statistics. Inferential Statistics focuses on making predictions about a population based on a representative sample. In the framework of business analytics, statistical analysis involves a lot of inferential statistics. Several inferential statistics tools such as hypothesis testing, analysis of variance, regression analysis can be used in business analytics. 2 Chapter: Descriptive, Prescriptive and Predictive Analytics This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 2.1 Descriptive analytics Descriptive Analytics helps us to answer: “What has happened?” Descriptive analytics does exactly what the name implies: they “describe”, or summarize, raw data and make it something that is interpretable. When we say that descriptive analytics describes the past, the ‘past’ can be any point of time when an event has occurred. Use descriptive analytics when you need to understand at an aggregate level what is going on in your company, and when you want to summarize and describe different aspects of your business. Descriptive analytics lets us learn from previous behaviors, so that you can comprehend their impact on present and future outcomes. The commonly used statistical procedures are fall under this category. For example, basic arithmetic like count, averages, percentages are used as descriptive analytics. Descriptive analytics can be applied to get insights about production, financials, processes, sales, finance, humanresource, inventory and customers. Some of the examples for the application of descriptive statistics could be total sales in five years, demand trends, average inventory, absenteeism, overall equipment efficiency over a period of time, and change in profits. 2.2 Predictive Analytics: Predictive analytics is used to answer: “What could happen?” The ability to “predict” what may happen in future is provided by predictive analytics. Predictive analytics is used to obtain actionable insights with the help of data. The concept of probability is the foundation of predictive analytics as no tool can predict the future with complete certainty. In companies, historical data can be obtained from systems such as enterprise resource planning (ERP) and Point of Sale (POS) to identify underlying patterns. Predictive analytics models help us to identify relationships between variables. Such relationships help the companies to look into the future. Predictive analytics has a wide variety of applications throughout various aspects of the businesses. For example, it can be used to forecast customer behavior, identify trends in absenteeism of the employees, demand forecasting, understand staffing needs etc. In finance sector, credit scores are used to determine the probability that the customer will make future credit payments on time. This is also an example of predictive analytics. Some of the most widely used predictive models are: This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 – Regression: Regression analysis predicts the relationships among variables. Key patterns in huge and diverse data sets can be found out by understanding how variables are related with each other. – Decision trees: Decision trees are a simple tool of dealing with multiple variables. Decision trees are created by algorithms that split data into branch-like segments and help to understand path of decisions. – Clustering model: A clustering model creates categories in data on the basis of on similar characteristics. Market segmentation can be done with the help of cluster analysis. In the upcoming weeks we will study linear regression, logistic regression and clustering models in detail. 2.2.1 Predictive Modeling Predictive modeling encompasses development of models that are useful for forecasting or predicting future events. In business analytics, these models are developed with the help of logic or data. As a part of our course we will be focusing on data-driven models. Logic driven models A logic-driven model is one based on experience, knowledge, and logical relationships of variables and constants connected to the desired business performance outcome situation. The question here is how to put variables and constants together to create a model that can predict the future. Model building requires an understanding of business systems and the relationships of variables and constants that seek to generate a desirable business performance outcome. A fishbone diagram could be an example of logic-driven model. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Source: Emiliani, 1998 Data driven models Data-driven models are created by using data collected from many sources to quantitatively establish the relationships. Logic driven models are generally used as precursor to data driven models. Some of the popular data driven models are regression analysis, correlation analysis, simulations and so on. In the following figure you can see an example of regression model. Source: Escobar 2.3 Prescriptive Analytics Analytics, which use optimization and simulation algorithms to advise on possible outcomes and helps to answer the questions: “Why it will happen” and “What should we do?” Prescriptive analytics is relatively new as compared to descriptive and predictive analytics. It allows us to “prescribe” a various actions and lead to a solution. In most simple terms, prescriptive analytics is useful for providing advice. It quantifies the effect of set of actions in to get a picture of possible outcomes, before making decision. While predictive analytics predicts what will happen, prescriptive analytics goes a step further and tells us “why it will happen.” Thus, recommendations regarding actions can be made. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 After carrying out descriptive and predictive analytics, one can undertake prescriptive analytics. Prescriptive analytics mostly utilizes decision science, management science, and operations research methodologies and focused on making best use of resources. The following figure provides an approximate depiction of prescriptive analytics methodologies and an overview of relationship between descriptive, predictive and prescriptive analytics. Prescriptive analytics methodologies. Source: Schniederjans et al. 3 Chapter: Steps of Business Analytics 3.1 Overview of the business analysis process It is important to understand the steps of the business analysis process so that you can systematically undertake business analytics projects in your organization. Business analytics can be utilized to solve issues as well as identify opportunities for improvement. Typically, problemsolving and identifying opportunities are both decision-making tasks. Identifying opportunities, can also be viewed as a “problem of choosing right strategy.” As shown in the figure below, the steps of a typical business analytic process can be mapped with the steps in typical organization decision making processes. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Comparison of business analytics and organization decision-making process. Source: Schniederjans et al. 3.2 Six step research approach for business analytics As a professional, it is not possible to work in silos and detach oneself from the wider aspects of organizational decision-making process. Therefore, it is practically advisable that while undertaking a business analytics process one should follow the six-steps business research approach which is shown in the figure below: This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Six step research approach. Source: Zikmund et al. Now we will see each of the steps in detail. Stage 1 Clarifying the research question This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Clarifying the research question. Source: Zikmund et al. This process is exploratory in nature. In this stage we can also run descriptive analytics to understand what is happening. The process begins with the management dilemma—the problem or opportunity that requires a business decision. The example of management dilemma could be: There is a decline in plant productivity. The next step is Management question. Management question is the management dilemma restated in question format. The example of management question is: How do we increase plant productivity, to bring it back to its former level? The management question leads to Research Questions. Research question best states the objective of the research. The question(s) will be addressed by the researcher. For example: What are the factors impacting productivity and how important is each factor? Stage 2 Research Proposal This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Research proposal. Source: Zikmund et al. Once the research question is defined, research is proposed in order to allocate resources to the project. Figure gives an overview of research proposal. Stage 3 Research design strategy Research design. Source: Zikmund et al. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 The research design is the outline for fulfilling objectives and providing the insight to answer management’s dilemma. The main sub-topics that we will be discussing as are census, sample, sampling design process and sampling techniques. Census Census is a count of all elements in a population. Sample A sample is a group of cases, participants, events, or records constituting a portion of the target population, carefully selected to represent that population. Sampling design process Under sampling design, we identify the target population (those people, events, or records that have the desired information and can answer the measurement questions) and then determine whether a sample or a census is desired. The sampling design process includes six steps which are shown in the figure. Sampling design process. Source: Nunan et al. Further, we move towards the sampling techniques as shown in the figure. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Sampling techniques. Source: Nunan et al. The scope, time frame, context of the study, the nature of data (qualitative or quantitative) is decided in research design stage. Here we also decide the type of research. For example, is it exploratory, descriptive, experimental, causal? The data collection instrument (For example, survey questionnaire, focused group interview questions, etc.) is designed in this stage. The instrument should also be checked for validity and reliability. Reliability refers to the consistency of a measure. If a measure is reliable, the results can be reproduced using the measure again and again under same conditions. Validity refers to the accuracy of a measure. A pilot test is conducted to identify weaknesses in research methodology and the data collection instrument. A pilot test may have from 25 to 100 subjects. Stage 4 Data collection and preparation The gathering of data includes a variety of data gathering alternatives. Questionnaires, standardized tests, and checklists are among the instruments used to record raw data. Secondary data are data generally collected to address a problem other than the one which requires the manager’s attention at the moment. Collected from published data. Primary data are data the researcher collects to address the specific problem at hand—the research question. It is collected by the researcher through questionnaires (observation method). Data are cleaned to ensure consistency across respondents and to locate omissions. Stage 5: Data analysis and interpretation This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 Researchers generate information and insights by analyzing data post its collection. Data analysis is the editing, data-reduction, summarizing, identifying patterns, and applying statistical methods to data. In this course we will be focusing mainly on this stage of the research process. Stage 6 Reporting the results. In this stage the analyst puts forth the findings, insights, and recommendations to the manager for the intended purpose of decision making. 4 Chapter: Linking business analytics and organizational strategy 4.1 Business Analytics at the strategic level In this chapter, a number of scenarios are presented that depict different degrees of coordination between the strategies in an organization and the role of business analytics (see figure). We begin by defining strategy. A strategy is a description of the overall way in which a business currently is, and is to be, run. The purpose of strategy is to adapt the organization’s business area, resources, and activities to the market in which the organization operates. As a rule of thumb, a strategy attempts to handle issues faced by the organization in the short run while at the same time trying to create competitive advantages in the long run. Link between strategy and BA. Source: Gert et al. Scenario 1 is that there is no formal link between the use of business analytics and business strategy. Firms that lie here are the ones without data or with limited data distributed over a large number of sources. These firms are generally not able to make a link between corporate strategy and business analytics. Data is not used for decision making at a strategic level in these companies. Instead, data is used to answer concrete questions along the way and This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 automate processes. But there is no link to business strategy. Generally, ad-hoc retrieval of data is utilized in such companies to answer specific questions. Scenario 2 is that business analytics supports strategy at a functional level. For example, in a firm, business analytics function performs monitoring of individual functions’ achievement of targets, we have coordination between strategy and business analytics. The analytics function is reactive in relation to the strategy function. In this case, the role of business analytics is merely to produce reports supporting the performance of individual departments (marketing, operations, etc.). Scenario 3 is dialogue between the strategy and the business analytics functions. If the organization makes sure that individual functions optimize its way of working based on business analytics information, but that the strategy function, too, takes part in the learning loop, we’ll get a business analytics function that proactively supports the strategy function. A learning loop is facilitated when the business analytics function is reporting on business targets and is providing analyses of as well as identifying differences between targets and actuals, with the objective of improving both future strategies and the individual departments’ performance. Scenario 4 is information as a strategic resource. The characteristic of the fourth scenario is that information is treated as a strategic resource that can be used to determine strategy. Companies that fit this scenario will systematically, while analyzing the opportunities and threats of the market, consider how information, in combination with their strategies, can give them a competitive advantage. Conclusion We’ve build a foundation for our course by understanding the basics of Business Analytics. We started by understanding the relationships and differences of business analytics with concepts such as data science, business intelligence and statistical analysis. We delved into types of analytics ― descriptive, prescriptive and prescriptive and developed a general understanding on how these can be applied to real business problems for better business decisions. The steps of business analytics have also been discussed in detail to empower the users to systemize their efforts. Further, the linkage of business analytics with organizational strategy has been clarified. In the next week we start with hands-on application of analytics tools with Stata along with detailed understanding of important tools. This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023 References: o Gavin, M. (2019). Business Analytics: What it is & why it’s important| HBS Online. Business Insights-Blog.https://online.hbs.edu/blog/post/importance-of-business-analytics o Stobierski, T. (2021) what's the difference between data analytics & data science? | HBS Online. Business Insights-Blog. https://online.hbs.edu/blog/post/data-analytics-vs-datascience o Chatterjee, D. (2022) Business Intelligence vs. Business Analytics: How to Distinguish Easily. Emeritus Blog. https://emeritus.org/blog/career-path-business-intelligence-vsbusiness-analytics/ o Schniederjans, M. J., Schniederjans, D. G., & Starkey, C. M. (2014). Business analytics principles, concepts, and applications: what, why, and how. Pearson Education. o Nunan, D., Malhotra, N. K., & Birks, D. F. (2020). Marketing research: Applied insight. Pearson UK. o Laursen, Gert Hn, Thorlund, Jesper (2016). Business analytics for managers: Taking business intelligence beyond reporting. John Wiley & Sons. o Emiliani, M. L. (1998). Continuous personal improvement. Journal of workplace learning, 10(1), 29-38. o Escobar (2016). Why do I need to have knowledge of multiple regression to understand sem? The analysis factor. https://www.theanalysisfactor.com/why-need-knowledge-ofmultiple-regression-to-understand-sem/ This document is prepared for use only in Prof. Chetna Chauhan’s Business Analytics course at Universidad de los Andes- Colombia (UniAndes) from March 2023 to May 2023