INTRODUCTION
In the era of globalisation and aggressive competition, data has become an essential part of any organisation. With the rise of several business intelligence tools, business professionals can easily access data for making data-driven decisions. Data-driven organisations consider data an important asset as it helps them to make the right
decision at the right time. However, in order to stay competitive, it is not enough for organisations to merely make accurate decisions; they must also be able to anticipate future trends. In such cases, predictive insights allow organisations to mitigate risks, analyse market trends, and address growth challenges proactively.
Business analytics is the scientific process of transforming data into insights to make better decisions. It is used for data-driven or fact-based decision-making, which is often seen as more impartial than other alternatives. Business analytics utilises descriptive analytics, diagnostic analytics, predictive analytics, and prescriptive analytics for performing analysis on business data and support decision-making processes.
In this chapter, you will learn about business analytics, explore different types of business analytics and understand the scope and nature of business analytics along with the various stages of business analytics project.
Nature and Scope of Business Analytics
1.4 Nature and Scope of Business Analytics
A Business analytics is like a smart tool for businesses because it uses data (like numbers and information) to help companies understand things better and make smart choices. It looks at patterns in data, predicts what might happen next and uses techniques to get important information from all that data.
Imagine you have a big box of different candies, and you want to know which one everyone likes the most. Business analytics is like sorting through the candies, figuring out patterns and predicting which candy will be the favorite. It is not just about looking at the past (like saying, “Yesterday, everyone liked chocolate”). It is also about guessing what might happen in the future (like saying, “Maybe tomorrow, more people will like gummy bears”).
The nature and scope of business analytics covers a wide range of activities that use data to drive decision-making across diverse functional areas within an organisation. It is a dynamic field that continues to evolve with advancements in technology and the availability of increasingly large and complex datasets.
The following points show the nature of business analytics:
- Data-driven decision-making: Business analytics relies on data to support decision-making processes. It transforms raw data into actionable insights that facilitate strategic and operational decisions
- Statistical analysis and modelling: Business analytics employs statistical methods and models to analyse past data, identify trends and make predictions about future outcomes. This include regression analysis, clustering and machine learning algorithms.
- Technological integration: Business analytics leverages advanced technologies, including data visualisation tools, machine learning algorithms and Big Data technologies, to process and analyse large volumes of data efficiently.
- Cross-functional application: It is not limited to a specific business function. Business analytics can be applied across various departments, such as finance, marketing, operations and human resources, to enhance overall organisational performance.
- Continuous improvement: Business analytics is an iterative process. Organisations continually refine their analytical models and strategies based on new data and changing business conditions.
The scope of business analytics are as follows:
- Customer experience: Ensuring a positive customer experience is vital for efficient business operations. Business analytics enables companies to comprehend the types of customers frequenting their establishments, understand their purchasing habits and adapt services to satisfy and build loyalty. This customisation i possible through insights gained from business analytics.
- Inventory management: Stream lining supply chains an reducing costs by understanding order frequency, demand for specific products and optimising supply chain operations. Business analytics allows organisations to scale their offerings sustainably, making strategic planning more effective.
- Sales and marketing: Designing targeted campaigns an identifying cross-sell and up-sell opportunities by analysing customer responses to product offers and marketing initiatives. Analysing demographics, average income and purchasing reasons helps predict patterns in consumer behavior, allowing businesses to focus product messages and launch timing according to customer needs.
- Hiring and recruitment: Seeking collaboration with HR official skilled in business intelligence and data analytics. This ensures that HR specialists can use data to bring on board skilled an qualified workers, contributing to the company’s success. This approach also reduces hiring and training expenses.
- Finance: Leveraging big data and business analytics enables businesses to manage finances more skillfully. Insights into marketing spending and a comprehensive understanding of all financial transactions empower businesses to make informed decisions, effectively managing their resources. According to McKinsey & Company, a globally integrated analytics strategy for marketing spending could potentially unlock up to S $ 20 billion. As organisations start exploring business analytics, it is crucial to navigate the ethical considerations surrounding the collection, processing, and use of data. The ethical considerations in business analytics are as follows:
- Privacy concerns: Maintaining the privacy of individuals i paramount. Businesses must adhere to data protection regulations, obtain informed consent and implement robust privacy policies to ensure the responsible handling of personal information.
- Transparency: Building trust through transparency in data practices. Organisations should communicate openly about this data collection and usage policies, providing stakeholders with a clear understanding of how their data is being utilized.
- Bias and fairness: Acknowledging and addressing biases in data and algorithms is crucial. Organisations must actively work t identify and rectify biases to ensure fair and equitable outcomes in decision-making processes, avoiding unintended consequences.
- Security measures: Safeguarding data from unauthorized aces and cyber threats is non-negotiable. Implementing stringent security measures, including encryption and access controls, protects sensitive information and mitigates the risk of data breaches.
- Social impact: Considering the broader social impact of their organisation analytics initiatives. This involves assessing ho data-driven decisions may influence society and actively working to minimize any negative consequences. Responsible an ethical analytics practices contribute to the overall wellbeing of communities and societies at large. Next, we will learn about the stages of a business analytics project.
SUMMARY
- Business analytics, using diverse tools, relies on data for decisions, necessitating adaptation to technological advances and increasing data complexity in organisations.
- Business analytics offers benefits like informed decision-making and competitive advantage, but faces challenges such as data quality, privacy concerns and organisational resistance to change.
- Business analytics spans from describing historical data for informed decisions to diagnosing past events, predicting future trends and prescribing actions for optimised outcomes in strategic decision-making.
- Business analytics scope covers customer experience, inventory management, sales, marketing, hiring, recruitment and finance.
- Stages of a business analytics project involve identifying challenges, gathering data, analysing with tools, forecasting trends, brainstorming solutions and making decisions based on data insights.
- Identifying information gaps in business analytics requires recognising areas lacking essential data, examining existing sources to understand and fill those gaps meticulously.
KEY WORDS
- Automation: It employs technology for tasks without human intervention, hence streamlining processes.
- Big data: It refers to vast and complex datasets that exceed the capabilities of traditional data processing methods.
- Cloud computing: It is a technology that enables the storage, access and processing of data and applications over the Internet.
- Data sourcing: It refers to the process of collecting and acquiring data from different internal and external sources to support analysis.
- Metrics: They are quantifiable measures that are used to assess, track and evaluate performance in various aspects of business.
- Simple linear regression: It analyses the relationship between two variables, predicting the dependent variable based on the independent variable using a linear equation.
- Business analytics, using diverse tools, relies on data for decisions, necessitating adaptation to technological advances and increasing data complexity in organisations.
-
INTRODUCTION
In the previous chapter, you have learned about business analytics, types of business analytics, nature and scope of business analytics and the stages of a business analytics project.
Data preprocessing and visualisation are some of the most important steps in business analytics. Preprocessing involves cleaning, transforming, and organising raw data ensuring higher quality and consistency using techniques like imputation which replaces missing values with estimates and helps data for accurate insights. Data visualisation, on the other hand, is the graphical representation of data through charts, graphs and dashboards. It allows businesses to quickly identify trends, patterns, and anomalies, which aids in
informed decision-making. Together, preprocessing and visualisation help businesses derive actionable insights from vast amounts of data, improving efficiency and strategy development.In this chapter, you will be introduced to the significance of data in analytics. Then you will learn about types of data, common issues with data, and imputation methods. Further, you will learn about data visualisation by using Excel.
Using Data in Analytics
2.2 Using Data In Analytics:
Data is collected to gain insights into business problems. Both profit and non-profit organisations heavily depend on reliable data to achieve their organisational goals, such as strategic planning, performance assessment, improving operations, and benchmarking against industry standards or competitors.
Some examples of how data is used in business include:
- Annual reports: Data is systematically presented alongside visualisations such as charts and graphs to provide stakeholders with a clear representation of the company’s financial performance, including profitability, revenue and market share.
- Audits and financial oversight: Accountants audit financial records to ensure that balance sheet figures are precise and reflect the actual data, maintaining accuracy in reporting and regulatory compliance.
- Financial analysis: Financial analysts compile and analyse datasets to estimate a company’s performance based on key indicators such as profitability, revenue growth, Return on Investment (ROI), Economic Value Added (EVA) and shar value.
- Business analysis: Business analysts process data to develop clear visualisations to convey complicated data in a way that is easy to understand. Business analysts are required to develop and maintain dashboards using data for tracking key business metrics.
- Artificial Intelligence (AI): AI-engineers are required to anal and interpret large amounts of data to train I models and t improve their accuracy, efficiency and scalability. To use data for business, we need to first collect the data from various sources. Next, we will study different data collection sources and their considerations while collecting data.
Data Collection Process
Data collection involves obtaining and evaluating information from various sources to address research challenges, provide answers, assess results and forecast trends. It plays a crucial role in research, analysis, and decision-making across sectors such as social sciences, business and healthcare. Maintaining high accuracy in data collection is essential for making business decisions and ensuring research validity. Analysts or researchers must determine the types of data, their sources, and appropriate methods before starting the data collection process. Data collection for research or analysis involves two main methods:
- Primary data collection: This method refers to gathering origina data directly from sources, allowing it to be customised to meet specific objectives. The key primary data collection sources include:
- Surveys and questionnaires: Surveys and questionnaires are used to collect data through face-to-face interviews, phone calls, mail or online platforms.
- Observations: Observation methods involve observing an recording behaviours or events in their natural setting without interfering.
- Experiments: Experiment methods involve researcher who manipulate variables to observe their effects, providing insights into cause-and-effect relationships.
- Secondary data collection: This method involves using existing data that was originally collected for other purposes. The key secondary data collection sources include:
- Published sources: Published sources include books, academic journals, magazines and newspapers.
- Online databases: Online platforms that provide access to variety of secondary data, such as research articles, statistics, and economic data.
- Publicly available data: Information shared on publish websites, platforms, or social media by individuals or organisations.
- Previous research studies: Data from past research, which can provide valuable insights or a foundation for new research.
The collection of data involves numerous ethical considerations that are meant to achieve fairness, accuracy, and regard for the rights and privacy of individuals. They include the following considerations
- Privacy and consent: The collection of information should respect the privacy of people, and explicit consent should be sought if it is appropriately needed. Only such information as absolutely required should be gathered, and participants should be informed of how their information will be used, stored and shared. That data which happens to be sensitive requires high levels of consent and security measures.
- Transparency: Organisations need to show their processes o data collection and cleaning clearly. Contributors need to know what information is being gathered and why. Transparency is also important in explaining how data will be processed so that the findings, which may affect a decision or might be a perception of the public, do not get misinterpreted.
- Bias and fairness: Depending upon the process used to collect the data, it may inherit biases in society or throughout history. For instance, a biased sampling that could lead to unfair or discriminatory outcomes might have been done if a certain group of people was underrepresented. Cleaning data for the elimination of bias is very challenging but necessary to make results fair and equitable.
- Accuracy and integrity: Ethical data cleaning seeks accuracy without alteration to ensure the cleanliness of data. Over cleanliness might eliminate the very valid data points one aims to keep, thereby producing biased results. Data integrity and an audit trail are, therefore, extremely important to make sure that all the data remain a good representation of reality
- Security and confidentiality: The data collected, especially containing personal identifiers, should be kept away from unauthorized access and breaches. More often, anonymised or pseudonymized data is used to ensure the security of a person’s identity.
- Purpose limitation: Data must be collected for specific, legitimate purposes. Use of data for purposes other than those for which it is collected and particularly for activities that may cause harm to individuals or groups is unethical. Regular audits of data use help to support this principle.
- Accountability: Organizations must be accountable to themselves about their data practices. It is part of the ethical responsibility that there are avenues provided for participants to seek information about their data, correct errors in the data, or revoke consent.
Having explored the various methods of data collection, it is now essential to understand the different types of data that are commonly utilised in analytics which we will learn next.
Data Cleansing: Common Issues with Data
2.4 Data Cleansing: Common Issues with Data
Data cleansing, also known as data cleaning or data scrubbing, involves identifying and correcting errors or inconsistencies within datasets to improve their accuracy. High-quality and error free data is critical for making informed decisions, conducting precise analysis, and maintaining the reliability of business operations. During the data cleansing process, several common challenges may arise, which require effective solutions to address them. The common issues concerning data are:
- Missing values: These occur when data entries are empty or null It could be due to oversight, errors in data collection, or systems not capturing information. Strategies to handle missing values include deletion, imputation (filling in missing values based on statistical methods), or using algorithms that can handle missing data.
- Duplicate entries: These arise when the same data is recorded more than once within a dataset. They skew analysis by overrepresenting certain information. Identification, which involves comparing entries for similarity and then removing or consolidating duplicate records.
- Inconsistent formatting: It leads to issues when merging datasets or performing calculations. For instance, dates entered in various formats (e.g., MM/DD/YYYY, DD-MM-YYYY) or categoric with different spellings (e.g., ‘USA,’ ‘U.S.A,’ ‘United St Standardising formats ensures uniformity across the dataset
- Outliers: These are the data points that significantly differ fro other observations in a dataset. They can distort statistical analyses and model predictions. Identification involves statistical methods such as Z-score, Interquartile Range (IQR), or visualisat techniques like box plots. Deciding whether to remove, adjust, or keep outliers depends on the context of the analysis.
- Inaccurate data: Data entry errors, typos, or incorrect information can affect the integrity of a dataset. Verifying data against reliable sources or cross-referencing with other datasets helps identify and rectify inaccuracies.
- Data imbalance: The data might be imbalanced in classification problems, meaning one class significantly outnumbers the others. Balancing techniques such as oversampling, under sampling, or using specialised algorithms help mitigate this issue.
- Data irrelevance: It happens when some data do not make a major impact on the analysis. Identifying and removing irrelevant data can streamline the dataset and improve analysis accuracy and efficiency.
Having learned about the common issues of data during the cleansing process, we next focus on addressing missing data and the presence of outliers.
Treating Missing Values
2.4.3 Treating Missing values
- Missing values in data are one of the common challenges encountered across various fields that rely on datasets. They can result from several factors, including human error during data collection and limitations in data gathering methods. Fortunately, there are strategies to manage missing data and mitigate its impact in an analysis. The two primary approaches are as follows:
- Deletion: This approach involves removing rows or columns that contain missing values. While this method is straightforward, it can create significant issues if a large portion of the data is missing. Excessive deletion may compromise the reliability of the conclusions.
- Imputation: This approach replaces missing values with estimates. There are several imputation techniques available, each with its advantages and disadvantages:
- Mean/median/mode imputation:In this approach, the missing entries are replaced with the average (mean), middle value (median), or most frequently occurring value (mode) of the relevant column. Though quick and simple, this approach can introduce bias if the missing data is not randomly distributed.
- K-Nearest Neighbors (KNN) imputation: This approach identifies the closest data points (neighbours) based on available features and uses their values to estimate the missing value. KNN is particularly effective when there is large dataset where the missing values are dispersed.'
- Model-based imputation: This approach involves certain a statistical model to predict missing values based on other features in the dataset. While powerful, it requires a higher level of expertise and can be computationally intensive.
SUMMARY
- Data guides decision-making and problem-solving by uncovering patterns and trends.
- The data collection process systematically gathers information from various sources, ensuring accuracy and reliability essential for research, analysis, and informed decision-making in diverse fields.
- Structured data is a type of data that is organised and easily managed using traditional data management tools such as spreadsheets, databases or tables.
- Unstructured data is the type of data that does not have a predefined format, making it difficult to manage using traditional data management tools.
- Examples of semi-structured data include XML and JSON files, which have some organisation but also contain elements of unstructured data.
- Big Data is a term used to describe large and complex data sets that cannot be processed using traditional data management tools.
- Data cleansing addresses issues such as missing values, outliers and inaccuracies, ensuring the reliability and accuracy of datasets for effective analysis.
- Imputation methods fill in missing data points with techniques like mean, median and mode.
- Data visualisation employs graphical representations to convey patterns, trends, and insights in a clear and understandable manner
- Analytics: It involves the systematic analysis of data and the systematic analysis of data and statistics to derive meaningful insights.
- Box plot: A box plot is a graphical representation of data distribution, showing the median, quartiles, and potential outliers, helping to visualise data spread and skewness.
- IQR: The IQR is the range between the 25th and 75th percentiles, representing the middle 50% of a dataset and used to identify outliers.
- Z-score: A Z-score measures how many standard deviations a data point is from the mean, indicating its relative position within a dataset.
- Data guides decision-making and problem-solving by uncovering patterns and trends.
