Page 47 - Banking Finance October 2025

ARTICLE

Different Types of Data Value: The benefits derived from the analysed data.

The fundamental prerequisite for effective data analytics is
understanding the different types of data. Data can be Data Analytics Lifecycle
categorized based on its structure, format, and source. The data analytics lifecycle provides a structured approach
to transforming raw data into actionable insights. It
a. Structured Data
encompasses several stages, each critical to the success of
Structured data is highly organized and easily searchable the analytics project.
within relational databases. It adheres to a predefined 1. Data Collection - Gathering relevant data from various
schema, making it straightforward to query and analyse. sources such as transactional systems, customer
Examples include: databases, external datasets, and real-time data
Transactional Data: Records of customer transactions, streams.
such as deposits, withdrawals, and loan payments.
2. Data Preparation - Cleaning and pre-processing data
Customer Profiles: Information like names, addresses, to ensure quality and consistency. This includes handling
account numbers, and contact details. missing values, correcting errors, and transforming data

Financial Statements: Balance sheets, income into a suitable format for analysis.
statements, and cash flow statements. 3. Exploratory Data Analysis (EDA) - Investigating data
b. Unstructured Data to uncover initial patterns, relationships, and anomalies
Unstructured data lacks a specific format or organization, through statistical and visualization techniques.
making it more challenging to analyse using traditional 4. Data Modeling - Applying statistical and machine
tools. However, it often contains valuable insights. Examples learning techniques to build predictive or descriptive
include: models that capture underlying data patterns.
Emails: Customer communications and inquiries.
5. Evaluation - Assessing the performance and validity of
Social Media Posts: Feedback, reviews, and comments models using appropriate metrics to ensure they meet
from social media platforms. the desired objectives.
Documents: PDFs, Word files, and scanned images 6. Deployment - Implementing the models into production
containing various forms of information. systems for real-time decision-making or strategic use.

c. Semi-Structured Data 7. Monitoring and Maintenance - Continuously tracking
Semi-structured data falls between structured and model performance and updating models as needed to
unstructured data. It contains organizational properties that accommodate new data or changing conditions.
make it easier to analyse than unstructured data but does
not conform to a rigid structure. Examples include: Types of Data Analytics
JSON and XML Files: Data formats that include tags
Data analytics can be categorized based on its purpose and
and keys to denote hierarchy and relationships.
the nature of insights it seeks to generate:
CSV Files: While primarily structured, they can contain 1. Descriptive Analytics - This type of analytics helps in
varied data types and missing values. understanding past data and identifying trends. For

d. Big Data example, a bank might use descriptive analytics to
generate monthly reports on the number of new
Big data refers to extremely large data sets that may be
unmanageable with traditional data processing tools. It is accounts opened, the volume of transactions, and
characterized by the "5 V". customer demographics.
Volume: The sheer amount of data generated. 2. Diagnostic Analytics - This type of analytics explains
why something happened. For instance, if a bank
Velocity: The speed at which data is generated and
notices a sudden drop in customer satisfaction scores,
processed.
diagnostic analytics can help identify the root causes,
Variety: The different types of data. such as long wait times or issues with online banking
Veracity: The reliability and accuracy of the data. services.

42 | 2025 | OCTOBER | BANKING FINANCE