DATA

ANALYTICS &
BUSINESS
INTELLIGENCE:
Driving Decisions in the Modern
Ghanaian Enterprise
 LEARNING OBJECTIVES
● Upon completion, you will be able to:
○ Define key data management concepts (databases, DW, ETL).
○ Compare Relational and NoSQL databases.
○ Explain the purpose and core functions of BI tools (Power BI, Tableau).
○ Differentiate between Descriptive and Predictive Analytics and their
techniques.
○ Identify best practices for data visualization and dashboard design.
○ Explain the link between data analysis and effective decision-making.
○ Recognize the application and challenges of DA & BI in the Ghanaian
context.
FOUNDATIONS OF DATA
MANAGEMENT

MODULE 1
 WHY DATA MANAGEMENT MATTERS

● Definition: Data Management encompasses the processes and policies

used to acquire, store, organize, protect, and use an organization's data.
● Importance: The foundation for ALL reliable analysis and BI for data-driven
analysis.
● "Garbage In, Garbage Out" (GIGO): Poor quality data leads to poor quality
insights and decisions.
● Key Concerns: Data Quality, Security, Governance, Integration.
● Context: Massive data growth globally and in Ghana (mobile data, financial
transactions, social media) makes robust management essential for
competitiveness.
 DATABASES: THE CORE STORAGE

Key Concepts
● Database: An organized collection of structured information, or data,
typically stored electronically.
● Database Management System (DBMS): Software used to create,
manage, and query databases.
● Two Main Database Types: SQL(or Relational) and NoSQL
Databases.
1. Relational Databases (SQL)

• Manages data using a collection of related tables

• Structure: Tables ( Rows, Columns). Schema-on-write (i.e. a
pred-determined structure for storing data).
• Examples: MySQL, PostgreSQL, Oracle, SQL Server.
• Use Cases: Transactional systems (Banking, Retail Point-of-Sale
(POS), HR records).
• Characteristics: ACID = Atomicity, Consistency , Isolation,
Durability
 A transaction is all or nothing. Either all operations
A Atomicity within it succeed, or none are applied. No partial
updates.
A transaction brings the database from one valid
C Consistency state to another. All data must follow defined rules
(e.g., constraints, relationships).
Transactions are independent. Even when multiple
I Isolation run at the same time, each behaves as if it’s the only
one accessing the database.
Once a transaction is committed, it is permanently
D Durability
saved, even if there’s a system crash or power failure.
2. NoSQL Databases:
• NoSQL = "Not Only SQL" – a family of databases designed for
flexibility, scalability, and performance, especially for
handling large volumes of unstructured or semi-structured
data.
• Use Cases: Big Data, Social Media Feeds, IoT Data, Real-time
Web Apps, Content Management
• Four main types of NoSQL databases: Document (e.g.,
MongoDB); Key-Value (e.g., Redis); Column-Family (e.g.,
Cassandra); Graph (e.g., Neo4j)
■ Structure: Flexible schema. Schema-on-read.
■ Key Concepts: BASE Properties (Basically Available, Soft state,
Eventually consistent), High Scalability & Availability.
■ Examples of NoSQL Databases:.
■ Document (e.g., MongoDB)
■ Key-Value (e.g., Redis)
■ Column-Family (e.g., Cassandra)
■ Graph (e.g., Neo4j)
 DATA WAREHOUSES & ETL (EXTRACTION,
TRANSFER, LOAD)

● Data Warehouse (DW):

○A central repository of integrated data from one or more disparate
sources.
○ Optimized for querying, reporting, and analysis (OLAP - Online
Analytical Processing).
○ Purpose: Provide a "single source of truth" for decision-making.
○ Characteristics: Subject-Oriented, Integrated, Time-Variant, Non-
Volatile.
● ETL (Extract, Transform, Load):

○ ETL is process by which messy, inconsistent data is extracted

into useful, structured information such as a data warehouse
○Extract: Retrieving data from various operational systems
(databases, logs, APIs, flat files).
○ Transform: Cleaning, standardizing, validating, aggregating,
and restructuring data. (Crucial for data quality!).
○ Load: Depositing the transformed data into the Data
Warehouse
INTRODUCTION TO BUSINESS
INTELLIGENCE TOOLS

MODULE 2
 WHAT IS BUSINESS INTELLIGENCE (BI)?

● Definition: Technologies, applications, and practices for the

collection, integration, analysis, and presentation of business
information.
● Purpose: To support better, more informed business decision-
making.
● Focus: Typically, historical and current data – "What
happened?" and "What is happening now?"
● BI vs. Analytics: BI often focuses on reporting and dashboards;
Analytics delves deeper into "Why?" and "What will happen?".
 KEY BI TOOL CAPABILITIES
● Data Connection (Databases, Files, Cloud Services, APIs)
● Data Preparation (Basic cleaning, joining tables, simple
transformations)
● Data Modeling (Defining relationships between data tables)
● Data Visualization (Creating charts, graphs, maps)
● Reporting (Creating static or interactive reports)
● Dashboarding (Consolidating multiple visualizations into a single
view)
● Sharing & Collaboration (Publishing reports, setting permissions)
 FOCUS ON TOOLS: POWER BI & TABLEAU
● Market Leaders: Two of the most widely used BI platforms globally
and increasingly in Ghana.
● Microsoft Power BI:
○ Part of the Microsoft Power Platform.
○ Strong integration with Excel, Azure, Microsoft 365.
○ Often perceived as having a lower entry barrier, especially for
organizations already using Microsoft products.
● Tableau (Salesforce):
○ Often praised for its powerful and flexible visualization capabilities
and intuitive user interface.
○ Strong community support.
 POWER BI VS. TABLEAU: FEATURE OVERVIEW
Feature Microsoft Power BI Tableau (Salesforce)
Ease of Use Often considered easier for Generally intuitive interface,
beginners, especially those familiar strong focus on visual
with Excel. exploration.
Visualization Good range of standard visuals, Excellent, highly flexible and
Capabilities custom visuals available. interactive visualizations are
a key strength.
Data Wide range of connectors, strong Very broad connectivity to
Connectivity integration with Microsoft sources diverse data sources.
(Azure, SQL Server).
 POWER BI VS. TABLEAU: FEATURE OVERVIEW
Feature Microsoft Power BI Tableau (Salesforce)
Data Power Query (shared with Tableau Prep Builder offers
Preparation Excel) provides robust data strong visual data
transformation capabilities. preparation features.
Integration Deep integration with Integrates with Salesforce
(esp. Microsoft 365, Teams, CRM, but less native
Microsoft) SharePoint, Azure. integration with Microsoft
Office suite.
 POWER BI VS. TABLEAU: FEATURE OVERVIEW
Feature Microsoft Power BI Tableau (Salesforce)
Pricing Free Desktop version, Creator/Explorer/Viewer
Model relatively low-cost Pro license, licenses, generally
Premium capacity options. considered more
expensive per user.
Common Often seen in organizations Popular in marketing,
Ghanaian using Microsoft suite (Govt consulting, data-heavy
Use Cases agencies, Banks, large corps). analysis roles, diverse
industries.
UNDERSTANDING DATA ANALYTICS
TECHNIQUES

MODULE 3
 WHAT IS DATA ANALYTICS?
● Definition: The science (and art) of examining raw data with
the purpose of drawing conclusions about that information.
● Goal: To discover meaningful patterns, trends, and insights to
inform decisions.
● Scope: Broader than traditional BI, often employing statistical
methods and machine learning. It's the engine that powers
BI insights.
 DESCRIPTIVE ANALYTICS: WHAT HAPPENED?
● Definition: Summarizing historical data to understand past events and
performance. The foundation of data analysis.
● Common Techniques:
○ Frequency: Count, Percentage (e.g., % of customers in Accra).
○ Central Tendency: Mean (Average), Median (Middle Value), Mode (Most
Frequent Value).
○ Dispersion/Variation: Range (Max-Min), Variance, Standard Deviation
(Spread around the mean).
● Use Cases: Standard Business Reports (Sales, Marketing, Operations), Website
Traffic Analysis, Social Media Engagement Reports.
● Example: A Ghanaian telco calculating Average Revenue Per User (ARPU) for
the last quarter.
 PREDICTIVE ANALYTICS: WHAT COULD HAPPEN?
● Definition: Using historical data, statistical algorithms, and machine learning
techniques to identify the likelihood of future outcomes.
● Purpose: To forecast future trends, behaviors, or events.
● Basic Concepts (Conceptual Introduction):
○ Regression: Predicting a continuous value (e.g., forecasting sales volume based on

marketing spend).
○ Classification: Predicting a category (e.g., identifying customers likely to churn:

Yes/No).
○ Clustering: Grouping similar items together without predefined labels (e.g.,

segmenting customers based on purchasing habits).

● Use Cases: Customer Churn Prediction, Fraud Detection, Demand Forecasting, Credit
Risk Assessment.
● Example: A Ghanaian bank predicting the probability of loan default for applicants.
DESCRIPTIVE VS. PREDICTIVE ANALYTICS SUMMARY

Feature Descriptive Analytics Predictive Analytics

Goal Summarize past data; Forecast future outcomes;
Understand what happened. Predict what could happen.

Key Question What were the sales last What will sales be next quarter?
quarter? How many users Which users are likely to churn?
churned?
Techniques Aggregations (Sum, Avg), Regression, Classification,
Frequencies, Basic Stats (Mean, Clustering, Machine Learning
Median, Std Dev), Reporting. Algorithms, Forecasting.
DESCRIPTIVE VS. PREDICTIVE ANALYTICS SUMMARY

Feature Descriptive Analytics Predictive Analytics

Examples Sales reports, Website Fraud detection, Demand
dashboards, Social media forecasting, Customer churn
metrics. prediction, Credit scoring.
Complexity Generally lower. Generally higher, requires
statistical/ML knowledge.
Typical Use Monitoring MoMo Predicting loan defaults,
(Ghana) transactions, Reporting Forecasting cocoa yields,
regional sales performance, Identifying high-risk insurance
Tracking website visits. customers.
EFFECTIVE DATA
VISUALIZATION & DASHBOARD
DESIGN

MODULE 4
 THE POWER OF VISUALIZATION
● Why Visualize?
○ Our brains process visuals much faster than text or tables.
○ Helps quickly identify trends, patterns, outliers, and
relationships.
○ Makes complex data accessible and understandable.
● Example: Anscombe's Quartet (Show the four datasets
visually – same stats, different plots). Data needs to be seen!
 COMMON CHART TYPES & USAGE
● Bar Chart: Comparing values across categories.
● Line Chart: Showing trends over continuous time or sequence.
● Pie Chart: Showing proportions of a whole (Use with caution! Best for
<5 categories).
● Scatter Plot: Showing the relationship between two numerical
variables.
● Histogram: Showing the distribution (frequency) of a single numerical
variable.
● Map: Showing geographic patterns or variations.
 COMMON CHART TYPES & USAGE
Chart Type Description Best Used For Example Application (Ghanaian
Context)

Bar Chart Uses rectangular bars (vertical or Comparing quantities Comparing market share of MTN
horizontal) to compare values across across different groups. MoMo, Vodafone Cash, AirtelTigo
discrete categories. Money; Sales per region.

Line Chart Connects data points with lines to Showing trends over time Tracking the monthly Cedi/Dollar
show trends or changes over a or sequence. exchange rate; Growth of internet
continuous interval (usually time). subscribers over years.

Pie Chart Circle divided into slices representing Showing proportions of a Breakdown of electricity
parts of a whole (percentages). whole (use cautiously, best generation sources (Hydro,
for <5 categories). Thermal, Solar); Budget allocation.
 COMMON CHART TYPES & USAGE
Chart Type Description Best Used For Example Application (Ghanaian
Context)
Scatter Plot Uses dots to represent values for Showing the relationship Plotting rainfall amount vs. maize
two different numeric variables. or correlation between yield; Customer age vs. average
Position indicates values on two numeric variables. spend.
horizontal/vertical axes.

Histogram Uses bars to show the frequency Understanding the Distribution of ages of mobile banking
distribution of a single numeric distribution of a single users; Distribution of exam scores.
variable. Data is grouped into variable.
continuous bins.
Map Displays data geographically. Uses Showing spatial patterns, Mapping regional internet
color, size, or symbols to represent concentrations, and penetration rates; Distribution of
values across regions or locations. regional variations. health facilities; Cocoa production by
 DASHBOARD DESIGN PRINCIPLES
● Dashboard: A visual display of the most important information needed to achieve one or
more objectives; consolidated and arranged on a single screen so the information can be
monitored at a glance. (Stephen Few)
● Best Practices:
○ Audience & Purpose: Design for specific users and the decisions they need to make.

○ Right Visuals: Choose appropriate charts for the data and message.

○ Simplicity & Clarity: Avoid clutter ("Less is More"). Use whitespace effectively. Limit colors

and fonts.
○ Context: Provide clear titles, labels, units, and comparison points.

○ Highlight Key Info: Use size, position (top-left is prime), and color strategically.

○ Accuracy & Timeliness: Ensure data is correct and up-to-date.

○ Layout & Flow: Organize logically. Group related metrics.

○ Interactivity (Optional): Use filters, drill-downs judiciously.

DATA FOR DECISIONS: BI, ANALYTICS &
DECISION SUPPORT

MODULE 5
 CONNECTING DATA TO DECISIONS
● The Goal: Move from raw data to information, then to
insights, and finally to informed actions/decisions.
● Data-Driven Decision Making (DDDM): The practice of basing
decisions on the analysis of data rather than purely on
intuition or experience.
● DDDM vs. Intuition: Data provides evidence to support or
challenge intuition, leading to more objective and often
better outcomes.
 DECISION SUPPORT SYSTEMS (DSS)
● Definition: Computer-based systems that help decision-makers
confront ill-structured problems through direct interaction with
data and analysis models. (Sprague & Carlson)
● Role: Provide the right information to the right person at the right
time to aid specific decisions.
● Modern BI Platforms: Often serve as the front-end or core
component of modern data-driven DSS.
● Types (Briefly): Model-driven (simulations, optimization), Data-
driven (OLAP, data mining), Knowledge-driven (expert systems).
 BENEFITS & CHALLENGES OF DDDM IN GHANA

● Potential Benefits:
○ Improved Operational Efficiency (e.g., optimizing delivery routes).

○ Enhanced Customer Understanding & Targeting.

○ Better Risk Management (e.g., credit scoring, fraud detection).

○ Identification of New Market Opportunities.

○ Increased Transparency & Accountability.

● Common Challenges in Ghana:

○ Data Availability, Quality, and Accessibility (especially integrating informal sector
data).
○ Shortage of Skilled Data Professionals.

○ Cost of Tools & Infrastructure (including reliable power/connectivity).

○ Organizational Culture & Resistance to Change.

○ Data Privacy & Security Concerns (Compliance with Data Protection Act, 2012).
DATA ANALYTICS & BI IN ACTION:
GHANAIAN CASE STUDIES

MODULE 6
 CASE STUDY 1: MOBILE MONEY & FINANCIAL
INCLUSION

● Context: High mobile penetration and MoMo adoption in Ghana (MTN,

Vodafone, AirtelTigo).
● Data Use:
○ Descriptive: Tracking transaction volumes, user growth, agent
distribution, popular services (P2P, bill pay).
○ Predictive: Fraud detection algorithms, customer churn analysis, credit
scoring for micro-loans, targeted marketing.
○ BI: Dashboards for monitoring service health, agent network
performance, regional uptake.
● Impact: Understanding user behavior, improving service reliability,
managing fraud risk, tailoring financial products, informing financial
inclusion strategies.
 CASE STUDY 2: AGRICULTURE & SUPPLY CHAIN
● Context: Key sector for Ghana's economy (Cocoa, Cashew, Maize, etc.).
● Potential Data Use (Illustrative):
○ Descriptive: Analyzing historical yields vs. weather patterns, tracking
market prices (Esoko platform example), mapping farm locations.
○ Predictive: Crop yield forecasting (using weather/satellite data),
pest/disease outbreak prediction, optimizing fertilizer/water use.
○ BI: Dashboards for Ministry of Food and Agriculture or COCOBOD
monitoring national/regional production, pricing trends, export volumes.
Logistics optimization for aggregators.
● Potential Impact: Increased farm productivity, reduced losses, better price
transparency, improved food security, efficient supply chains.
 CASE STUDY 3: RETAIL & CUSTOMER BEHAVIOUR

● Context: Growing formal retail (Supermarkets like Shoprite, Game) and E-

commerce (Jumia, local platforms).
● Data Use:
○ Descriptive: Sales analysis (by product, category, store, time), Market Basket
Analysis (what items are bought together?), customer demographics.
○ Predictive: Customer segmentation, personalized
recommendations/promotions, inventory demand forecasting, optimizing
store layout/website design.
○ BI: Sales performance dashboards, inventory level monitoring, customer
lifetime value tracking.
● Potential Impact: Reduced stockouts/overstocking, effective marketing
campaigns, improved customer loyalty, increased revenue.
ETHICAL CONSIDERATIONS IN DATA
ANALYTICS

MODULE 7
 DATA ETHICS & RESPONSIBILITY
● Key Issues:
○ Privacy: Responsible collection, use, storage of personal data.
Compliance with Ghana's Data Protection Act, 2012 (Act 843).
Importance of consent, anonymization/pseudonymization.
○ Bias: Algorithms can learn and perpetuate existing societal biases (e.g.,
in loan applications, hiring). Need for fairness checks and mitigation
strategies.
○ Transparency: Explainability of analytical models ("Black Box" problem).
Ensuring decisions can be understood and challenged.
○ Security: Protecting sensitive data from unauthorized access, breaches,
and misuse.
● Core Principle: Ethical considerations must be integrated throughout the
entire data lifecycle, from collection to deployment.
 CONCLUSION & KEY TAKEAWAYS
● Data is a strategic asset, but requires effective Management.
● BI Tools (Power BI, Tableau) provide access to insights through reports and
dashboards.
● Analytics Techniques (Descriptive, Predictive) uncover patterns and forecast trends.
● Visualization makes data understandable; good Dashboard Design makes it
actionable.
● The ultimate goal is Data-Driven Decision Making (DDDM).
● DA & BI offer significant opportunities but face specific Challenges in Ghana.
● Ethical Responsibility is paramount.
● Takeaway: These skills are increasingly essential for professional success in Ghana
and globally.