Learning Objectives

By the end of this lesson, learners should be able to:

• Define Business Intelligence.
• Explain the relationship between BI, data analytics and decision-making.
• Describe the major components of a BI environment.
• Distinguish operational, analytical and strategic reporting.
• Explain the purpose of dashboards.
• Identify characteristics of effective dashboards.
• Design KPI-driven business dashboards.
• Explain dashboard filtering and drill-down.
• Identify common dashboard design failures.
• Evaluate dashboards from an executive decision-making perspective.

1. Introduction to Business Intelligence

Business Intelligence (BI) refers to the technologies, processes and practices used to collect, integrate, analyze and present organizational information to support better decision-making.

BI brings together:

• Data.
• Databases.
• ETL/data preparation.
• Analytics.
• Reporting.
• Visualization.
• Dashboards.
• Decision-making.

A simplified BI process is:

Data Sources → Data Preparation → Data Storage → Analysis → Visualization → Decision

2. Why Organizations Use Business Intelligence

Organizations use BI to answer questions such as:

• What is happening?
• Why is it happening?
• Where is performance changing?
• Which customers are most valuable?
• Which branches are underperforming?
• Are targets being achieved?
• Where should management intervene?

BI therefore turns organizational data into decision-support information.

3. Business Intelligence Versus Data Analytics

The concepts overlap but are not identical.

Business Intelligence traditionally emphasizes the structured reporting, monitoring and analysis of organizational performance.

Data analytics is broader and may include:

• Descriptive analytics.
• Diagnostic analytics.
• Predictive analytics.
• Prescriptive analytics.

BI often forms part of the wider analytics environment.

4. Business Intelligence Architecture

A simplified BI architecture may contain:

Data Sources

Examples:

• ERP systems.
• CRM systems.
• Accounting systems.
• Websites.
• Spreadsheets.
• Transaction systems.

Data Integration

Data is extracted, transformed and loaded or otherwise prepared for analytical use.

Data Storage

Examples:

• Data warehouses.
• Data marts.
• Analytical databases.

Analytics Layer

Analysts query and analyze the data.

Presentation Layer

Information is delivered through:

• Reports.
• Dashboards.
• Scorecards.
• Visualizations.

5. Operational Reporting

Operational reports support day-to-day activities.

Examples:

• Today’s sales.
• Current inventory.
• Outstanding customer orders.
• Daily production.
• Pending payments.

They generally focus on immediate operational requirements.

6. Tactical Reporting

Tactical reports support middle-management decisions.

Examples:

• Monthly branch performance.
• Department budgets.
• Sales team performance.
• Customer segment analysis.

They help managers monitor performance and adjust operations.

7. Strategic Reporting

Strategic reporting supports senior leadership.

Examples:

• Revenue growth.
• Profitability.
• Market expansion.
• Strategic KPI performance.
• Long-term financial trends.

Strategic reports generally focus on broader organizational objectives.

8. What Is a Dashboard?

A dashboard is a visual interface that presents selected information and performance indicators in a consolidated view.

A dashboard may include:

• KPI cards.
• Charts.
• Tables.
• Filters.
• Alerts.
• Trend indicators.
• Targets.

Its purpose is to help users monitor and understand important information efficiently.

9. Dashboard Versus Report

A report may contain detailed information across multiple pages.

A dashboard generally emphasizes:

• High-level information.
• Monitoring.
• Visual communication.
• Interactivity.
• Rapid interpretation.

The two are complementary rather than interchangeable.

10. Executive Dashboards

An executive dashboard should focus on strategic indicators.

For example:

Revenue

KSh 240M

Revenue Growth

+8.7%

Operating Margin

21.4%

Customer Retention

89%

Executives generally do not need every transaction on the first screen.

They need information that helps them understand organizational performance and decide where attention is required.

11. KPI Selection

A dashboard becomes less effective when it contains too many KPIs.

KPIs should be:

• Relevant.
• Measurable.
• Understandable.
• Aligned with objectives.
• Actionable.

The analyst should ask:

“If this KPI changes significantly, what decision could management make?”

If the answer is unclear, the KPI may not belong on the primary dashboard.

12. Leading and Lagging Indicators

Lagging indicators

Measure outcomes that have already occurred.

Examples:

• Annual profit.
• Revenue.
• Customer churn.

Leading indicators

May provide early signals of future outcomes.

Examples:

• Sales pipeline.
• Website enquiries.
• Customer complaints.
• Production backlog.

A strong dashboard may combine both.

13. Targets and Variance

A KPI becomes more useful when compared against a benchmark.

For example:

Actual Revenue: KSh 9.2M

Target: KSh 10M

Variance: −KSh 0.8M

Without the target, the significance of KSh 9.2M is uncertain.

14. Dashboard Filters

Filters allow users to change the scope of displayed information.

Examples:

• Date.
• Region.
• Product.
• Customer segment.
• Department.

For example:

Select “Nairobi” to view Nairobi-specific performance.

Filters can make dashboards flexible but excessive filtering options can increase complexity.

15. Drill-Down

Drill-down allows users to move from summary information to more detailed levels.

For example:

Total Revenue

↓

Region

↓

Branch

↓

Product Category

↓

Product

This allows users to investigate the source of a performance change.

16. Drill-Through

Drill-through takes users from one analytical view to another detailed page or report.

For example:

Clicking a branch may open a detailed branch-performance report.

This is different from simply changing the level of aggregation within the same visual.

17. Dashboard Layout

A dashboard should have a clear information hierarchy.

A common structure is:

Top

Key KPIs.

Middle

Major trends and comparisons.

Bottom

Detailed analysis or supporting information.

The exact design depends on the business context.

18. Dashboard Navigation

A large BI solution may contain multiple pages:

• Executive Overview.
• Sales.
• Finance.
• Customers.
• Operations.

Navigation should be intuitive.

Users should understand:

• Where they are.
• What information they are viewing.
• How to move to related analysis.

19. Dashboard Interactivity

Interactive dashboards may allow users to:

• Filter.
• Drill down.
• Select periods.
• Compare segments.
• Highlight data.
• View tooltips.

Interactivity is valuable when it supports meaningful exploration.

It should not be added merely because the BI platform makes it possible.

20. Dashboard Refresh

A dashboard is only useful if the information is sufficiently current for its purpose.

Different dashboards may require different refresh frequencies:

• Real-time.
• Hourly.
• Daily.
• Weekly.
• Monthly.

A monthly strategic dashboard may not require real-time data.

21. Data Freshness

Analysts should clearly understand:

When was the data last updated?

A dashboard showing yesterday’s information without indicating its freshness may cause users to assume that the information is current.

Displaying a “Last Updated” timestamp can improve transparency.

22. Dashboard Performance

A dashboard that takes several minutes to load may discourage users from using it.

Performance can be affected by:

• Large datasets.
• Complex calculations.
• Too many visuals.
• Poor data models.
• Inefficient queries.
• Excessive filters.

Dashboard design therefore includes technical considerations as well as visual design.

23. Dashboard Consistency

Organizations should use consistent definitions.

For example, if one dashboard defines:

Active Customer = customer who purchased within 90 days

another dashboard should not silently use:

Active Customer = customer who purchased within 180 days.

Inconsistent definitions can cause management to make decisions using conflicting information.

24. Single Source of Truth

A BI environment may aim to establish a single source of truth for important organizational metrics.

This does not necessarily mean one physical database.

It means that important metrics should have:

• Agreed definitions.
• Controlled calculations.
• Reliable data sources.
• Consistent reporting.

25. Dashboard Governance

Governance addresses questions such as:

• Who owns the dashboard?
• Who can access it?
• Who approves KPI definitions?
• How often is it reviewed?
• What happens when source data changes?

Governance is important because dashboards can influence major organizational decisions.

26. Security and Access Control

Not every employee should necessarily see every metric.

For example:

• HR information may be restricted.
• Salary information may be confidential.
• Financial information may have access controls.
• Customer information may require protection.

BI systems should therefore incorporate appropriate access controls.

27. Role-Based Dashboards

Different users may need different information.

Sales Manager

May need:

• Sales pipeline.
• Conversion rate.
• Revenue.
• Salesperson performance.

Finance Manager

May need:

• Revenue.
• Expenses.
• Cash flow.
• Budget variance.

CEO

May need:

• Strategic KPIs.
• Revenue growth.
• Profitability.
• Market performance.

One dashboard does not necessarily serve every audience effectively.

28. Dashboard Storytelling

A dashboard should guide users toward meaningful questions.

For example:

Revenue ↓ 12%

↓

Retail Segment ↓ 22%

↓

Nairobi Region ↓ 28%

↓

Inventory availability ↓ 19%

This sequence helps users investigate the potential cause of a performance problem.

29. Exception-Based Reporting

Executives often benefit from seeing exceptions rather than reviewing every normal result.

Examples:

• Branches below target.
• Products with abnormal returns.
• Expenses above budget.
• Customers at high churn risk.

This allows attention to focus where intervention may be required.

30. Dashboard Design Mistakes

Common problems include:

• Too many KPIs.
• Too many charts.
• Poor labeling.
• Inconsistent colors.
• Unclear targets.
• Excessive filters.
• Slow performance.
• Missing data-refresh information.
• Unclear metric definitions.
• Decorative rather than analytical design.

31. Dashboard Evaluation Framework

A dashboard can be evaluated using five questions:

Relevance

Does it contain information users actually need?

Accuracy

Are the numbers correct?

Clarity

Can users understand the visuals?

Context

Are targets, comparisons and time periods available?

Actionability

Can the information support a decision?

32. Example Executive Dashboard

Consider a manufacturing company.

The dashboard could contain:

KPIs

• Revenue.
• Gross margin.
• Production volume.
• Defect rate.

Trend

Monthly revenue.

Comparison

Actual versus budget.

Operations

Production versus capacity.

Exception

Factories exceeding defect-rate thresholds.

This provides a balanced view of financial and operational performance.

33. BI and Decision-Making

Business Intelligence does not make decisions automatically.

Instead, it provides evidence that supports human judgment.

For example:

Dashboard → Sales decline detected → Analyst investigates → Cause identified → Management chooses response.

The quality of the decision depends not only on the dashboard but also on:

• Data quality.
• Analytical reasoning.
• Business knowledge.
• Management judgment.

34. BI Limitations

Business Intelligence may fail to provide useful insight when:

• Data is incomplete.
• Data definitions conflict.
• Source systems are unreliable.
• KPIs are poorly designed.
• Users lack analytical skills.
• Dashboards focus on appearance rather than decisions.

Technology alone does not create business intelligence.

Lesson Summary

Business Intelligence transforms organizational data into information that supports monitoring, analysis and decision-making.

Effective BI requires:

• Reliable data.
• Well-defined metrics.
• Appropriate analytical models.
• Effective visualization.
• Good dashboard design.
• Governance.
• Security.
• Clear business objectives.

A successful dashboard should answer not only:

“What is happening?”

but, where possible:

“Where should management focus next?”