Learning Objectives

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

• Distinguish between major chart and graph types.
• Select visualizations according to analytical objectives.
• Compare categorical, numerical and time-series data visually.
• Use charts to identify trends, patterns, distributions and relationships.
• Explain the strengths and limitations of common visualization techniques.
• Apply appropriate visualization techniques to business datasets.
• Identify inappropriate or potentially misleading chart choices.
• Evaluate charts from both analytical and communication perspectives.

1. Introduction to Business Charts

Charts convert numerical or categorical information into visual structures that allow patterns to be recognized more quickly.

A business analyst may work with:

• Sales data.
• Customer data.
• Financial data.
• Employee data.
• Operational data.
• Marketing data.

Each type of data may require a different visualization technique.

The objective is not to select the most sophisticated chart.

The objective is to select the chart that best answers the business question.

2. Categories of Visualization

Business visualizations can broadly be classified according to what they communicate.

Comparison

Used to compare categories.

Example:

Sales by branch.

Trend

Used to show change over time.

Example:

Monthly revenue.

Distribution

Used to understand how values are spread.

Example:

Customer transaction values.

Relationship

Used to investigate relationships between variables.

Example:

Advertising expenditure and sales.

Composition

Used to show how a total is divided.

Example:

Expenses by department.

Geographic

Used to compare values across locations.

Example:

Revenue by county.

3. Bar Charts

Bar charts are among the most useful business visualization techniques.

They compare values across discrete categories.

Example:

Product	Sales
Product A	850,000
Product B	620,000
Product C	490,000
Product D	310,000

A bar chart allows the analyst to quickly identify the ranking of products.

4. Horizontal Versus Vertical Bars

Vertical bars are useful when category names are short.

Horizontal bars can be more appropriate when:

• Category names are long.
• There are many categories.
• Ranking is important.

For example, department names with long descriptions may be easier to read horizontally.

5. Grouped Bar Charts

A grouped bar chart compares multiple measures across categories.

Example:

Actual revenue versus budgeted revenue for each branch.

Each branch can contain two bars:

• Actual.
• Budget.

This makes category-level comparison possible.

6. Stacked Bar Charts

A stacked bar chart divides each bar into components.

For example:

Total revenue by region, divided into product categories.

It can show:

• Overall size.
• Composition of the total.

However, comparing segments that do not share a common baseline can become difficult.

7. 100% Stacked Bar Charts

A 100% stacked bar chart shows the proportional composition of categories.

For example:

Percentage of customers in each income segment across different branches.

Every bar represents 100%.

This makes proportions easier to compare but hides differences in absolute totals.

8. Line Charts

Line charts are particularly effective for time-series data.

Example:

Monthly revenue from January to December.

They can reveal:

• Trends.
• Growth.
• Declines.
• Seasonal patterns.
• Turning points.

9. Multiple-Line Charts

Multiple lines can compare several series over time.

For example:

Revenue trends for Nairobi, Mombasa and Kisumu.

However, too many lines can make interpretation difficult.

A chart containing 20 overlapping lines may technically contain information but communicate very little.

10. Area Charts

Area charts extend the line chart by filling the area beneath the line.

They can be useful for:

• Showing magnitude over time.
• Showing cumulative composition.
• Comparing broad trends.

Stacked area charts can show how different components contribute to a total over time.

11. Pie Charts

Pie charts show parts of a whole.

For example:

Departmental contribution to total expenditure.

They are most effective when:

• There are relatively few categories.
• Categories form a meaningful whole.
• Differences are reasonably clear.

They become less effective with many small categories.

12. Doughnut Charts

A doughnut chart is similar to a pie chart but has a central hole.

It can be visually attractive, but the same limitations generally apply:

• Difficult comparison of similar segments.
• Problems with many categories.
• Less precision than a well-designed bar chart.

The choice should therefore be based on communication needs rather than appearance.

13. Histograms

A histogram shows the distribution of numerical data.

Suppose a company has 10,000 transaction values.

Instead of displaying every transaction, a histogram can group them into ranges:

• KSh 0–10,000.
• KSh 10,001–20,000.
• KSh 20,001–30,000.
• And so on.

This allows the analyst to see the shape of the distribution.

14. Histogram Versus Bar Chart

This distinction is frequently tested.

Bar chart

• Used for categories.
• Bars represent separate categories.
• Ordering may be changed without altering the meaning.

Histogram

• Used for numerical distributions.
• Values are grouped into intervals.
• The order of intervals matters.

15. Scatter Plots

Scatter plots display observations using two numerical variables.

For example:

X-axis: Advertising expenditure

Y-axis: Revenue

Each point represents an observation.

The analyst can look for:

• Positive association.
• Negative association.
• Clusters.
• Outliers.
• Nonlinear patterns.

16. Scatter Plot and Correlation

If points generally move upward from left to right, the variables may have a positive relationship.

If they move downward, they may have a negative relationship.

However:

Correlation does not establish causation.

Other variables may influence both measurements.

17. Bubble Charts

A bubble chart extends a scatter plot by using bubble size to represent a third numerical variable.

For example:

• X-axis = Revenue.
• Y-axis = Profit Margin.
• Bubble size = Number of Customers.

Bubble charts can communicate multiple dimensions, but excessive complexity can make interpretation difficult.

18. Box Plots

Box plots are useful for comparing distributions.

They can reveal:

• Median.
• Quartiles.
• Spread.
• Potential outliers.

For example:

Compare employee salaries across five departments.

A box plot can show whether one department has a wider salary distribution than another.

19. Heatmaps

Heatmaps represent values through varying visual intensity.

Example:

Sales performance by product and month.

The analyst can quickly identify areas of:

• High performance.
• Low performance.
• Concentration.
• Change.

Heatmaps are particularly effective for large two-dimensional tables.

20. Waterfall Charts

A waterfall chart shows how sequential positive and negative changes contribute to a final value.

For example:

Opening Profit

• Revenue increase

− Cost increase

− Tax

• Other income

= Closing Profit

Waterfall charts are useful in:

• Financial analysis.
• Budget variance analysis.
• Profit bridge analysis.
• Cash movement analysis.

21. Funnel Charts

Funnel charts are commonly used for processes where values decrease through stages.

Example:

10,000 Website Visitors

↓

3,500 Product Views

↓

1,000 Carts

↓

400 Purchases

This can help businesses analyze conversion processes.

22. Geographic Visualizations

Geographic charts display information according to location.

Examples:

• Sales by county.
• Customers by region.
• Branch performance.
• Delivery volume by location.

Maps can be highly effective when geography is central to the business question.

However, a map should not be used merely because location information exists.

23. Choropleth Maps

A choropleth map uses different shades or intensities to represent values across geographic areas.

For example:

Revenue per county.

However, raw totals can be misleading when geographic areas differ substantially in population or size.

Normalized measures may therefore be more appropriate.

24. Tree Maps

A treemap represents hierarchical categories using nested rectangles.

For example:

Total Revenue

→ Electronics

→ Furniture

→ Clothing

Each category can contain subcategories.

Treemaps are useful when analysts need to show hierarchical composition and relative magnitude simultaneously.

25. Radar Charts

Radar charts display several variables around a circular structure.

They can be used to compare profiles across dimensions.

However, when there are many categories or small differences, radar charts can become difficult to interpret.

A bar chart or table may provide clearer comparisons.

26. KPI Visualizations

KPI visualizations focus attention on specific performance indicators.

Examples:

• Revenue.
• Profit.
• Customer retention.
• Cost per acquisition.
• Inventory turnover.

A KPI should ideally show context such as:

Current value + target + previous value + variance.

27. Bullet Charts

A bullet chart can compare:

• Actual performance.
• Target.
• Performance ranges.

For example:

Actual sales = KSh 9.2M
Target = KSh 10M

This can be more informative than a simple KPI card.

28. Sparklines

A sparkline is a compact mini-chart showing a trend.

For example:

Revenue: KSh 12M ↑
[small six-month trend]

Sparklines are useful in tables and executive dashboards where space is limited.

29. Choosing Charts by Question

Consider the question:

“Which branch generated the most revenue?”

Bar chart

Question:

“How has revenue changed over 12 months?”

Line chart

Question:

“What is the distribution of transaction values?”

Histogram

Question:

“Is advertising spending associated with revenue?”

Scatter plot

Question:

“How is total expenditure divided among departments?”

Pie chart or stacked bar

Question:

“Which combinations of product and month have unusually low sales?”

Heatmap

The analytical question should drive chart selection.

30. Chart Selection Matrix

Question	Recommended Visualization
Compare categories	Bar chart
Compare categories over time	Grouped bar / line
Show trend	Line chart
Show distribution	Histogram
Identify outliers	Box plot
Examine relationships	Scatter plot
Show composition	Stacked bar
Show sequential financial changes	Waterfall
Show geographic differences	Map
Show hierarchical composition	Treemap
Monitor KPI	KPI card / bullet chart

31. Small Multiples

Small multiples use a consistent chart format repeated across several categories.

For example:

One small line chart for each region.

This can be more readable than putting 20 regional lines into a single chart.

Consistency allows users to compare patterns across panels.

32. Dual-Axis Charts

A dual-axis chart uses two vertical scales.

For example:

• Revenue.
• Profit margin.

Dual axes can be useful when variables have very different scales.

However, they can also create misleading visual relationships if scales are chosen carelessly.

They should therefore be used sparingly.

33. Visualization of Financial Data

Financial data often requires particular care.

Common visualizations include:

• Revenue trends.
• Budget versus actual.
• Profit bridges.
• Expense composition.
• Cash flow trends.
• Margin analysis.

Financial charts should clearly identify:

• Currency.
• Period.
• Units.
• Whether values are nominal or percentage-based.

34. Visualization of Operational Data

Operational dashboards may visualize:

• Production volume.
• Delivery time.
• Defect rates.
• Inventory levels.
• Service response time.
• Capacity utilization.

Operational metrics often benefit from targets and thresholds.

35. Visualization of Customer Data

Customer analytics may involve:

• Customer acquisition.
• Retention.
• Churn.
• Purchase frequency.
• Customer lifetime value.
• Segment performance.

Different customer questions may require different charts.

For example, churn over time is better represented using a line chart, while customer segments can be compared with bars.

36. Visualization of Marketing Data

Marketing analytics may visualize:

• Impressions.
• Click-through rates.
• Conversion rates.
• Cost per acquisition.
• Campaign revenue.
• Customer acquisition channels.

A funnel can be useful for conversion stages, while a bar chart can compare campaign performance.

37. Visualization of HR Data

HR analytics may include:

• Employee turnover.
• Absenteeism.
• Salary distributions.
• Recruitment performance.
• Training hours.
• Department headcount.

Box plots can be useful for salary distributions, while line charts can show turnover trends.

38. Interactivity

Modern visualization tools often allow users to:

• Filter.
• Drill down.
• Hover over data points.
• Change date ranges.
• Select categories.
• Explore detailed records.

Interactivity can improve exploration but should not compensate for poor underlying design.

39. Drill-Down

A dashboard may initially show:

Total company revenue.

The user can then drill down into:

Company → Region → Branch → Product → Transaction

This allows management to move from a high-level indicator to detailed evidence.

40. Visualization and Accessibility

Effective visualizations should consider users with different abilities.

Good practices include:

• Adequate contrast.
• Clear labels.
• Avoiding color as the only indicator.
• Readable text.
• Meaningful titles.
• Appropriate font sizes.

Accessibility is part of effective communication, not merely a technical requirement.

Lesson Summary

The choice of visualization should depend on:

• Data structure.
• Analytical objective.
• Audience.
• Number of categories.
• Required level of detail.

There is no chart that is universally superior.

A strong analyst asks:

What comparison, trend, relationship, distribution or composition does the decision-maker need to understand?

The answer determines the visualization