Unit 8.4

Descriptive Analytics: Understanding What Happened

IT 233: Business Information Systems

Learning Objectives

By the end of this session, you will be able to:

✅ Define Descriptive Analytics and its primary purpose.
✅ Identify the key question that descriptive analytics answers.
✅ Describe the main techniques used, such as reporting, aggregation, and visualization.
✅ Understand the role of dashboards in monitoring business performance.

The Foundation of Business Insight

Descriptive Analytics: The process of summarizing historical data to understand what has happened in the past.

It provides a retrospective (backward-looking) view of business operations.

It's the essential first step. Before you can predict the future, you must understand the past.

🎬 Watch: What Descriptive Analytics Is

The Core Question: "What happened?"

Descriptive analytics turns raw data into answers for past events. It helps us understand performance by answering questions like:

What were our total sales last quarter?
Which of our products was the most profitable?

How many customer complaints did we receive last month?
What was our website traffic last week?

Interactive: Is It Descriptive Analytics?

Click each question card to reveal whether it belongs to Descriptive Analytics.

Key Techniques & Tools 🛠️

We use a range of techniques to summarize data and find insights.

Reporting

Routine summaries of key metrics (daily, weekly, monthly).

Data Aggregation

Gathering and summarizing data (e.g., sum, average, count).

Data Visualization

Presenting data in graphical formats like charts and graphs.

Technique 1: Reporting & Aggregation

📊 Standard Reporting

The creation of routine reports that provide a summary of key business metrics on a regular basis.

Example: A monthly sales report emailed to all department heads.

🔢 Data Aggregation

The process of gathering data and summarizing it in a simple, statistical format.

Common aggregations:

Sum, Average (Mean), Median, Count, Percentage

Interactive: Data Aggregation Calculator

Enter monthly sales figures (comma-separated) and see live aggregations.

Sales Figures (NPR 000s):

Aggregation	Value

🧮 Worked Example: When the Average Lies

Illustrative: monthly pay (NPR thousand) at a small Kathmandu firm: 30, 35, 40, 45 for four staff, and 350 for the owner.

Mean (average): (30 + 35 + 40 + 45 + 350) ÷ 5 = 500 ÷ 5 = 100.
Median (middle value): sorted list 30, 35, 40, 45, 350, so the median is 40.
An advert saying "average pay NPR 1 lakh" would be true, and misleading.

With an extreme value (an outlier), report the median, or both.

🧪 Quick Activity: Mean or Median?

Typical house price in a district where a few mansions sold. → Median
Average daily units produced by a machine that runs steadily. → Mean
Typical delivery time when a few parcels took weeks. → Median
Average exam mark in a class with no extreme scores. → Mean

Technique 2: Data Visualization

"A picture is worth a thousand rows of data."

Data Visualization: The practice of presenting data in a graphical format, such as charts, graphs, and maps, to easily identify trends, patterns, and outliers.

Common types include bar charts, line charts, pie charts, and heat maps.

🗺️ Real Case: A Map That Stopped an Epidemic

In 1854, cholera killed hundreds of people in Soho, London. Most doctors blamed "bad air".

Doctor John Snow recorded where each death happened and marked them on a street map.
The deaths clustered around one public water pump on Broad Street.
He persuaded officials to remove the pump handle, and cases near it fell.

Pure descriptive analytics: count, place, visualise. The picture showed "what happened" when words couldn't.

John Snow's Cholera Map, 1854

Map of Soho marking cholera deaths around the Broad Street pump, 1854.

🎬 Watch: John Snow and Cholera

📈 Why Visualize? Anscombe's Quartet

In 1973, statistician Francis Anscombe created four small datasets.

All four have almost the same average, spread and correlation.
On a chart they look completely different: a straight line, a curve, a line with one outlier, and a single extreme point.
Summary numbers alone would have called them identical.

Always look at the data, not just its averages.

Anscombe's Quartet

Four scatter plots with identical summary statistics but very different shapes.

Interactive: Pick the Right Chart (1/2)

For each scenario, select the most appropriate chart type.

Interactive: Pick the Right Chart (2/2)

For each scenario, select the most appropriate chart type.

Bringing It All Together: Dashboards 🎯

A Dashboard is a data visualization tool that displays a real-time summary of Key Performance Indicators (KPIs) and other important business metrics on a single screen.

Dashboards are a very common application of descriptive analytics.

They allow managers to monitor the health of the business at a glance.

🎬 Watch: What a KPI Dashboard Is

🎯 What Makes a Good KPI?

A Key Performance Indicator is a measure that shows progress towards an important goal. Good KPIs are SMART: Specific, Measurable, Achievable, Relevant, Time-bound.

❌ Weak
"Improve customer happiness."

✅ Strong
"Raise the average satisfaction score from 3.8 to 4.2 out of 5 by Asar end." (illustrative)

Interactive: Business KPI Dashboard

A simulated descriptive analytics dashboard. Click "Refresh Data" to simulate loading a new reporting period.

Descriptive Analytics in the Real World

📈 Sales

A sales manager uses a dashboard to view total revenue by region, product, and salesperson for the previous quarter.

📢 Marketing

An analyst creates a report showing the performance of a recent ad campaign, including click-through rate and cost per acquisition.

🌐 Web Analytics

A website manager looks at a Google Analytics report to understand visitor numbers, popular pages, and traffic sources.

👥 Human Resources

An HR manager reviews a report on the employee turnover rate for the past year, broken down by department.

Local Context: Descriptive Analytics in Nepal

Scenario: An e-commerce site in Nepal (e.g., Daraz)

Managers would use a descriptive analytics dashboard to answer:

🔍 Which products were the top sellers during the Dashain festival season?
🔍 What is the average delivery time for orders within Kathmandu Valley vs. outside?
🔍 Which payment method (eSewa, Khalti, Cash on Delivery) is most popular?
🔍 How many new customers did we acquire last month?

A Key Limitation ⚡

While extremely useful, descriptive analytics has its limits.

It tells you WHAT happened, but it does not explain WHY it happened or WHAT WILL happen next.

Answering "why" requires Diagnostic Analytics.

Answering "what will happen" requires Predictive Analytics.

Interactive: Which Analytics Type Answers This? (1/2)

Match each business question to the correct analytics type: Descriptive, Diagnostic, or Predictive.

Interactive: Which Analytics Type Answers This? (2/2)