Business Analytics Types

Interactive lecture demonstration — Philippine sari-sari store

Descriptive → Diagnostic → Predictive → Prescriptive

Philippine Business Scenario

Tindahan ni Ana operates fictional small retail outlets in Cagayan de Oro, Iligan, and Malaybalay.

Management wants to understand sales, explain differences, anticipate future sales, and decide what actions to take.

Classroom dataset: 400 synthetic transactions from June–July 2026. Values are fictional and are provided only for teaching.

1. Descriptive Analytics — What happened?

Descriptive analytics summarizes historical data. It answers questions such as “What happened?” using totals, averages, counts, rankings, tables, and charts.

Lecture question: How much did the stores sell, and which products generated the most sales?

Sales by Product

2. Diagnostic Analytics — Why did it happen?

Diagnostic analytics examines differences and patterns in historical data to investigate possible reasons for an observed result.

Lecture question: Why are sales different across branches and product categories?

Select branch: Sales: Transactions: Average:

Selected Branch: Sales by Category

Important: Diagnostic analytics identifies patterns and possible explanations. A difference in sales does not automatically prove that one factor caused another.

3. Predictive Analytics — What might happen?

Predictive analytics uses historical patterns and analytical models to estimate likely future outcomes.

This demonstration uses a simple 14-day moving-average baseline. It is intentionally simple for first-year students.

Forecast idea: Use recent daily sales to estimate a likely daily sales level for the next period.

4. Prescriptive Analytics — What should we do?

Prescriptive analytics uses data, predictions, business rules, or optimization to recommend actions.

Lecture question: Given what we know about sales, what action should management consider?

Simple Restocking Recommendation

This classroom demonstration uses a simple rule: prioritize products with the highest observed sales.

Real-world reminder: A real recommendation should also consider current inventory, profit margin, supplier lead time, product expiry, seasonality, available cash, and business objectives.

Distinguishing the Four Types

TypeMain QuestionTypical OutputPhilippine Retail Example
DescriptiveWhat happened?Summary, KPI, chartTotal sales last month
DiagnosticWhy did it happen?Comparison, drill-down, patternWhy one branch sold less
PredictiveWhat might happen?Forecast or probabilityExpected sales next week
PrescriptiveWhat should we do?Recommendation/actionWhich products to prioritize for restocking

The Analytics Chain

What happened? → Why? → What might happen? → What should we do?

This sequence is a useful introductory way to distinguish the four types. In real projects, organizations may move back and forth among these forms of analysis.

Download the dataset: The downloaded CSV contains the same synthetic transactions used by this application.