Topic 1.1 - 商业分析的概念

1. 商业分析的定义

Business analytics is the use of data and analytical tools to support better decision-making:

• By combining:

  • (1) Statistical thinking
  • (2) Technological capability
  • (3) Business insight

• Core mission: Help organizations answer the fundamental question that drives business strategy

Foundation of business analytics:

1. Thinking with data

   • Understand variation, uncertainty, and patterns
   • Use statistical reasoning to support decisions

2. Technological capability

   • Access, clean, and analyze data using modern tools
   • Work with data and code

3. Interpretation and communication

   • Draw meaning from results in business context
   • Tell clear, persuasive stories to guide action

“Let the data tell its story.” - Gregory Crawford, Chief Economist at Zalando

• The story is ultimately what business analytics is trying to uncover

  • We help businesses listen to what the data is saying and translate that into actions

  • For example:

    • Customer needs
    • Competitive dynamics
    • Operational bottlenecks

• Our aim is to move beyond educated guesswork toward informed action

2. 商业分析的步骤

Business analysis is a recurring and repetitive process that follows a general steps:

1. Defining the Problem: For example:

   • What decision needs to be made?
   • What are the options?
   • What does success look like?

2. Collecting and Cleaning Data:

   • Gather relevant data sources
   • Ensure data is accurate, complete, and reliable

3. Analyzing Data Appropriately: For example:

   • Identify patterns
   • Test hypotheses
   • Quantify trade-offs

4. Communicating Insights:

   • Tell persuasive, relatable stories with data
   • Clarify the decision at hand and guide action.

5. Acting to Create Value:

   • The goal is not analysis for its own sake
   • The goal is to generate insights that drive strategy, improve operations, or better serve customers

3. 商业分析的类型

(1) 商业分析分类的重要性

Fundamental goal: To make smarter, more confident choices in an uncertain world

• Different business problems require different types of insight
• Therefore, we need different analytical tools for different questions

(2) 三种商业分析类型

(a) 描述性分析 - Descriptive Analytics: What happened?

Descriptive Analytics: Summarizing and reporting what has already occurred in a business.

Examples:

• What were our total sales last quarter?
• Which products had the highest return rates?
• What were the top 10 ASX-listed companies by shareholder return in 2024?

Tools:

• Summary statistics
• Tables
• Visualizations and dashboards

Role: Organize past data in a way that helps managers:

• Understand performance
• Spot trends
• Identify anomalies

(b) 预测性分析 - Predictive Analytics: What might happen next?

Predictive analytics: Uses past data to forecast future outcomes.

Examples:

• How many customers are likely to churn next month?
• How are annual depreciation expenses likely to change over the next three years?
• Will the RBA cut interest rates next month, and if so by how many basis points?

Tools: Techniques that identify patterns and extrapolate them forward.

Role: Give decision-makers a probabilistic view of likely outcomes.

(c) 因果分析 / 规范性分析 - Causal / Prescriptive Analytics: What will happen if we act?

Causal Analytics: What would happen if we did X instead of Y?

Examples:

• What impact will a 10% price increase have on demand?
• How will a merger affect the takeover firm’s share price?
• Would switching to a new supplier improve delivery times?

Tools: Cleverly crafted comparisons to figure out what actually causes what.

Role:

• The decision engine of analytics
• It helps businesses not just anticipate the future, but change it deliberately.

(3) 三种商业分析类型的比较

Type	Core Question	Typical Tools	Business Use
Descriptive	What happened?	Summaries & Visualizations	Reporting and monitoring
Predictive	What might happen?	Forecasting & Classification	Anticipating future trends
Causal (Prescriptive)	What will happen if we act?	Experiments & Causal Inference	Strategy and decision-making

4. 商业分析的发展历程

The evolution of business analytics is basically consistent with there kinds of business analytics types 👆.

(1) Stage 1 - 直觉决策: Gut Instinct (The HIPPO Era)

Stage Background: For most of the 20th century, decisions were dominated by experience, instinct, and authority of leadership.

Stage Characteristics:

• The most senior person in the room often had the final say (Highest Paid Person’s Opinion)
• Lacks transparency and difficult to learn from both successes and failures

(2) Stage 2 - 简单数据支持: Simple Data Support

Stage Background: With the rise of digital systems, businesses began generating and storing data and using it to create basic reports and forecasts.

Stage Characteristics:

• Create basic reports and forecasts
• Descriptive Analytics helped explain what had happened
• For example, monthly sales reports or annual budget reviews.

(3) Stage 3 - 高级分析: Advanced Analytics

Stage Background: As computing power increased, so did analytical ambition.

Stage Characteristics:

• Predictive models and machine learning to forecast future outcomes and identify hidden patterns
• Gives accurate predictions of the future
• Doesn’t answer strategic/ what if questions well

(4) Stage 4 - 实验与因果学习: Experimentation and Causal Learning

Stage Background: The frontier today is not just prediction - it’s learning what actually works.

Stage Characteristics:

• Compare how decisions affect outcomes to learn what works best
• Answer questions like: what happens if we do X instead of Y?
• Modern business analytics is as much about learning and adapting as it is about measuring and predicting

5. 商业分析的局限性

There are two major limitations of business analytics:

1. Good decisions depend on good data

   • Principle: Garbage in, garbage out

     • Analytics doesn’t make decisions, it supports decision-making
     • The quality of that support depends on the inputs we provide, and the care with which we interpret the results

   • Implication: How data is collected, stored and managed influences the quality of analytics that follows

2. Analytics doesn’t eliminate uncertainty or judgment

   • Principle: All models are simplifications

     • Models help us focus on what matters
     • However, models also leave things out

   • Implications: Even with data & models:

     • (1) Uncertainty remains: We make decisions under:

       • Incomplete information - source is imperfect
       • Ambiguous results - result is imperfect

     • (2) Trade-offs persist:

       • No model can fully resolve business tensions
       • These are managerial choices, not statistical ones
       • For example, short-term gains vs long-term investments, or efficiency vs innovation

     • (3) Judgment is essential:

       • Good analysts don’t just perform analysis
       • Good analysts think critically, question results, and connect the data to the broader context