Article From:

A Preface

Lack of thinking – causing “no knowing, no certainty” (does the problem occur? What’s the problem? Why? Not sure, right? Uncertain execution results? I don’t know if the boss is satisfied with the salary increase.

There are three kinds of core thinking:

  1. Structured
  2. Formulaic
  3. Business oriented

Seven skills of data analysis thinking: 1. Quadrant method 2. Multidimensional method 3. Hypothesis method 4. Index method 5. 28 rule 6. Contrast method 7. Funnel method

Exercise Analytical Ability in Business Time–Curiosity

two or three kinds of thinking details”


Focus: Find out the core arguments, break them down one by one, and then break them down to exhaustion (top to bottom of the pyramid)

Tools: Single Operations – Mind Mapping Team Operations – Card, Brainstorming

2.Formula-Everything Energy

Upper and lower are mutually computational, left and right are mutually related, and all structures are quantifiable and minimum inseparable.


+ :Addition of different types of services- Subtraction is often used to compute business logic relationshipsMultiplication and division are ratios and proportions.

Transform the structured thinking (mind map) into formula form – — transform the index that is not easy to quantify into the index that is easy to quantify, and screen out the importance of the index.

3.Business Thinking

Judging whether the analysis is suitable for business 3: Do you think from a business perspective? Really analyze the reasons? Can you put the analysis results to the ground?

To distinguish between phenomena and causes, phenomena are not the real causes of things. We should find out the real reasons.

Data is the embodiment of a certain result, but it does not represent the reason. We need to use business thinking to dig up a layer more carefully.

Increasing Business Thinking Method: Close to Business, Transposition Thinking

Summary: Structured thinking (smoothing ideas) – – structured data (digitalizing it) – – structured business data (landing, fit business)

Three, Data Analysis Thinking 7 Skills Detailed

1.Quadrant Method: Core – Strategy Driven

How to divide quadrants is uncertain. It is adjusted according to its own strategic objectives. Usually, there are average/median/fixed value and so on.

2.Multidimensional method

Many dimensions can be counted.

It can be divided into high school and low school, belong to the statistics of large data volume, for the statistics of rich dimensions.

Disadvantage: It is easy to hide the results of comprehensive statistics from the content of subdivision conclusions.

Coping methods: subdivision, drilling

3.Hypothetical method

Assuming a conclusion, consider what phenomena and circumstances this conclusion will bring to verify.

Application scenario: It is common to assume a conclusion for reverse inference when data is not available.

Usage: Assume conclusions, infer desired results based on experience and custom values.

4.index method

Application scenario: When there is a lot of data, but the data is open and don’t know how to use it – – use the index to solve the problem of measurement

Application: Create an indicator for comparison

Computing method:

  1. Linear Weighting – Simple Calculations of Additive or Additive Multiplicative Weights
  2. The inverse proportional method – – y = 1/x, y = 1-1/x, y = K / x, K can assign any value or y = x /(x + 1). [The purpose of the formula is to converge the larger data and reduce the difference between them], the method is the same as the sum of multiplication weights based on the importance degree.
  3. log Method – Use log function to reduce the number (which also converges to larger data) and then assign weight to multiply and sum.

5.28 Rules – — Focus only on the Key Points

20% of the variables in the data will directly produce 80% of the results. Data analysis focuses on these 20% of the articles.

Keeping an eye on TOPN data is a very good habit, especially in some industries.

Although there are many indicators, some indicators are more valuable. The 28th Law can not only analyze data, but also manage data.

【Data analysis thinking should not give up the overall situation or be structured and globalized. Otherwise, it is easy to become narrow-minded.)

6.Contrast Method–A Thinking Way of Mining Data Rules

A qualified analysis must use n comparisons.

The comparison of competitors, categories, characteristics and attributes, the comparison of the year-to-year cycle, the transformation of the comparison, the change before and after the comparison and so on.

7.Funnel method – single funnel method is useless. It should be combined with contrast method for analysis.

It is a process-oriented way of thinking, involving changes and processes can be applied.

Fourth, how to train data analysis thinking in business time

The data analyst’s essential ability is curiosity.

Analytical thinking is a habit that can be analyzed everywhere in daily life.

Return to the General Directory


Link of this Article: Week 1: Data Analyst Thinking

Leave a Reply

Your email address will not be published. Required fields are marked *