Operational Manual: How to Improve Data Insights?

The author of this article will teach you how to improve your data insight through several simple cases and help you develop thinking habits in the process of data analysis. I hope this article can inspire you in your data analysis work.

Many students complain that they face a lot of numbers every day but can't make sense of them. On the contrary, some business people can make accurate judgments immediately after looking at a few numbers. What's going on? Not feeling anything when looking at the data is a sign of lack of data insight. Data insight has nothing to do with operating tools, it is completely a thinking habit.

After it is established, it is not only helpful in work, but also very useful in life. Today we will explain it systematically.

1. Feel what insight is

Numbers themselves have no meaning. Only when numbers are combined with business scenarios can they have specific business meanings (as shown in the following figure):

Note that the handsome guy in the picture above went berserk not because the girl was 180cm tall, but because the girl was 180cm tall and made him look too short (and was ridiculed for this). "Comparison" is the key to the problem. Therefore, data alone cannot form a judgment, only data + standards can form a judgment. If you want to understand the meaning of data, you must look at the specific business scenario and the standards for business judgment (as shown in the figure below).

• Only with data, business scenarios, and judgment criteria can we form basic data insights. These three are indispensable.
• Without data, we will fall into the dilemma of "I saw a black apple, so all apples in the world are black."
• Without business scenarios, embarrassing things like “one woman takes ten months to give birth to a baby, but ten women can give birth to one in one month” will occur.
• Without a standard for judgment, we will be talking at cross purposes. After talking for a long time, we will find that the "good/bad" we are talking about are not the same thing at all.

2. Basic ideas for cultivating insight

Since insight comes from the combination of data, business scenarios, and judgment criteria, cultivating insight also starts from these three directions, including:

• Find data when something happens
• Understand the business scenarios in detail
• Clear judgment criteria
• Accumulate conclusions based on data judgment in specific scenarios
• Use the conclusions in new scenarios to test the effects
• Continue to accumulate correct conclusions and correct wrong conclusions

This statement looks very official, but it is very simple to implement, and we all practice it every day. For example, when looking for a partner, ignorant boys always complain that Reba is too fat and Mi Mi is bald, and fantasize about finding a fairy. But after dating and chasing a few girls, I found that "Oh, it turns out that it is so difficult to find a beautiful girl in reality!"

Then, when he really finds a "beauty" to spend some time with, he finds that personality, hobbies, living ability, and work ability are more important than appearance. In the middle of the night, when the guy is alone smoking a cigarette, facing the moon, thinking: "Why should I spend money and effort to invite a lady to come back to serve me, do I deserve to be beaten!", his insight will have a qualitative leap. Even if he sees a beautiful girl in the future, he will immediately understand: This is not my type!

In real life, the key to insight is often data. Because of the serious information asymmetry problem in life, it is too difficult to collect data, and it also costs time, money, and even prospects and future.

So in life, we often adopt the strategy of limited rationality. We try to make decisions with as little data as possible within the feasible range. Or we simply adopt the strategy of following those who are better than us. But in the enterprise, it is a completely different scene.

3. Difficulties in Developing Data Insights

In corporate work, the biggest difficulty in cultivating data insight is that data, business scenarios, and standards are separate from each other.

• Students who do data analysis do not understand the business scenarios and can only make blind guesses based on the data;
• People in the business department are confused or have their own agenda, and deliberately distort the judgment criteria;
• Not paying enough attention to data, not collecting complete basic data, and preferring to talk about individual cases when encountering problems instead of looking at the overall picture of the data;

These bad conditions will make it difficult for students doing data analysis to accumulate experience.

Therefore, we often find that the most insightful person in the company is often the boss. Because the three factors are transparent to the boss, even if he does not operate the basic data, he can still see the slightest details.

But this is not a good thing for data analysts. Because the boss is still waiting for our opinions, and letting the boss run ahead of us in everything will cause dissatisfaction. Therefore, students who work with data still need to exercise their own insight.

Steps to Develop Data Insights

When many students talk about improving their insight, they like to do these three things:

• Find "XX Industry 2020-2025 Panoramic Insight Report (Heavy and In-depth!)"
• Find the mind map of the data indicator system of the XX industry, pick the most densely packed one and save it in the D drive - dry goods folder
• Join various data analysis groups and ask: "Is there any cool data analysis report with insights? Please send it to me."

These three methods are completely useless. It's like a guy who wants to fall in love and looks at beautiful pictures on the Internet every day. If you don't practice and think concretely, it's impossible to improve your insight. Never do it, never. You have to find a way to do it yourself. And often these things have too much content, and you will never look at the stuff saved in the D drive. So it's best to start from a specific point.

1. Start with one scenario and one indicator

The advantage of students who work with data is that they have data at hand and can check it at any time. The disadvantage is that they do not understand the business scenario. Therefore, combining data with the business scenario is the key to solving the problem. It is best to start with a business that you are familiar with and a department where you have good friends. Start by understanding the result indicators (as shown below).

2. From extremes to intermediate values

After understanding the business meaning of the indicator, if you want to make a judgment, you can start with the White Rhino - first look at the maximum and minimum values ​​of the indicator. What are the scenarios in these situations, what problems occur, and what are the responses.

With the understanding of extreme values, you can master the basic judgment criteria and accumulate analytical assumptions and analysis logic. When you encounter less extreme situations, you can follow the accumulated analysis logic to understand. If you really can't interpret it, you can also choose to observe it again to see which extreme direction the data develops (as shown in the figure below).

3. From static to dynamic

When we have accumulated enough insights into static scenarios, we can interpret dynamic scenarios. In essence, dynamic scenarios are just a collection of static scenarios. It is important to remind you that business changes often have regularity. A continuous regularity itself has business significance. By accumulating regularity in periodic forms, we can improve our insight from point to line.

4. From single indicator to multiple indicators

After accumulating insights into single indicators, you can expand to multiple indicators. After mastering the judgment of result indicators, you can look at them together with process indicators. Note: Multiple indicators are not the accumulation of single indicators. When put together, the more indicators, the better. When multiple indicators are combined, they will form specific forms in specific business scenarios. The interpretation based on the forms can make more accurate judgments (as shown in the figure below).

After mastering the basic forms, you can continue to observe the changes in forms and accumulate more experience. In this way, you can gradually move from simple to complex, and as you accumulate more experience, you will naturally be able to draw inferences about other situations.

It should be noted that the specific scenarios will change when changing industries, companies, products, or development stages. Therefore, attempting to pursue the "eternal truth of data analysis" will only make you go further and further on the path of metaphysics. If you want to improve your insight, you should accumulate more specific scenario fragments and improve your specific analysis ability. Specific problems require specific analysis. This sentence will never go out of date.

Author: Down-to-earth Teacher Chen

Source: WeChat public account "Down-to-earth Teacher Chen (ID: gh_abf29df6ada8)"