The road to a higher salary: Eight essential skills for senior data analysts

This article summarizes the eight abilities that a senior data analyst needs based on the abilities that a data analyst needs. It is recommended for students who are engaged in related work.

What abilities do senior data analysts need? This is a question in many people's minds. Popular on the Internet are Excel, SQL, Python, and data analysis from 0 to 1, but what should be done from 1 to 100? Teacher Chen has prepared a series of sharing on "Year-end Inventory of Data Analysis" for everyone. Today, let's start with "ability", which few people talk about.

First of all, let’s clarify the difference between knowledge, skills and abilities:

1. Knowledge, such as statistics, mathematics, operations research, and machine learning. Knowledge has a theoretical system and can be learned from books.

2. Skills, such as Excel, SQL, Python, Tableau and other software. These can be improved through repeated practice.

3. Abilities, such as business understanding, communication, logical thinking, reporting, etc., are very important but there is no standard tutorial.

So what abilities should a data analyst have in order to achieve growth from 1 to 100?

1. Ability 1: Business understanding

This is the first step to get rid of the low-level SQLboy and become a data acquisition tool person. Data analysts above the team leader level must communicate directly with the business about the project. The director of the data department must attend the business analysis meeting. If you don't understand the business, you can't pass it. A company's business has three levels: strategy, tactics, and combat. The understanding of data analysts is also divided into three levels (as shown below):

If you want to fully understand the business, you may need to stay in the industry for many years. However, using the value chain model + financial statements, you can quickly understand the main revenue and cost structure of an industry, quickly establish cognition, and cope with interviews. For specific business process sorting, you can use the company's SOP + business system process to quickly understand.

2. Ability 2: Sorting out indicators

Because business processes are very flexible and changeable, senior data analysts often have to sort out indicators by themselves (rather than memorizing AARRR and the like) to sort out the indicator system. Sorting out the indicator system needs to be combined with specific business processes (i.e. the combat layer in business understanding). Data indicators have two basic forms: total score type and process type (as shown below), but often the business process itself is very complex and has many steps, so you need to be able to sort it out carefully.

3. Capability 3: Design Labels

The ability to design labels is as important as combing indicators, but it is often overlooked. Senior data analysts are not only good at pulling out cities, genders, and ages from dimension tables, but they can also keenly capture words such as "senior users", "long-term operations", and "defensive products" from business mouths, and then try to quantify them with data.

These business-meaning labels, commonly known as "jargon" and "jargon", are an important criterion for measuring a data analyst's understanding of the industry. Of course, being able to design a new label at any time is a reflection of high ability (as shown below):

4. Ability 4: Communication needs

With indicators and labels, you can start to extract data. However, the raw data requirements that have not been processed are often messy, such as:

"Hey! Give me a number, just a regular number, quick!"

“Give me a forecast for next year’s sales. It doesn’t have to be 100% accurate, just 99% is fine.”

"Help me figure out which user won't buy the product, so that he will buy it when I go there."

A qualified data team leader or department manager has the ability to organize these confusing expressions into standardized indicators and dimensions, and can understand the purpose of the business looking at the data. This is a very important ability! Many students are tortured to death because their leaders do not have this ability.

An excellent data department manager can discover opportunities to develop dashboards and models from routine data acquisition needs, thereby gaining more credit from the department. This is a more valuable advanced ability.

5. Ability 5: Analytical Logic

The so-called analytical logic is how to use data to explain the problem. There are two typical logics:

1. Split logic, start from a main indicator, split layer by layer, and find the answer to the problem
2. Hypothesis logic: first propose business assumptions, then use data to verify the hypothetical relationship

Generally, when data analysts look at data, they often use splitting logic to find problem points from large to small. When discussing with the business department, the business may propose analytical hypotheses, which need to be verified in an onion-like manner. Analytical logic is the core ability of data analysis. If the analytical ability is not strong, then you can only write something like: "Activity is low, we need to increase it." If you can only write that you need to increase it, you basically have no chance of getting a high-paying job.

6. Ability 6: Designing Experiments

In addition to using existing data, testing is also a common tool. But be careful! Designing experiments is a test of one's ability. Because the statistical two-sample T test will only tell you whether the sampled numbers are significantly different, but why do you compare these two sets of numbers; why do you compare this indicator; and what business problems are indicated by the significant difference, you have to figure it out yourself.

Often, when doing AB testing, the business side does not consider whether the two user groups and the two versions are comparable, and does not control interference factors, making the test results difficult to interpret. The differences between the previous and subsequent tests are all due to this. We often say that a good experiment must be based on the business. Who is it for? What indicators are measured? What factors are controlled? You have to think clearly about all of these, which is the embodiment of reasoning ability.

7. Capability 7: Project Management

Senior data analysts certainly need to have project management capabilities, because data analysis work itself intersects with various businesses. If the front-end tracking is not done well, the business rules are not clear, and the front-line staff operate blindly, it will affect the data quality, the accuracy of the prediction, the launch of the data dashboard, and the implementation of the model. Therefore, you need to have project coordination and communication skills to promote project implementation. Project management is a common ability for all senior positions.

8. Ability 8: Summary and Report

Reporting ability is also a common skill for senior positions, but it is particularly important for senior data analysts, because they have to face the bosses of various departments directly, and the data rigor, scientificity of the summary, and rigor of the reasoning are all very high. And they often have to adjust the reporting direction at any time according to the intentions of the bosses (follow the trend), which requires very high personal ability.

But on the other hand, if your bosses are willing to take you to meetings of all sizes, your promotion is just around the corner.

The above is an introduction to the eight major capabilities. This is also the reason why we see SQLBoy in the cubicle typing away on his keyboard until smoke comes out, while team leaders, managers, and directors are all busy in meetings because there are too many non-technical tasks to handle.

Author: Down-to-earth Teacher Chen WeChat public account: Down-to-earth Teacher Chen