What is a data analyst? Explains the meaning, necessary aptitude, and the theory of “work to disappear?”

1. What is a data analyst?

What exactly does a data analyst mean?

This time,

Let’s take a closer look at each.

1.1 Definition of data analyst

A data analyst is a job that proposes improvement measures for products / services and management issues while utilizing data and data analysis models . The working styles of data analysts are divided into “consulting type” and “engineer type”. Specifically, there are two types: a “consultant” type that encourages more advanced service operations and various decisions at the site, and an “engineer” type that enhances the performance of existing services and products.

1.1.1 Consulting data analyst

Consultant-type data analysts make hypotheses to solve problems faced by companies, set analysis objectives, select necessary data, mine big data, propose concrete solutions, and provide advice and consulting. It ‘s a job to do.

The main places of employment include consulting firms and marketing companies.
Unlike consultants who propose near the management level, we formulate specific problem-solving plans and business execution policies closer to the site.

1.1.2 Engineer-type data analyst

Based on the results of data mining and machine learning, engineer-type data analysts find certain regularities such as user behavior characteristics, analyze and report analysis results, and aim to improve the quality of services provided. ..

The main places of employment include social game companies and in-house media management companies.
We will consider what the analyzed data and machine learning results represent, whether there is regularity in consumer trends, etc., consider specific improvement measures in product development, and even implement them.

1.2 What the data analyst does

Data analysts can generally be described as “a task that uses statistics and IT skills to analyze a huge amount of data, find meaning from the data, and use it to improve management and products.” Although there are small differences in the work of each step between the consultant type and the engineer type, the work is roughly carried out in the following steps.

1.2.1 Analyze data to discover issues

Analyze big data and discover challenges. Big data is defined as ” data for deriving useful knowledge for business ” in the ” 2012 Information and Communication White Paper ” of the Ministry of Internal Affairs and Communications . The following is an example.

Much of the big data is often collected over the internet, and the data is updated and analyzed in real time. It is necessary to process the enormous amount of accumulated data, discover issues for your company, and make a “hypothesis” when discovering issues. To acquire hypothetical thinking skills

It is important to repeat the above process

1.2.2 Hypothesis making for solving problems

We will make a hypothesis to solve the problem you found. It is important to think about a set of possible “hypotheses” (why the problem is occurring) and “solutions” for the problem .

1.2.3 Hypothesis testing

Test the hypothesis.

For example, if the conversion rate of your application from free membership to paid membership is low, the hypothesis is as follows.

In this way, we will test various hypotheses.

1.2.4 Reporting

Finally, reporting. We will summarize the results of hypothesis testing, collaborate with the field and management, and consider the next move.

1.3 Differences between data analysts and data scientists

Data analysts and data scientists tend to have vague and mixed business divisions and definitions.
The specific difference is

Let’s take a closer look at each.

1.3.1 Data scientists implement algorithms and build models

Data scientists use machine learning to implement algorithms and build models based on the data processed by data analysts.

An algorithm is, in a broad sense, a “procedure or rule for solving some problem”, and is implemented by applying machine learning based on the data processed and molded by a data analyst.

Model construction is performed in 4 steps of data preparation → data preprocessing → model creation → model evaluation, and if problems are found, they are corrected and repeatedly verified until a satisfactory result is obtained.

1.3.2 Data analysts are closer to the field

Data analysts provide consulting, data analysis and processing to solve problems from a position closer to the field . In addition to working as a data analyst, I also work as an artificial intelligence (AI) engineer, including machine learning.

There is no strict line between data analysts and data scientists, so some companies hire data scientists as data analysts.2. Skills and appropriateness required for data analysts

There are four main skills and aptitudes required for data analysts.

Let’s take a closer look at each.

2.1 Statistical skills

As a prerequisite for machine learning and data analysis

Learn statistical skills such as.

If you want to become a data analyst and start data analysis and statistics, let’s perform typical statistical analysis and machine learning.
First of all, move your hand and try it. We recommend that you learn languages ​​such as R and Python, and actually move your hands using books such as “Calculus” and “Linear Algebra (Matrix)” for college students.

2.2 Programming skills

Programming skills are also required to learn data analysis using R, Python, etc.

Data analysts need to learn “statistical analysis” and “time series analysis”. R is strong in statistical analysis, and for time series analysis, the R language, such as the forecast package, has an overwhelmingly rich lineup of packages.

Statistical analysis refers to “analysis of data accumulated based on statistical theory”, and time series analysis refers to “analysis of data on phenomena that fluctuate over time, such as temperature, earthquakes, and stock price fluctuations.”

Many research companies use the R language because it is convenient to understand whether it is statistically significant from the analysis results of the questionnaire data.

Python has the advantage of “prediction” through machine learning. For example, it is strong in predictive models such as house prices and horse racing.

2.3 Hypothesis building ability

Skills for building hypotheses for problem finding and building hypotheses for problem solving are also required . Make a hypothesis before gathering or analyzing information.

Thinking styles and thinking habits that consider the overall picture and conclusions of a problem from the stage where there is little information are called “hypothetical thinking.” With this hypothetical thinking skill, your work will go smoothly and be more accurate.

2.4 Communication skills

Communication skills are also important. Unlike consultants who carry out their duties in a position close to the management team, they often act concretely in a position close to the site .
Therefore, it is important to win the trust of the field, and “humility” and “respect for the opinions of the other party” are also important.

3. How to proceed with the work of data analysts and tips

Next, I will explain the tips for smooth business as a data analyst.

Specifically, the following can be mentioned.

Let’s look at each one.

3.1 Technical skills such as database operation and programming are “premise”

Utilization of big data utilizing R and Python libraries is a prerequisite. By learning how to use Web API and scraping, it is possible to pull huge amounts of data from scraping on various websites, and to incorporate trained models into Web API format into services.
In addition, you need the skill to derive the answer that gives a solid answer to the question you asked by analysis.

Also, API and scraping are important for getting good quality data. It is meaningless if the data itself is a mixture of missing or low quality data, or if the population parameter is small. Collecting the “data” that is the material is very important.

The importance is as follows.

When working as a data analyst, skills in Web API and scraping, utilization of R and Python libraries, and technical skills such as DB operation are prerequisites.

3.2 Thorough hypothetical thinking

Let’s acquire hypothetical thinking thoroughly.

By learning hypothetical thinking, you can improve the quality of decision making. The result is less wasted work, not only getting the work done faster, but also improving the quality of the work.

3.3 Cooperation and communication with on-site staff

A data analyst is in a position to discover issues, make hypotheses, and verify them from a position closer to the site.

Therefore, cooperation and communication with on-site staff is important. The larger the project scale, the more difficult it is for a data analyst to verify the effect alone, so it is important to work in collaboration with on-site staff.

3.4 “Execution speed” and “verification speed” are more important than the accuracy of the hypothesis itself

“Execution speed” and “verification speed” are more important than the accuracy of the hypothesis itself.

Since the 2000s, the world economy has undergone rapid globalization, and the market has also undergone rapid evolution. Especially since 2010, it has been said that the ” era of VUCA ” has arrived in the world economy.

What is VUCA?

A combination of the above acronyms, it is a term that describes the modern “unpredictable” economic environment.

It is difficult to accurately grasp in advance whether the hypothesis is “correct or incorrect”, and even if “it was correct at the time of analysis”, the situation can change from moment to moment.
Therefore, we will execute various hypotheses at high speed and verify the effect.
Then, we repeat the process of stopping ineffective measures and brushing up while leaving effective measures.

4. Will there be no data analysts? Unnecessary work?

With the development of AI (artificial intelligence), it will be possible to predict the future based on the collection, analysis, classification, etc. of huge amounts of data, and there is a possibility that more accurate AI will appear in the future and work will be taken away.

As a result, some people are worried that data analysts will disappear or that it will be an unnecessary job.

4.1 Ambiguous definition

Data analysts have an ambiguous division of roles with data scientists, data engineers, etc. “I thought that hiring a data scientist would solve various problems, but that was not the case,” said the employer. The disagreement between aspirations and the skills of human resources is becoming a problem. Therefore, it is important to clarify the definition in the future.

4.2 Data processing and model building automation may increase

There is also the idea that the work currently being done by data scientists will become unnecessary due to the spread of AI platforms that can apply prediction models using machine learning without specialized skills.

In fact, some services have been deployed on the AI ​​development platform, and the machine learning model is already built in, so users can perform data analysis and prediction simply by uploading data without building.

For example, ” MatrixFlow ” is a cloud-based platform that allows you to build AI without programming. Both deep learning and numerical algorithms are available, and sample data is abundant, so users who “do not have the data but want to move it for the time being” can also use it.

4.3 “How do you want to utilize the data” is important

As data processing and model building automation progress, it becomes difficult to evaluate “data processing ability” itself, such as database operations and simple programming, as a skill set for data analysts.

Therefore, what you want to do with the data will be more important. Data analysis ability itself + alpha skill is required.

For example:

As mentioned above, let’s aim to become a human resource who can provide added value other than data analysts.

5. Estimated salary for data analysts

The following is a guideline for salaries of data analysts.

The average annual income of data analysts is 6.49 million yen, which is higher than the average annual income of Japan.

Looking at the salary distribution of regular employees, the volume zone is 6.7 to 7.85 million yen, and the average annual income belongs to a position lower than the volume zone. The overall salary range is 406 to 11.1 million yen, and as you can see from the above, it is expected that the income will change significantly depending on the place of employment, experience, required skills, etc.

Source: Data Analyst Job Annual Income / Hourly Wage / Salary Information | Job Box Salary Navi (Updated: January 6, 2021)

6. To become a data analyst

To become a data analyst, first clarify your career path as to whether you want to be a “consultant data analyst” or an engineer data analyst.
On top of that, learn the basics of statistics and programming, thoroughly practice hypothesis thinking in various tasks that you are in charge of on a daily basis, and hone your hypothesis building skills.

If you want to become a data analyst from inexperienced, if you are an inexperienced person in the IT industry as well as a data analyst, you should be prepared for the difficulty of changing jobs.

Data analysts are required to have high skills related to data analysis, and also to be human resources who can greatly promote projects in a position close to the site, and high skills are required. From a long-term perspective, there is also the idea of ​​building up as an engineer that even beginners can easily get.

7. Summary

This time, I explained what a data analyst is, what kind of work it is, what skills it requires, and its future potential.
Data analysts are required to have high skills in data analysis, so if you are aiming from inexperienced, the threshold will be high. It is also a good idea to start with an engineer who is easy to get even for beginners.
We hope you read this article to get a better understanding of data analysts.