What is the average annual salary of a data scientist? Explains the future you care about

Of course, programming skills are also important, but it is clear that a wide range of skills are required, from setting tasks in the first place, selecting appropriate analysis methods, and operating them.

1.2 58% of companies struggle to secure data scientists

Data scientists who possess the above three skills, especially those that can be applied to business, are rare, and many companies are struggling to hire them.

In fact, in a questionnaire survey of domestic companies in 2019, among the companies that planned to secure data scientists in the past year, 58 companies in total answered that they could not secure the target number of people. %There was also. Furthermore, by industry, “information and communication industry” is low at 34%, but “manufacturing industry” is high at 68%.

Data scientists do not yet have official qualifications, but they require knowledge of statistics, artificial intelligence (AI), machine learning, IT, programming and much more. From these figures, we can see that there is not enough human resources with sufficient skills to meet the needs of corporate data scientists .

2. What is the average annual income of data scientists?

What is the actual average annual income of data scientists who are required to have such knowledge / skills in various fields?

In conclusion, the example of presenting a high compensation level when hiring digital human resources with high skills related to overseas AI and data science is boosting, and the annual income of Japanese data scientists is also increasing.

2.1 The average annual income of data scientists is about 7.69 million yen

According to a survey conducted by the Data Scientists Association in 2019, the average annual income of data scientists is 7.69 million yen. Around 2016, the ratio of young human resources was relatively high, and the average annual income tended to be kept at 7.29 million yen.

The increase in annual income in recent years

Seems to be influential.

In some cases, they are offering an annual income of 10 million yen or more. For example, DENA says that an annual income of 10 million yen is appropriate for new graduates and AI human resources, and Fujitsu and NTT DoCoMo offer an annual income of around 30 million yen as professional human resources with high expertise and experience.

2.1.1 Percentage of older data scientists

According to the 2019 survey results, 28% of the current age group of data scientists are in their 50s or older. Although the core group is in their 30s and 40s, the percentage of people in their 50s and above has been on the rise for the third consecutive year since 2017.

2.1.2 [Reference] What is the average annual salary of office workers?

According to the National Tax Agency’s ” Statistical Survey of Private Salaries for Two Years of Reiwa, ” the average salary of office workers was 4.33 million yen.

Since the annual income of data scientists is about 7.69 million yen, it can be said that it is a relatively high level compared to general office workers. For those who want to earn a high income, it may be attractive to be a data scientist.

3. Data scientists are very popular among students at the University of Tokyo and Kyoto University

According to the employment rankings of 2693 people (including graduate schools) who are planning to graduate or complete Reiwa 3rd year at the University of Tokyo and Kyoto University, the consulting companies are monopolized from 1st to 7th.

There is no description about the ranking by job type, but it is clear that the popularity of companies called general / IT consulting is increasing.
For example, Accenture Co., Ltd., which ranks first, lists data scientists as one of the recruiting positions, so it is inferred that there are many applicants for data scientists among the students of the University of Tokyo and Kyoto University who aspire to be consulting companies. Will be done.

4. Actual conditions of working styles of data scientists

How do data scientists actually work?
We investigated the type of business of the company you are enrolled in, the type of data scientist you are looking for, and what kind of infrastructure environment you work in.

4.1 Actual status of enrollment of data scientists by industry

Although it is a high-demand data scientist, the total number of companies with one or more employees is very small at 29%.

By industry, data scientists engaged in the IT telecommunications industry accounted for the largest percentage at 59%, while data scientists belonging to the construction industry and the transportation / postal industry accounted for 11% and 14%, respectively. Was there.

From this result, we can see that the situations in which data is handled differ depending on the type of industry.

4.2 “Types of data scientists” required by companies

Data scientists have the following three types of human resources.

The most in demand is the “marketer type”, which accounts for 40% of the total demand, followed by the “engineer type” at 36% and the “analyst type” at 24%. Although corporate needs differ depending on the type of industry, it can be seen that many marketer-type data scientists with high business power are currently required.

Reference: Customer contact data (2) What is analysis to deepen customer understanding? | ADK Marketing Solutions

4.3 Actual state of analytical infrastructure in enterprises

According to a 2019 survey, 44% of data scientists were “dissatisfied with the equipment of analysis / analysis infrastructure (database / data warehouse, etc.)” and “dissatisfied with the equipment of analysis / analysis tools (statistical analysis software, BI tools, etc.)” There is a high percentage of data scientists at 38%.

Reference: Data of Data Scientest Series vol.9 “44% -Increasing Dissatisfaction with Analytical Infrastructure”

While the need for AI is increasing and management’s understanding of data utilization is increasing, the issue that the environment for data scientists to analyze in the field is not yet ready is highlighted.

5. Future potential of data scientists

A data scientist who is receiving so much attention. Is it a profession with high expectations for the future?
Let’s take a closer look at the survey results of students and people engaged in data science.

5.1 What do data scientists think about the future of their jobs?

According to a survey conducted by the Data Scientists Association in 2020, 81% of the respondents answered that they felt the future / rather the future.

Looking at each age group, nearly half of teens and 20s answered that they feel the future.
The profession of data scientist seems to be a promising and longing profession.

Source: Data of Data Scientist series vol.14 “81% -Percentage of people who feel the future of their work as a data scientist”

5.2 Universities are also serious about training data scientists

The University of Tokyo and Kyoto University are not the only ones focusing on training data scientists. Many universities are embarking on the establishment of faculties because even new graduates are expected to earn a high annual income of 10 million yen or more.

The national and public universities that founded the “Faculty of Data Science” are Shiga University in 2017 and Yokohama City University in 2018. Although there is no faculty at Tokyo Institute of Technology, we started an initiative to educate all graduate students in data science and AI in April 2020.

As a private university, Musashino University also established the Faculty of Data Science, and the number of applicants has increased significantly, so it is likely that more and more universities will focus on specialized education in data science.

6. Programming skills required of data scientists

It seems that the programming languages ​​required of data scientists are often “R language” and “Python”. In addition to programming languages, it would be useful to have knowledge of machine learning and statistics as necessary skills.

6.1 R language

R is a free software programming language specializing in statistical analysis.
It is possible to draw complicated graphs after analysis, and until now it has often been used mainly for forecasting in fields such as finance and aerospace. Since it can also be used for machine learning, in recent years it has also been used for business applications such as staffing of organizations and evaluation of marketing measures.

On the other hand, compared to Python, R language is specialized in data analysis, so it seems to be a drawback that it is not possible to create websites.

6.2 Python

Python is a programming language that allows you to do various things, such as:

It is characterized by its simplicity and versatility, and is also used for large-scale Web service development. It’s a famous story that YouTube is made of Python.

However, Python is not all-purpose, and its disadvantage is that its execution speed is slower than that of C language, so care must be taken when processing large-scale data with a high computational load.

6.3 What skills should I acquire other than programming?

In addition to programming, skills that should be acquired include problem definition, hypothesis design, data collection, and the ability to correctly select a model suitable for analysis.

For example, it is difficult to make use of the R language in business based on the results of programming, and statistical knowledge is required to unravel the data. Machine learning solutions such as AutoML have also evolved, and programming is expected to be automated in the future.

Instead of sticking to programming alone, it is good to acquire “the ability to prepare the first stage for automation”.

7. Trends in freelance jobs related to data scientists

According to the recruitment of professional engineers, many data scientists have a monthly income of 600,000 to 800,000 yen, and there are also projects with a maximum monthly income of 1.5 million yen and an annual income of 18 million yen. Even if you are a fledgling person, your monthly income is 300,000-400,000 yen, which is attractive. Let’s take a look at some of the actual projects.

7.1 R language system

There is no description about the ranking by job type, but it is clear that the popularity of companies called general / IT consulting is increasing.
For example, Accenture Co., Ltd., which ranks first, lists data scientists as one of the recruiting positions, so it is inferred that there are many applicants for data scientists among the students of the University of Tokyo and Kyoto University who aspire to be consulting companies. Will be done.

4. Actual conditions of working styles of data scientists

How do data scientists actually work?
We investigated the type of business of the company you are enrolled in, the type of data scientist you are looking for, and what kind of infrastructure environment you work in.

4.1 Actual status of enrollment of data scientists by industry

Although it is a high-demand data scientist, the total number of companies with one or more employees is very small at 29%.

By industry, data scientists engaged in the IT telecommunications industry accounted for the largest percentage at 59%, while data scientists belonging to the construction industry and the transportation / postal industry accounted for 11% and 14%, respectively. Was there.

From this result, we can see that the situations in which data is handled differ depending on the type of industry.

4.2 “Types of data scientists” required by companies

Data scientists have the following three types of human resources.

The most in demand is the “marketer type”, which accounts for 40% of the total demand, followed by the “engineer type” at 36% and the “analyst type” at 24%. Although corporate needs differ depending on the type of industry, it can be seen that many marketer-type data scientists with high business power are currently required.

Reference: Customer contact data (2) What is analysis to deepen customer understanding? | ADK Marketing Solutions

4.3 Actual state of analytical infrastructure in enterprises

According to a 2019 survey, 44% of data scientists were “dissatisfied with the equipment of analysis / analysis infrastructure (database / data warehouse, etc.)” and “dissatisfied with the equipment of analysis / analysis tools (statistical analysis software, BI tools, etc.)” There is a high percentage of data scientists at 38%.

Reference: Data of Data Scientest Series vol.9 “44% -Increasing Dissatisfaction with Analytical Infrastructure”

While the need for AI is increasing and management’s understanding of data utilization is increasing, the issue that the environment for data scientists to analyze in the field is not yet ready is highlighted.

5. Future potential of data scientists

A data scientist who is receiving so much attention. Is it a profession with high expectations for the future?
Let’s take a closer look at the survey results of students and people engaged in data science.

5.1 What do data scientists think about the future of their jobs?

According to a survey conducted by the Data Scientists Association in 2020, 81% of the respondents answered that they felt the future / rather the future.

Looking at each age group, nearly half of teens and 20s answered that they feel the future.
The profession of data scientist seems to be a promising and longing profession.

Source: Data of Data Scientest series vol.14 “81% -Percentage of people who feel the future of their work as a data scientist”

5.2 Universities are also serious about training data scientists

The University of Tokyo and Kyoto University are not the only ones focusing on training data scientists. Many universities are embarking on the establishment of faculties because even new graduates are expected to earn a high annual income of 10 million yen or more.

The national and public universities that founded the “Faculty of Data Science” are Shiga University in 2017 and Yokohama City University in 2018. Although there is no faculty at Tokyo Institute of Technology, we started an initiative to educate all graduate students in data science and AI in April 2020.

As a private university, Musashino University also established the Faculty of Data Science, and the number of applicants has increased significantly, so it is likely that more and more universities will focus on specialized education in data science.

6. Programming skills required of data scientists

It seems that the programming languages ​​required of data scientists are often “R language” and “Python”. In addition to programming languages, it would be useful to have knowledge of machine learning and statistics as necessary skills.

6.1 R language

R is a free software programming language specializing in statistical analysis.
It is possible to draw complicated graphs after analysis, and until now it has often been used mainly for forecasting in fields such as finance and aerospace. Since it can also be used for machine learning, in recent years it has also been used for business applications such as staffing of organizations and evaluation of marketing measures.

On the other hand, compared to Python, R language is specialized in data analysis, so it seems to be a drawback that it is not possible to create websites.

6.2 Python

Python is a programming language that allows you to do various things, such as:

It is characterized by its simplicity and versatility, and is also used for large-scale Web service development. It’s a famous story that YouTube is made of Python.

However, Python is not all-purpose, and its disadvantage is that its execution speed is slower than that of C language, so care must be taken when processing large-scale data with a high computational load.

6.3 What skills should I acquire other than programming?

In addition to programming, skills that should be acquired include problem definition, hypothesis design, data collection, and the ability to correctly select a model suitable for analysis.

For example, it is difficult to make use of the R language in business based on the results of programming, and statistical knowledge is required to unravel the data. Machine learning solutions such as AutoML have also evolved, and programming is expected to be automated in the future.

Instead of sticking to programming alone, it is good to acquire “the ability to prepare the first stage for automation”.

7. Trends in freelance jobs related to data scientists

According to the recruitment of professional engineers, many data scientists have a monthly income of 600,000 to 800,000 yen, and there are also projects with a maximum monthly income of 1.5 million yen and an annual income of 18 million yen. Even if you are a fledgling person, your monthly income is 300,000-400,000 yen, which is attractive. Let’s take a look at some of the actual projects.

7.1 R language system

▸ Project information: Fashion app data scientist ★ 3 days a week ~

This is a job as a data scientist for fashion apps that automatically proposes outfits from the clothes you have. Specifically, it will be the proposal and implementation of CVR and retention rate improvement measures, and the reference implementation and design for incorporating the improvement measures into the product.

Required skills include data analysis in services for toC, experience in improving KPI, skill in collecting and aggregating tens of millions of actual data using SQL, statistical analysis, data mining, etc. Expertise / research achievements, analysis experience in either Python / Go / R / Java.

It is a job unique to the fashion industry that uses AI to propose clothes that suit the person and create new trends. Working 5 days a week from 10:00 to 19:00, monthly income starts from 400,000 yen, and remote control is possible depending on the skill.

I am a data scientist at a company that develops TV commercial effect measurement and marketing effect analysis platforms. Specifically, it is the analysis and development work of analytics and the development of the data infrastructure part.

The required skills are to be able to extract data from normalized data using SQL, and to have experience in data analysis in the marketing context using BI tools and Python / R, or in database design / operation experience / data analysis infrastructure development. Understanding and experience of ETL processing.

If you have experience in building or analyzing large-scale data infrastructure and web development using script development languages ​​such as Ruby and Python, the possibility of adoption will increase further. From 10:00 to 17:00, the core time is a flextime system, and overtime is 20 hours a month, working 5 days a week, and monthly income ranges from 800,000 yen to 1 million yen. Remote work is also possible at Korona-ka.

Professional engineers have introduced many other data scientist-related jobs, so if you are interested, please check them out below.

▸ Click here for professional engineer job listings

8. Summary

This time, we have summarized the work contents, skill sets, and annual income of data scientists.

Data scientists are promising professions with increasing demand and potential for career advancement.

If you are interested in machine learning, statistics, problem definition, hypothesis design, or data scientists, why not try programming with R language or Python?