How can we transform into a data-driven company that makes business decisions based on data? Many companies are still not fully implementing data-driven management. When data utilization is not progressing, the reason is often due to communication issues between management and data professionals.

In October 2021, he will manage a private equity fund.D Capital Co., Ltd.Megumi Matsutani, who joined the company, has consistently worked on data science, which uses huge amounts of data to solve business problems in areas such as improving aircraft safety, trading financial derivatives, and fashion e-commerce. We asked Matsutani about the challenges that company executives must face when putting data science into practice at Japanese companies, and the efforts that are necessary to solve them.

Partner, D Capital Co., Ltd.
Megumi Matsutani

Graduated from the Department of Aeronautics and Astronautics, Faculty of Engineering, University of Tokyo. Completed doctoral studies at the Department of Aeronautics and Astronautics, Massachusetts Institute of Technology. After studying aircraft control theory at NASA's Langley Research Center and other institutions, he worked as a quantitative strategist at a US investment bank (Tokyo and New York) where he engaged in research and development into trading automation. After that, he promoted AI research and development and data strategy as CSO at a fashion e-commerce platform operator. The "Open Source Development and Implementation Platform for Social Decision-Making Algorithms," developed in collaboration with Yale University, won the Prime Minister's Award at the Japan Open Innovation Awards sponsored by the Cabinet Office. In October 2021, he joined D Capital Co., Ltd. as a partner.

The penetration of data science requires technological literacy among managers

── Data science has been attracting attention for some time now, but what do you think about the current state of data science in Japanese companies?

Data science is a powerful tool that can be used in a wide range of areas, including management, finance, human resources, marketing, and sales.In Japan, the adoption of AI is relatively advanced in sectors that already have large amounts of accessible data, such as advertising and e-commerce, and in sectors where technology is at the core, such as manufacturing. However, I feel that compared to North American companies, the adoption is still far from being widespread.

One of the reasons why data science has not taken root in Japan isThe technological literacy of business managers is still insufficientExamples include:.In North America, it is becoming common for students to study computer science as a liberal arts subject, just as they study history and culture.

However, in Japan, the division of science and humanities courses during university entrance exams may also have an impact. For example, once students realize that they are in the humanities, they tend to refrain from working on their knowledge of science subjects. In many cases,Communities and ways of thinking themselves are still divided, such as between science and humanities, or business people and engineers.Communication itself has been cut off. This situation has led to extremely few opportunities for the exchange of knowledge between business people, engineers, and academia.

In addition, business competition in North America is much more intense, and apart from some old-fashioned managers,"If you don't use data, you're not going anywhere"This is why individuals and companies alike are desperately trying to catch up, regardless of their original business domain, but I think that in Japan, which is geographically independent, this is not yet close to reality.

Managers need to communicate more with data scientists

── What should managers do to overcome this situation?

What is most needed isDon’t label data scientistsDon't assume that the other person won't understand after listening to you, but communicate directly. This will lead to new discoveries for both of you and an understanding that will allow you to tackle business issues together.

I think there are specialists in various fields within the company, and data scientists are one of those specialists. When you hear the word "science," you tend to get an academic impression, butData science is about using data to solve business problems.The attitude towards tackling a problem is science, but calling the methodology academic is slightly different from the original meaning.

Similarly,A data scientist is not an academic, but rather a businessperson who uses data analysis techniques to tackle business problems.Therefore, first of all, it is essential to communicate well with data scientists, just as you would with other specialists in the company. Setting the right problem is based on communication between management and data scientists. Without setting the right problem, there is no point in actually running any analysis.

In addition to sufficient communication,It is also effective for managers themselves to learn the basic concepts of data science and try it out for themselves..

There is a misconception here, but data science is not a difficult field to get into. Just as people who used to create documents on paper now do so naturally on computers,Data science is actually easy to understand, and the way of thinking in data science becomes natural through sufficient communication with the field.I would like you to think of it that way.

Nowadays, various courses are available online, and you can also learn by attending classes. There are also many tools with friendly UIs. As you can see, there are many options for learning the basic concepts as an entry point, and there are also a variety of books and specialized books on data science depending on the basic knowledge you need.

Just as there is a need to change management awareness, there are things data scientists can do to improve communication.

That is, Don't limit yourself to just model development and data analysis.These are just a small part of our business."Deciding the business problem you want to solve with data is itself an important part of a data scientist's job."You should have that mindset. It is no exaggeration to say that more than half of a data scientist's work is problem setting. Data scientists should be fully aware of this, and if they are not, they should question whether the problem they are working on is appropriate.

The transformation into a data-driven company is up to the management

── In order for managers to utilize data science in their own decision-making, is there anything else they should do besides deepening their communication with data scientists?

Creating an organization that allows sufficient direct communication between data scientists and managementCreating a data-driven culture requires organizational change and hiring the right talent.

First, the organization. As a result of the data analysis,An organization where the work of data scientists can be reflected in management decision-makingSpecifically, is the data science department located just below or at an equal level to the management team? If the organization is not like this, it will be impossible to tackle data-driven management. It will be difficult to hire data scientists who can bring about change, and even if they are able to do so, they will not be able to demonstrate their capabilities.

In addition, Are there people in management with a technology background?This is a major point. This is important for data scientists in the field. Having people in management who have technical knowledge and understand the importance of using data will make the impact of data science on business even more certain.

Furthermore, many Japanese companies still employ people through a membership-based system, where the emphasis is on which company they join, rather than on the type of job they do. Specialists such as data scientists are employed on a job-based basis, and there are few organizations where people can work under an appropriate evaluation system.It's difficult to chart a career path as a specialistThis also contributes to the low productivity and lack of human resources compared to North American companies.

── Where would be a good place to start when it comes to change?

Just having ideals won't get you started,Identify on-site operations that are likely to be effective and easy to implement, and streamline or improve those areas.If you don't have a good grasp of the current situation with the necessary granularity, start by obtaining the data. If you can make improvements based on the data and make them happen,The benefits of data can be shared between management and the field, fostering an atmosphere where people can tackle even bigger challenges.I think.

Changing corporate culture is not something that can be achieved quickly. However, to be reborn as a data-driven company, it is necessary to make a thorough effort. This thoroughness is a task that should be taken on at the leadership level, not just by individual management.

Data-driven efficiency will strengthen Japanese companies

── What kind of future does await Japanese companies that have incorporated data science as a matter of course?

For companies, it is no longerData science is not a nice-to-have, it's a necessityI think so. If we stand still while other companies are moving forward with their initiatives, it is clear that we will be left behind.

Managers should already be looking at some kind of data every day and reflecting it in their decision-making. The idea is to add data that gives a deeper understanding of on-site operations to the information used in those decisions. If there isn't enough detailed data to answer the questions you want to know, obtain it and analyze it. Such efforts will enable data-driven decision-making.

Japanese companies offer high-quality, stylish services and value. If they combine this with data-based efficiency, they can increase their competitiveness, and I would like to see this happen.