Nishimatsu Construction Co., Ltd.

Masuda

I come from an accounting background, and my involvement with company-wide data began when I transferred to the Corporate Planning Department in 2013. Management frequently requested analysis of various operational data, but the data required for analysis was not readily available and was scattered across various departments. It took a significant amount of time to request and collect data from each department, and I realized that this was a critical issue that hindered the speed of management decision-making.

Consequently, I proposed the creation of a specialized department to examine ICT strategy and determine how to organize data and systems company-wide. As a result, I was transferred to lead this initiative myself. Since then, every time I have proposed initiatives such as data utilization or DX promotion, I have ended up taking charge of them myself (laughs).

Masuda

Data management methods varied by department, and even when we collected data, the granularity and accuracy were inconsistent, leaving it in a state that was far from ready for analysis. Therefore, the preparatory work, such as data cleansing prior to analysis, was an extremely laborious task.

For example, when trying to aggregate data for a specific project using a single axis, the coding systems differed depending on the department or system. Even in cases where a client and a supplier were the same company, the data was not linked, making it impossible to see the relationship. Conducting effective data analysis was extremely difficult.

Masuda: First, regarding data, I realized that there are different stages to data utilization, such as what kind of data is needed, in what situations, and at what speed. Therefore, I presented these stages along with our company's current status, and finally, I outlined our future direction and roadmap. When explaining this to the management team, I first gave a presentation to all members of the President's Office, including my then-supervisor, the Head of the President's Office, stating that DX requires the development of three key areas: data, human resources, and organizational structure.

Regarding the importance of data in particular, I started from the fundamentals, such as what data and information are, and how the information required differs depending on one's position, such as information necessary for management versus information necessary for daily operations, to advocate for the necessity of data. As a result, the Head of the President's Office showed understanding and supported us in prioritizing infrastructure development when we presented it to the President.

To prepare for implementation, we first conducted interviews with each department regarding their data-related challenges. We began by identifying where data was located within the company (As-Is), considered a model for data retention (To-Be) (see figure), and proceeded to design the architecture.

■ Model diagram of the data utilization infrastructure design at Nishimatsu Construction Co., Ltd. The top shows the current state (As-Is), and the bottom shows the target state (To-Be)

For a common data utilization infrastructure, it was necessary to organize client, employee, and construction project data as master data. Among these, for client data, it was necessary to connect the client data held by each system using a unified code system and perform data cleansing and matching to ensure accurate and reliable data. That is why we decided to consider the introduction of uSonar (LBC code) (see figure).

■ The client master for the data utilization infrastructure was cleansed and matched using uSonar's LBC code

Masuda

To begin with, our philosophy is that data infrastructure development is an upfront investment. Since the return on investment depends on how the established infrastructure is utilized, I believed from the start that it would be difficult to explain and justify the data infrastructure project to management based on cost-effectiveness alone. I persuaded them by explaining that by integrating with uSonar data, we could establish a 'Golden Record' (the true master data) for our client information, which would then enable us to advance our data utilization.

Masuda

Building the data utilization infrastructure took about 10 months, but the uSonar implementation portion was completed in just over two months, excluding the data cleansing process. I believe it was very smooth because you checked the requirements definition in advance and responded flexibly to our progress.

Masuda

Although we have not measured the quantitative effects, we are now able to retrieve data smoothly whenever necessary. Furthermore, we feel that value beyond mere operational efficiency is being created, such as the ability to visualize and analyze data from new perspectives. Specifically, it has become easier to perform data analysis and extraction using diverse aggregation axes, such as by individual client or client group, which is proving useful for making proposals, decisions, and establishing management policies. I consider this a necessary expense for maintaining, updating, and ensuring the accuracy of our corporate data.

Masuda

Interviews revealed that when employees wanted to perform analysis, they did not know where the data was located. I was not the only one struggling; in fact, everyone was facing the same issue. With the introduction of the data utilization platform, we are now able to provide data, which has shifted the mindset to "If that is the case, let's use the data." I feel that because we are now able to collect and analyze internal data more comprehensively than before, the accuracy of our discussions has improved.

However, regarding our client master data, while it has reached a level where we can match, consolidate, and utilize it, I feel there is a need for further refinement to make it immediately actionable for business operations. To elevate the data to a professional, operational level, we must not simply match and cleanse it, but rather assign LBCs at the time of client master registration to organize the information. We have already begun this process in some areas, and once fully implemented, I believe the data will be usable within our business systems.

Our data utilization platform has only just begun, but we have received inquiries from various companies, including other general contractors, asking, "We want to work on MDM, so could you teach us how?" This makes me feel that our initiatives are somewhat advanced. Moving forward, I intend to continue taking on challenges in collaboration with partners, including uSonar, to ensure the success of our data utilization and, ultimately, our DX.

Masuda

Data is essential for DX. Some may be struggling to gain management approval due to debates over initial ROI. In an era where data is so critical, it is necessary to view it as a necessary expense and a strategic investment from the perspective of protecting the company's valuable assets. More importantly, it is vital to use the organized data to promote DX and generate results. We are still in the early stages ourselves. Let's work hard together.