YAMASHITA adopts Microsoft Fabric as a platform to drive data democratization

Nurturing more DX personnel becomes essential for achieving business goals 

YAMASHITA started as a linen supplier in 1963 in Shimizu City, Shizuoka, expanded into Japan’s Kanto and Kansai regions in 1976, and began renting and selling welfare equipment 10 years later. Its contract retention rate for linen supply is an extraordinary 99.9 percent, and its share of equipment rental and sales sits at the top of the industry. The enterprise is also renowned as the only operator in its industry to provide 365-days-a-year support.

YAMASHITA’s Long-Term Vision 2030 announced in December 2022 articulated a plan for dynamic growth through superior CX (customer experience) derived from EX (employee experience) and a goal of becoming a home care platform provider with holistic, user-centric services. One of its key measures is data democratization through more DX personnel.

“Technologies incorporating AI have appeared and their evolution will only accelerate,” says Kuniharu Ogawa, YAMASHITA’s Head of Digital Transformation. According to Ogawa, fostering more DX personnel capable of fully leveraging AI is pivotal to the enterprise’s growth over the next decade. A data platform to support this increase is also critical.

“Becoming a home care platformer required several services beyond our existing businesses,” says Ogawa. “We have to grasp market response in real time and get deeper customer insights so our services can deliver better customer experiences. This required a seamless data platform transcending the distinct channels of individual service sets. We also needed to democratize data by fostering more citizen data scientists capable of using the platform and its data to enhance business.”

Ogawa set his sights on Microsoft Fabric as the data platform to achieve the goal, adopting it a month before its official release in November 2023. He explains the reasons below.

“I was intrigued by the ability to leverage natural language-based data through the AI capabilities built into Fabric. It synergizes well with Microsoft Power Platform, which we use for low-code and no-code development, and we see it playing a key role in expanding development in the future.”

Ogawa also viewed Fabric as a flagship system for YAMASHITA’s data democratization.

Tackling the most complex and sophisticated areas of data analysis first

In October 2023, YAMASHITA began building a data platform with a preview version of Fabric. The firm also enlisted Microsoft-recommended partner Zeal for its considerable knowledge of Microsoft solutions and extensive build experience. In 2021, Zeal had become the first company in Japan to obtain an Advanced Specialization in analytics using Azure, which was a major factor in the selection.

Besides recommending Zeal, Microsoft proposed Azure Innovate, a program it offers to accelerate cloud and AI adoption and which includes financial support.

“Fabric is a new product, and we were an early adopter, so we didn’t have any examples to calculate the ROI,” says Hirofumi Nakayama, YAMASHITA’s DX Promotion Manager. “But financial support from Innovate lowers the initial hurdles by lowering deployment and build costs. This was a very persuasive aspect for our president.”

YAMASHITA launched a development project in the Home Care Division, which rents and sells welfare equipment, in November 2023. Kazuki Komatsuzaki of the Systems Division explains the reasons for this starting point below.

“Because our Home Care Division deals with nursing care insurance for individuals, we need an accurate grasp of constantly fluctuating policy-holder counts. The core of our system is spreadsheets that track the number of new and additional contracts for rental and sales of welfare equipment. It’s a complex mix of data and the logic for data analytics is extremely advanced. In short, we started with the most challenging area in the company.”

The project team proceeded to organize data until December 2023 and provided the first dashboard to the Sales Department in February 2024. More recent dashboards visualize the number of customers per sales representative for each level of care and provide total sales analysis. Zeal took charge of creating many of these dashboards and preparing data, but YAMASHITA made the final dashboard for total sales analysis.

No more data creation work, fresh data ready for decision making every morning

“In the past, a team of four people, including myself, would spend the morning creating data for the Sales Department using Excel spreadsheets,” recalls Natsumi Maruki of the Home Care Business Center. For example, the spreadsheet on new contracts for equipment rental and sales was initially visualized on a dashboard and required an hour of work every day for daily reports. Inevitably, with limited time to spend on data creation, only the most important data was daily and the rest dropped to weekly or monthly reporting.

“Moving from Excel to Fabric eliminated the need to create data spreadsheets in the old way,” says Maruki. “Now, when sales personnel arrive in the morning, they have a dashboard with the most up-to-date data in the Fabric workspace. A sales manager told me he can quickly make decisions by looking at the data every morning. Data we used to provide weekly or monthly is now always available and up to date, which greatly reduces the work required to provide every data set.”

Systems Division workloads are also lighter because it no longer has to prepare the data from which the Business Center created Excel spreadsheets. According to Komatsuzaki, “We used to work an extra hour of overtime on weekdays to extract data for daily reports. For random requests, we had to process data for another one or two days, which made timely provision strenuous.” The new system has eliminated this burden and all requests now receive an immediate response.

In the past, each sales location would process data based on Excel spreadsheets created by the Business Center, but the dashboard has eliminated this task as managers can easily analyze each sales real situations.

“Most data processing at sales offices was done by office managers,” says Maruki. “Eliminating this work has made it easier for them to focus on strategic planning, which is their main job.”

Linen Supply Division follows Home Care Division’s success

In June 2024, YAMASHITA’s Linen Supply Division began developing Fabric-based dashboards. “We started a project to identify the cost and revenue of each customer using standardized costs in April 2024,” explains Accounting Division section manager Ryo Tanimura, who is overseeing dashboard development for the Linen Supply Division.

“Our linen supply business is B2B. Although we don’t have to grasp fluctuations in the number of users on a daily basis, we need to understand and visualize the differences in costs and profits on sales for each business partner. We were visualizing data with Excel and Microsoft Access, but the Home Care Division was already enjoying positive results with Fabric, so we adopted it too.”

In August 2024, the division launched a dashboard showing revenue per customer. Zeal aggregated data into Fabric, but Tanimura created the dashboard himself.

“Although it was my first time using Fabric, making dashboards was effortless,” Tanimura explains. “The intuitive GUI made it easy to create screens tailored to specific user needs. Once the data to be used was organized as per the data mart, I could create a screen in about 30 minutes between my main tasks. Excluding design elements, it would have been even quicker.”

As of November 2024, the necessary dashboards have only just been completed, and according to Tanimura, “The results are primed to appear.” However, sales revelations are already emerging.

Fabric praised as the best data platform for short-term and medium-to-long term business applications

Komatsuzaki details YAMASHITA’s Fabric initiatives below.

“Our objective for using Fabric was to give business units the power to develop the dashboards that field workers’ needs. The System Department is only responsible for providing data models. We only provide structured data at present, but we would also like to prepare unstructured data to make better use of Fabric's data lake.”

Currently, over a dozen YAMASHITA employees, mainly from the DX Promotion Department and the Data Analytics Team, can create dashboards. “We want to increase this number in the future and have at least one citizen data scientist in all business locations and sales offices by March 2025,” says Ogawa.

“At the moment, we mainly analyze sales KPI data, but the great endorsements we’re receiving internally drive us to further expand the scope of application and consolidate unstructured data,” Ogawa continues. “We also intend to promote the use of generative AI by leveraging data stored in Fabric. In our view, Fabric is the best data platform for both short-term and medium-to-long term business applications.”