"We installed the tools, but the digital transformation isn't lasting."
Many people probably face these kinds of challenges. DX projects that started up suddenly stall or fail to continue. In fact, Macnica has been grappling with the same problem for the past three years, constantly experimenting and refining our approach.
At Macnica, approximately 20 departments participated in about a year, more than 70 ideas were gathered, and 11 systems and services were released. The catalyst for this change was not the introduction of new tools, but rather the creation of a system centered around the CoE (Center of Excellence = a company-wide driving organization).
This article traces the process of creating such a system, starting from the failures along the way. We hope it will provide some hints for those struggling with DX promotion as they consider their next steps.
This article is a reconstruction based on a presentation given by Mr. Tone of Macnica 's Innovation Strategy Division.
Three years of trial and error. Once the project ends, people and knowledge return to their respective departments.
"From QCD to DCQ"
The starting point was a sense of crisis. The market environment, including generative AI, is changing very rapidly. If we don't continue to innovate, we will lose to the competition and be eliminated from the market. That sense of crisis was present within the company.
The very concept of system development needs to be re-evaluated. That's where Macnica 's CIO, Ando, came up with the phrase, "From QCD to DCQ." This represents a shift from a development approach that prioritizes quality above all else to one that starts with speed.
In waterfall development, the market changes by the time requirements are finalized and development is complete, and by the time of release, business models, operations, and customer needs have changed. That's why what's needed is to release with the minimum requirements first and then continuously develop to keep up with the changes.
That said, we're not sacrificing quality. Security measures and other quality-related aspects are standardized as company-wide components and distributed to each project to ensure quality. We'll explain this system in more detail later.
Proliferation of unofficial projects
Macnica began its initiative to change not only processes but also culture and mindset three years ago. The first step was "promoting understanding," that is, letting people experience the value of digital technology firsthand. So, they distributed translation tools and other resources to sales staff who had frequent contact with overseas partners, starting with an experience of how their work would change.
Next, we gathered representatives from approximately 20 departments. We launched system development projects based on the real-world business challenges of each department. Up to this point, everything was going smoothly.
However, once the project ends, the members return to their respective departments. As a result, the entire process of idea generation → development → business process improvement comes to a halt. The reproducibility of the initiative that had just started to run smoothly is lost.
On the other hand, under the management vision for 2030, each department was trying to advance DX independently. Before we knew it, we had a proliferation of rogue projects, with each department using different tools and development processes. This made it impossible to expect rapid transformation that would leverage the strengths of the entire company.
*The "Digital Execution Factory" shown in the diagram above is a DX support service for the manufacturing industry that includes support for building a CoE organization, in order to generate continuous DX.
To break free from this chaotic situation, Macnica chose to create a Center of Excellence (CoE) organizational structure.
The Center of Excellence (CoE) was initially established as a virtual organization without a dedicated department, with members contributed from the IT department and the Innovation Strategy Business Unit.
The goal was to establish a system and functions that would promote DX in a reproducible manner. To that end, we first integrated all DX projects from all departments into a single point of contact and put control over them. Rogue projects and rogue applications must be prevented. However, control is not the goal itself, but rather a guardrail to prevent inconsistencies and maintain reproducibility.
If control is the defensive strategy, then improving productivity is the offensive strategy. At the time, Macnica, to varying degrees across departments, often outsourced development to external consulting firms and system integrators.
We brought as much of that process in-house as possible. Furthermore, the use of low-code and generative AI technologies increased the productivity of the development process itself. As a result, development man-hours decreased dramatically, and development costs also fell.
Furthermore, by consolidating the points of contact, the acquired know-how can be accumulated in one place. Even if one project fails, it can be prevented from being repeated in another project. This creates a situation where such mistakes are avoided.
A system for developing DX talent
The Center of Excellence (CoE) is not an organization that handles development on behalf of others. The people from the business divisions that bring in the ideas are the ones who actually drive the project forward, acting as product owners (those responsible for what is created) and developers.
The Center of Excellence (CoE) also provides training mechanisms such as onboarding and on-the-job training (OJT) for these members. This is so that when they complete a project and return to their department, they will be able to take on leadership roles and lead their department.
Through real-world project experience, we cultivate individuals who can excel as DX (Digital Transformation) professionals. As our people develop, the company-wide DX process will accelerate.
One-fifth of the solution is tools. The "5Ps" methodology for building a robust organizational structure.
The foundation of Macnica 's Center of Excellence (CoE) organization lies in the 5Ps concept. Originally, these were best practices for digital transformation (DX) systematized by Siemens and Mendix based on their vast practical knowledge. Macnica then optimized these practices through its own implementation, adapting them to the culture and business practices of Japanese companies.
"Tools are only a small part of DX transformation." Of the five Ps, tools like Mendix handle the Platform aspect. However, transformation won't progress with tools alone. People, Process, Portfolio, and Promotion—the remaining four must all be in place for things to really start moving.
Each P
"People" refers to individuals. We cultivate talent capable of planning and developing agile business improvements and new services through collaboration between business and IT. Furthermore, we define appropriate team structures to increase project success rates.
The process is the way things are done. Before development, we quantify the value of the idea using ROI, prioritize it, and select the optimal tools. During development, we proceed using an agile methodology that integrates business and IT. Finally, we establish a standard process that includes maintenance after development.
Portfolio is a theme. We generate and collect ideas that can lead to internal business improvements and new services through workshops and other means. We aim to enable proactive proposals not only from business divisions but also from the IT division.
A Platform is the foundation. It combines various technologies such as low-code, AI, and IoT to create a common development and operational platform that supports transformation.
Promotion is about communication. It's an awareness-raising activity that maintains team motivation and promotes cultural change.
Furthermore, there are certain types of "P" (Property) that tend to be lacking depending on the type of DX promoter, and these often become obstacles to DX implementation. For more details, please see this article.
The most important thing is "Promotion".
So, which of the five is the most important? When Siemens representatives and Macnica members discussed this framework, the answer they came up with was Promotion.
These kinds of projects can quickly fizzle out and even be forced to disband if the members' motivation drops. Those responsible for DX (Digital Transformation) have other main duties and are pushing these projects forward while being extremely busy. On top of that, they may even receive inquiries from their own departments asking what they're doing.
Nevertheless, if the company recognizes that this activity will lead to profits, the individual will be rewarded and encouraged to move forward. At Macnica, we continue to refine this promotion process by involving management in communication and making results visible.
Transforming the 5 Ps into CoE functions
Macnica has incorporated these 5Ps into the structure and functions of its Center of Excellence (CoE). For example, in the Platform category, they select recommended tools. Depending on the nature of the project, Macnica 's CoE uses different tools such as Power Platform, Mendix, UiPath, and code generation AI (Clude Code).
Of these components, Mendix is responsible for the application layer of the Platform.
So why is Mendix handling the application layer? There's also the option of entrusting development to the much-talked-about code generation AI. However, Macnica believes there are still hurdles to overcome when it comes to building enterprise-level systems while ensuring company-wide governance and security.
The reasons behind this choice are explained in detail in the following article.
These organizational functions responsible for promoting and controlling the use of generative AI are also called "AI CoEs" (Center of Excellence). At Macnica, the CoE for company-wide digital transformation (DX) also incorporates this function.
Once the tools are selected, the next step is distribution. Templates and security modules that conform to the company's design standards are distributed to all projects. This system prevents individual projects from using their own optimal tools and methods. The quality mentioned at the beginning is also guaranteed by these common components.
Did the system function effectively? A year in which all departments participated.
If you anticipate opposition, you'll find that all departments are involved.
Each department can submit ideas at their own pace, the Center of Excellence (CoE) receives them, and quickly produces output in collaboration with business units. After operating this cycle as an organizational function for a year, approximately 20 business and support departments are now participating in this initiative. The number of ideas on the dashboard has exceeded 70. Nine business improvement systems and two new services have been released, resulting in a reduction of 44,000 hours.
However, the biggest takeaway wasn't the numbers themselves. At the start, there were concerns that there might be opposition. But, to our surprise, all business divisions and support departments were involved.
Case study: From in-house development of core system integration to new services using generative AI.
One example is the in-house development of an IT budget management application. Initially, they were using an expensive SaaS solution for budget management. The IT department replaced this with an in-house solution using Mendix. Although this area required integration with SAP, they released and are operating it as a single product. This has become a major success story of developing a core business system in-house using low-code and putting it into production use.
Other examples include standardizing and systematizing the DX promotion process itself (idea generation/requirements → development → release → maintenance), and company-wide and departmental KPI management. Furthermore, a new service was created that eliminates the difficulty of setting prompts for generating AI from the user's awareness. This past year gave us the confidence that valuable services can be created if we properly capture the ideas in the minds of business and support departments.
Two conditions for making DX a "continuing activity"
I'd like to share two lessons I learned over the past year. The first is aligning perspectives with top management. In this initiative, the CIO and CEO acted as sponsors, agreeing on which business metrics would be impacted by digital technology. Without this agreement, it would be impossible to judge whether the results were good or bad.
DX initiatives sometimes end up as "projects" that are disbanded after only a year. On the other hand, Macnica 's management targets are set for 2030 and beyond. By linking activities to these targets, the activities will never end as long as the targets continue. In other words, they become continuous activities.
The second is the empathy of the people who actually do the work. Organizations cannot function through top-down instructions alone. It is each and every person in the business and IT departments, including ourselves, who empathizes with the strategy and takes action.
For example, the IT department might see the business divisions as somewhat unapproachable and difficult to communicate with. Conversely, the business divisions might see IT as an entity that won't help no matter what they say. In many companies, these kinds of misunderstandings may be happening on a daily basis.
If this gap can be overcome with a shared vision and a common process, profits can be generated in the shortest possible time. Aligning the perspectives of top management and gaining the empathy of those who are actually doing the work—only when these two conditions are met can a "reproducible process" where business and IT are integrated begin to operate.
What's lacking isn't the tools, but the system.
What's missing from DX that loses its reproducibility isn't the tools, but the system. This is a common saying, but Macnica has grasped its essence through practical experience. An organizational structure that continuously generates DX, the 5Ps, and the circulation of people and know-how. At Macnica, our Center of Excellence (CoE) organization was able to produce over 70 ideas and 11 releases in one year.
You don't need to get everything in order at once. Macnica 's first step was also to help people experience the value of digital technology through translation tools, starting with small steps to promote understanding.
The structure of a CoE organization is organized into the "5P" methodology introduced in this article. Macnica has also experienced the high reproducibility of CoE organizations through its own implementation and support for many clients (CoE organization building support). Why not consider these 5Ps as the next step in creating reproducibility for your own DX?