Fragmented Investment in Digital Platforms: 4 Powerful Layers to Stop Costly Waste

TL;DR: Fragmented investment in digital platforms happens when many agencies and organizations build separate digital systems without a common architecture. Each system may meet its own need well, but connectivity, data sharing and reuse are constrained while investment and operating costs rise.

Vietnam's National Digital Architecture Framework, Version 1.0, issued under Decision No. 3090/QD-BKHCN of the Ministry of Science and Technology (MOST) in 2025, answers fragmented investment with four shared layers: infrastructure and cybersecurity, data and core platforms, shared applications, and interaction channels with performance measurement.

Investing more in technology does not necessarily mean that a digital system will operate more effectively. Fragmented investment in digital platforms can emerge when multiple agencies and organizations implement digital transformation without a common architecture. A paradox can then emerge: more systems are built, more digital platforms receive investment, and more data is generated, yet the components are difficult to connect, difficult to share, and may even perform similar functions.

This article explains what fragmented investment is, the five costs it creates, how the four layer architecture replaces it, how fragmented investment compares with architecture investment, and what changes for ministries, localities, vendors, enterprises and data teams.

What Is Fragmented Investment in Digital Platforms?

Imagine a system in which multiple units all need several basic capabilities: data storage, user authentication, data sharing, analytics, reporting, or online service delivery. Instead of using a shared platform, each unit may build its own system.

• Unit A builds its own database.
• Unit B builds another database.
• Unit C continues to invest in a similar system.

Each system may operate well within its own scope. But when the systems need to exchange information, problems begin to emerge.

This can be understood as fragmented investment in digital platforms: technology resources are deployed across multiple separate systems, units, or projects without a sufficiently unified architecture and coordination mechanism.

Duplicated investment is more serious in another respect: multiple different systems are funded to address similar needs or functions instead of leveraging a shared component. Duplication is therefore the most expensive form of fragmented investment.

This does not mean that every standalone system is unnecessary. Agencies still have specialized business processes that require dedicated systems. The issue is that components that could be shared are built multiple times, or standalone systems are not designed from the outset to be interoperable.

The National Digital Architecture Framework requires agencies to develop architectures appropriate to their functions while still complying with common principles for interoperability, connectivity, sharing, and shared use.

What does a simple example of fragmented investment look like?

Suppose 10 systems all need:

• data storage;
• user authentication;
• API (application programming interface) connectivity;
• backup;
• security;
• data analytics;
• management dashboards.

If each system builds all of these components independently, resources are lost to fragmented investment. But the problem is not only that they are "built 10 times"; the 10 systems may use:

• 10 different data standards;
• 10 different identification methods;
• 10 different connectivity mechanisms;
• 10 different security processes;
• 10 different reporting systems.

At that point, the greatest cost of fragmented investment may no longer be the cost of building the systems, but the cost of connecting them and maintaining a fragmented digital environment.

What Are the Consequences of Fragmented Investment in Digital Platforms?

When digital platforms develop through fragmented investment and duplicated capabilities, at least five major issues can emerge. Each one compounds the others, which is why fragmented investment is a strategic problem and not only a budget line.

1. Resource waste from fragmented investment

Building a technology capability multiple times means that financial, human, and time resources are also used multiple times. Instead of investing once in a platform capable of serving multiple units, resources are divided among many smaller projects.

Therefore, the requirement to "limit duplicated, overlapping investment and waste" is directly incorporated into the implementation of the National Digital Architecture Framework. According to VnExpress reporting on the implementation conference (September 16, 2026), the central government plans to invest in 79 national shared digital platforms for ministries, sectors, and localities to reuse.

2. Data fragmentation caused by fragmented investment

This may be an even more important consequence than the cost of fragmented investment. The same information may appear in multiple systems but in different formats. For example, the same organization may be recorded as "ABC Company" in one system but as "ABC Co., Ltd." in another. Without a unified identifier standard, computers cannot easily recognize that the two records refer to the same entity.

When data is not standardized, data integration, large scale analytics, and AI (artificial intelligence) applications become more difficult. This is also why the Digital Architecture Framework defines data as needing to be governed, standardized, connected, and shared according to the principles of being "correct, complete, clean, live, unified, and shared."

3. Difficulty interoperating across systems under fragmented investment

A platform may work well when operating independently but become less effective when it needs to exchange data with another system. Organizations then have to build additional middleware layers, dedicated connections, or perform manual data conversion.

The result of fragmented investment is an increasingly complex architecture: the more independent systems there are → the more connection points are required → the higher the maintenance cost.

4. Difficulty creating a unified data picture after fragmented investment

This is particularly important for governance. If each system maintains its own dataset, leaders may receive multiple reports without necessarily having a "single source of truth", a unified data source for comparison and decision making.

The question is then no longer "Do we have data?" but rather "Which data is correct, which data is the most up to date, and are the systems referring to the same entity?"

DataCore has raised a similar issue in its analyses of data infrastructure: the problem for many organizations is not necessarily a lack of data, but data being trapped in silos, using different definitions, and being updated on different schedules, causing answers from different systems to conflict.

5. Reduced ability to leverage AI under fragmented investment

AI needs more than models. AI needs high quality, structured, accessible data with context. If fragmented investment leaves data scattered across multiple systems and inconsistent in format, AI may struggle to:

• consolidate information;
• identify the same entity;
• analyze trends;
• forecast;
• automate processes;
• generate reliable metrics.

In other words, a fragmented digital system can generate a great deal of data without generating a corresponding amount of "intelligence."

How Does the 4 Layer Digital Architecture Replace Fragmented Investment?

This is an important point in the approach of the National Digital Architecture Framework. Instead of viewing each technology project as an independent system, which is the root of fragmented investment, the overall architecture places them within a common structure. Four layers are defined:

• Layer 1: Shared digital infrastructure and cybersecurity
• Layer 2: Data and core platforms
• Layer 3: Shared applications and business functions
• Layer 4: Interaction channels and performance measurement

This is not simply a way of dividing a system into four technical layers. It creates a logic: Infrastructure → Data → Application → Interaction and Measurement.

Starting with physical infrastructure and security, data is organized at the core layer; applications leverage the core layer; finally, results are delivered to users and measured. According to Decision No. 3090/QD-BKHCN (MOST, 2025), Version 1.0 of the National Digital Architecture Framework is the basis for building, developing, upgrading, and operating digital infrastructure, information systems, digital platforms, databases, and digital services in a synchronized and unified manner, with connectivity, interoperability, sharing, and shared use.

Layer 1: Shared digital infrastructure and cybersecurity, the "foundation"

Without infrastructure, the layers above cannot operate. The first layer provides basic capabilities such as:

• computing capacity;
• storage;
• network connectivity;
• data centers;
• cloud computing;
• monitoring;
• safety and cybersecurity.

This layer can be compared to a city's transportation, electricity, and water infrastructure. An application does not need to build its own "road" if shared transportation infrastructure already exists. Similarly, a digital system does not necessarily need to build all of its infrastructure from scratch if it can use shared infrastructure capabilities that meet its requirements. That is the first place where fragmented investment can be reduced.

But "shared" does not mean "used without control." This layer must address fundamental requirements such as:

• Who is authorized to access the system?
• Where does the data travel?
• Can the system recover when an incident occurs?
• Is information protected throughout its lifecycle?

The architecture framework also emphasizes that cybersecurity, data protection, personal data, privacy, and state secrets must be safeguarded from the design stage and throughout implementation.

Layer 2: Data and core platforms, the "heart" of the architecture

If Layer 1 is the foundation, Layer 2 can be considered the center of the architecture. According to the Digital Architecture Framework, this is where data is created, stored, shared, and processed into knowledge. Data must be standardized according to the principles of being "correct, complete, clean, live, unified, and shared."

The key point here is that it is not enough to store data; data must be turned into an asset that can be reused. This requires:

• Standardization: systems must interpret data in the same way.
• Integration: data from multiple sources can be connected.
• Sharing: once data has been created, it does not necessarily need to be entered again into other systems.
• Governance: it must be clear who owns the data, who can use it, for what purposes, and how quality is controlled.
• Utilization: data must be usable for reporting, analytics, forecasting, and AI.

An important principle reflected in digital architecture frameworks is the concept of "enter data once, use it in many places." This is how data is transformed from a by product of individual applications into a shared asset of the entire system, and it is the layer where fragmented investment does the most lasting damage when it is ignored.

Layer 3: Shared applications and business functions, "where data creates value"

Data does not create value by itself unless it is incorporated into business processes. The third layer includes applications and platforms serving common business functions that span multiple units.

The goal is not to turn every agency into exactly the same system. Instead, what is common should be shared, and what is unique should be developed separately. This is an important principle for avoiding a situation in which each unit builds a different version of the same function. That is fragmented investment at the application level.

When a shared platform already exists, units can focus their resources on genuinely differentiated business processes instead of rebuilding basic components. Under the Digital Architecture Framework, the shared application and business function layer is built on the principle of shared use to avoid duplication and waste.

Layer 4: Interaction channels and performance measurement, "where value is validated"

A digital system cannot be evaluated solely by the number of servers, applications, or databases built. Ultimately, it must answer:

• Are users actually using it?
• Are services better?
• Are processes faster?
• Are costs lower?
• Is data being used effectively?

That is the role of the fourth layer. This layer includes user interaction channels and tools for tracking, monitoring, and measuring outcomes. In a digital architecture, dashboards, KPIs (key performance indicators), and real time data can transform operations from "reporting after an event has occurred" into continuous monitoring for detection and adjustment.

According to the same VnExpress report (September 16, 2026), the national digital architecture management system is expected to use data automatically collected from existing systems, thereby reducing repeated data entry, limiting manual reporting, and enabling real time monitoring of results.

This is also the final connection point of the architecture: Data → Analysis → Action → Measurement → New data. A loop is formed. Without this loop, fragmented investment cannot even be detected, because nobody measures it.

Why Is a Data Centric Approach Important for Ending Fragmented Investment?

There is a major shift in digital transformation thinking. In the early stages, organizations often ask "What application do we need to build?" But as systems grow, the more important question becomes "What data already exists, and how can we use that data?"

This is the shift from application centric to data centric. If applications are at the center, each application can form its own "data repository." If data is at the center, applications become one of the ways to leverage data. The difference is significant.

With an application centric approach: Application A → Data A; Application B → Data B; Application C → Data C. The result is data divided into silos, the typical footprint of fragmented investment.

With a data centric approach: Data sources → Shared data platform → Multiple applications. The result is that the same data asset can be leveraged across multiple business functions.

This is why DataCore, in its analyses of data infrastructure, also emphasizes that data needs to be structured, validated, and connected rather than simply collected.

A system may own terabytes of data and still lack analytical capabilities if the data is unstructured, lacks common standards, or cannot be connected. More data from fragmented investment does not necessarily mean more valuable data. Value emerges when data can be found → understood → connected → analyzed → used → measured.

How Does Fragmented Investment in Digital Platforms Affect AI?

AI is often described as a technology capable of automation and creating new ways of working. But AI does not operate in a vacuum. AI needs data, a point DataCore also makes in its coverage of Vietnam's AI strategy to 2030. A model can be highly capable, but if its input data is:

• incomplete;
• incorrect;
• duplicated;
• inconsistent;
• outdated;
• lacking context;

then the output can also be affected. Fragmented investment makes each of these problems more likely. Therefore, placing AI and data alongside each other in an architectural framework does not mean "putting AI into everything." It means building a system in which AI can leverage well organized data to perform appropriate tasks.

The National Digital Architecture Framework identifies AI as capable of supporting analytics, forecasting, and process automation. At the same time, it emphasizes that for decisions affecting the lawful rights and interests of organizations and individuals, authorized persons remain primarily responsible and AI does not replace human decision making responsibility.

This creates an important principle: AI can augment decision making capabilities, but AI quality depends on the quality of the underlying data system and governance mechanisms. Fragmented investment weakens exactly that foundation, because every silo adds one more version of the truth that a model has to reconcile.

How Do Fragmented Investment and Architecture Investment Compare?

The table below summarizes how fragmented investment works, as described so far. It compares the two approaches dimension by dimension; it does not contain measured figures.

Takeaway: fragmented investment looks cheaper project by project, but architecture investment is cheaper system by system, because shared data and shared components stop the cost of connection from growing with every new platform.

What Does a Worked Example of Fragmented Investment Look Like?

Illustrative only: the scenario below is hypothetical. The province, departments, counts and outcomes are invented to show the mechanism, and no part of it is reported fact.

Imagine a province where six departments each keep their own database of registered businesses. Under fragmented investment, every department that needs data from another department builds a direct connection. If every pair of departments connects, the province maintains 6 × 5 ÷ 2 = 15 separate connections.

Now suppose one department changes how it records company names. Up to five connections may need to be checked or rewritten. Meanwhile the same company might appear as "ABC Company" in one database and "ABC Co., Ltd." in another, so a report that counts businesses across departments can double count it.

When a seventh department joins, fragmented investment adds six more connections, bringing the total to 21. Under fragmented investment, each new system makes the next one more expensive to integrate.

Under an architecture approach, the same seven departments each connect once to a shared data platform, which means 7 connections instead of 21. A shared business identifier lets the platform recognize that "ABC Company" and "ABC Co., Ltd." are one entity, so every department sees the same record.

The general mechanism is simple arithmetic. With point to point integration, the number of possible connections grows roughly with the square of the number of systems. With a shared platform, it grows in a straight line. That is why fragmented investment becomes more costly the more successful digital transformation is.

Why Should the Four Layers Not Be Viewed as Four Separate Systems?

The real value of the architecture lies in the relationships among the four layers. The entire system can be viewed as a chain:

• Layer 1, Infrastructure: provides computing, storage, connectivity, and security capabilities.
• Layer 2, Data: standardizes, integrates, governs, and shares data.
• Layer 3, Applications: turns data and platforms into specific processes, services, and business functions.
• Layer 4, Interaction and Measurement: delivers services to users, captures feedback, and measures outcomes → new data → returns to the data layer for further analysis and utilization.

This is precisely what makes a unified architecture different from simply "building more systems." The goal is not to have as much technology as possible. The goal is to make existing technologies operate as one system. If the four layers were procured as four separate projects, the architecture itself would become another case of fragmented investment.

What Changes for Whom When Fragmented Investment Ends?

Moving away from fragmented investment is not only a technical change. It shifts responsibilities and incentives for several groups.

• Ministries and central agencies: the question moves from "Which system does my agency need?" to "Which shared platform already covers this, and what is genuinely unique to us?" Budget proposals are more likely to be judged on reuse, connectivity and measured outcomes.
• Provinces and localities: local digital transformation projects can reuse national shared platforms instead of rebuilding basic components. DataCore has covered this shift in its article on Vietnam's digital transformation platforms in 2026.
• Technology vendors and system integrators: selling a standalone system becomes harder than selling a component that plugs into shared infrastructure, follows common data standards and exposes clean interfaces, as buyers learn to spot fragmented investment.
• Enterprises, banks and financial institutions: as public data becomes more standardized and connected, the quality of data they receive through official channels may improve, and their own internal architectures face the same fragmented investment question.
• Data and analytics teams: work shifts from reconciling the conflicting records that fragmented investment leaves behind toward governing shared data, defining identifiers and measuring quality. DataCore discusses the enterprise side of this in enterprise AI data infrastructure in Vietnam.

How Do Organizations Move From Technology Investment to Architecture Investment?

The issue of avoiding fragmented investment in digital platforms and duplicated capabilities is therefore not merely a budget issue. It raises a larger question: are we investing in individual systems, or are we building reusable digital capabilities?

To avoid fragmented investment, a good platform should not be evaluated only by how many functions it has. It should also be evaluated by these questions:

• Can it connect to other systems?
• Is the data standardized?
• Where can it be reused?
• Can it scale?
• Is it secure?
• Can it generate valuable data for other systems?
• Can its outcomes be measured?

This is also why the National Digital Architecture Framework is defined not merely as a technical document but also as a governance tool for digital transformation. Implementation effectiveness is considered in relation to connectivity, sharing, shared use, reduced fragmented investment and duplicated investment, resource savings, security, and improved service quality.

The same logic applies inside companies. DataCore, a Vietnamese financial and business data platform, works on the data layer this framework describes: structured, validated and connected data on Vietnamese companies. Teams that want to see connected entity data in practice can explore DataCore services or try the DataCore company data trial.

Is the Future of Digital Transformation About the Number of Platforms?

Digital transformation may begin with an application. But a digital ecosystem cannot be built simply by adding more and more applications. When each unit builds its own system, fragmented investment can cause data to become fragmented. When data becomes fragmented, connectivity declines. When connectivity declines, analytics and AI applications also become more difficult.

Conversely, a unified architecture creates a different logic: shared infrastructure → unified data → reusable applications → measured interactions → continuously improved data. This is why the four layers of the National Digital Architecture Framework have significance beyond a technical diagram: they are a practical defence against fragmented investment.

It shifts the question from "What system does each unit need to build?" to "What capabilities does the entire system need that can be shared, what data needs to be connected, and how should the value created be measured?"

In this model, data becomes the central asset, while AI becomes a capability for leveraging that asset at greater scale. Ultimately, the goal of digital architecture is not to create more systems. The goal is to create a system in which what has already been built can continue to be reused, connected, and generate value across multiple contexts.

Frequently Asked Questions About Fragmented Investment

What is fragmented investment in digital platforms?

Fragmented investment in digital platforms is the deployment of technology resources across many separate systems, units or projects without a unified architecture and coordination mechanism. Each system may work on its own, but connecting, sharing and reusing data across them becomes difficult and costly.

Is fragmented investment the same as duplicated investment?

Not exactly. Fragmented investment describes resources spread across disconnected systems. Duplicated investment is its most expensive form, where several systems are funded to solve the same need instead of sharing one component.

What are the four layers of the National Digital Architecture Framework?

The four layers are shared digital infrastructure and cybersecurity; data and core platforms; shared applications and business functions; and interaction channels and performance measurement. Together they form one chain from infrastructure to measured outcomes.

Why does fragmented investment make AI harder to use?

AI depends on structured, consistent data with context. Fragmented investment scatters data across systems with different formats and identifiers, so models receive incomplete or conflicting inputs and their outputs become less reliable.

Does ending fragmented investment mean every agency uses the same system?

No. The principle is that what is common should be shared and what is unique should be developed separately. Agencies keep dedicated systems for specialized processes, but those systems are designed to interoperate with shared platforms.

How many national shared digital platforms are planned?

According to VnExpress reporting on the implementation conference (September 16, 2026), the central government plans to invest in 79 national shared digital platforms for ministries, sectors and localities to reuse.

References

• VnExpress, "Avoid fragmented and duplicated investment when developing digital platforms", September 16, 2026. vnexpress.net
• Ministry of Ethnic and Religious Affairs, "National Digital Architecture Framework: Unified, interoperable, and data centric", August 17, 2026. bdttg.gov.vn
• Ministry of Science and Technology, Decision No. 3090/QD-BKHCN on the National Digital Architecture Framework, 2025. thuvienphapluat.vn