_
Lets work together
_
Custom is the New Standard
Rune Galschiøt
04.06.2026
A New Perspective on Enterprise Software in the Age of AI Driven Development
_
A Shift Already Underway
For more than two decades, enterprise software has been shaped by one fundamental logic: standardization at scale.
SaaS platforms have become the backbone of both digital businesses and public sector solutions. They have made it possible to implement systems faster and adopt mature capabilities without having to build them from the ground up. In many cases, this has been an effective way to accelerate digitalization.
However, recent years point to a more complex reality. Across organizations, we see how platforms gradually introduce dependencies and cost structures that become difficult to balance over time. The emergence of MACH and hybrid architectures often reflects this exact pressure.
At the same time, the cost picture has changed. Software and cloud now account for an increasingly large share of IT budgets, and many organizations report challenges in managing their SaaS spend¹. Consumption based pricing models, especially in relation to AI functionality, make it harder to predict and control operating costs².
_
When Operating Costs Shape Strategy
As platforms become central, their economic logic starts to influence strategic decisions.
Licenses, consumption, specialized capabilities, and integration layers gradually shift costs into operations. Over time, this reduces the room to invest in new initiatives. In practice, we often see development budgets increasingly being used to maintain and extend existing platforms.
This development also affects organizational freedom of action. When core capabilities are tightly connected to platforms, it becomes harder to change direction, both technically and organizationally.
_
A New Foundation for Building Software
At the same time, new technological conditions for building software have emerged.
AI driven development is influencing the entire delivery chain. It makes it possible to move faster from idea to implementation, validate solutions earlier, and automate parts of maintenance and quality assurance.
Empirical studies point in the same direction. In a controlled study conducted by GitHub, developers completed specific tasks up to 55% faster when using AI assistance³. Similar analyses from McKinsey & Company show significant productivity gains across several parts of the software development process⁴.
This is gradually changing the economics of custom software. It is becoming possible to build and evolve solutions faster and with a different cost profile than before.
_
The Role of the Platform Is Changing
In this tension, organizations are beginning to reconsider the role of the platform.
Platforms still play an important role, especially where standardization and scale are needed. But in areas with rapidly changing requirements, such as cybersecurity, their limitations become more apparent. Here, the ability to adapt solutions is often more important than following a fixed product roadmap.
The same applies to areas closely linked to an organization’s differentiation. Here, flexibility and control become critical parameters.
The result is a more nuanced approach, where platforms and custom solutions are combined based on specific needs.
_
What We See in Practice
In our work, we see this development taking shape across organizations.
Platforms are increasingly used selectively. Functionality is moved out when costs or limitations exceed the value. At the same time, custom layers are established around existing platforms, with business logic, integrations, and user experiences placed outside the platform.
In new initiatives, custom software is increasingly considered a real alternative from the start. This is especially true in situations with a high need for change or a requirement for differentiation.
We are also seeing a shift in how maintenance is understood. AI driven development makes it possible to work more systematically and efficiently with continuous development and operations, reducing the historical disadvantage of custom solutions.
_
From 02.18 to a New Procurement Landscape
This development is also reflected in the public procurement structure.
The previous SKI framework agreement 02.18 made it possible to procure relatively comprehensive IT solutions. When it expired in 2023, it was not replaced by one single framework agreement, but by a more divided landscape consisting of several agreements, each with its own focus⁵.
This provides flexibility, but also shifts responsibility for coherence and architecture to the organization itself.
_
A Broader Perspective: Dependency and Control
This development also has a strategic dimension.
Europe is highly dependent on global software vendors, particularly from the United States. This dependency applies to both platforms and AI technologies⁶.
As it becomes easier to build and maintain software in-house, new opportunities emerge to take control of parts of the digital architecture. This creates a more balanced foundation for strategic decision-making.
_
Custom as the New Default
This development does not suggest that platforms are disappearing, but that their role is changing.
Custom software is increasingly becoming a real option in situations where flexibility, speed, and control are critical.
To realize this potential, organizations need to actively make room for it. In the public sector, it is especially important to bring these considerations in early. Otherwise, procurement frameworks can lock in decisions for several years.
The new generation of SKI 02.06 points to an opening through an increased focus on market dialogue⁵. But without actively using this opportunity, new approaches will rarely become part of the decision basis.
For us as a system integrator, this is a concrete development. We work with platforms, while also investing in understanding and applying AI driven development in practice.
It does not change the need for platforms.
But it does change the starting point for how solutions are designed.
Contact
Want to Explore the Fit?
Want to Explore the Fit?