Navigeren in de nieuwe IT-realiteit

A few years ago, the direction of travel was clear: move as much as possible to the cloud. The debate focused largely on how quickly organisations could get there and which route they should take. Today, the conversation has changed. Organisations want to know where their data reside, how dependent they have become on individual providers, how they can demonstrate business continuity, and what shifting costs and lead times may mean for their long-term IT strategy.

The cloud remains an important part of the equation, but it is no longer the automatic answer for every workload. Instead, more organisations are reassessing the balance between public cloud, private cloud, and on-premises infrastructure. The reason is simple: control, flexibility, and resilience have become more important than ever.

In this article, I explore seven topics that regularly surface in discussions about data centre strategy. Although each presents its own challenges, they are closely connected. The underlying question remains the same: which decisions will still give your organisation the freedom to adapt a few years from now?

Over the past decade, many organisations have migrated significant parts of their infrastructure to US hyperscale cloud providers in pursuit of greater flexibility, resilience, and operational simplicity. Today, however, data sovereignty has become a far more prominent concern as European requirements around privacy, compliance, and strategic autonomy continue to evolve.

Under certain circumstances, the US CLOUD Act may allow US authorities to request data from American cloud providers, even when that data are stored within Europe. For organisations in sectors such as government, healthcare, financial services, and industry, this creates important questions around control and governance. As a result, more organisations are reassessing where their data reside, who has access to them, and how much control they want to retain over critical information. Europe’s digital sovereignty journey is still evolving, but these considerations have become part of the strategic discussion. The growing focus on sovereignty is also driving renewed interest in hybrid infrastructure models.

The cloud is no longer the automatic choice it once seemed to be. For years, many organisations followed an almost cloud-first approach, moving applications to public cloud platforms in pursuit of scalability, agility, and speed. A clearer picture is now emerging of the long-term costs, dependencies, and operational implications that come with those decisions.

Many organisations are encountering similar challenges:

• Rising monthly cloud costs
• Limited predictability of consumption
• Complex licensing models
• High data-egress costs
• A lack of a clear exit strategy

What initially appeared highly flexible can prove remarkably difficult to reverse in practice. As a result, organisations are increasingly shifting from a cloud-first mindset to a workload-first approach. The focus is no longer on where workloads should run by default, but on which environment is best suited to each application, dataset, and business case. That shift is prompting more organisations to embrace hybrid architectures in which public cloud, private cloud, and on-premises infrastructure coexist and complement one another.

What happens when a critical part of your IT landscape depends on a single provider? For many organisations, the VMware pricing changes following Broadcom’s acquisition brought that question sharply into focus, highlighting just how dependent they had become on a single vendor.

Infrastructure environments that have evolved over many years around VMware can be difficult to migrate. Knowledge, tooling, processes, and operational dependencies often become deeply embedded throughout the organisation. Alternative hypervisors are available, but not every option offers the same maturity, ecosystem support, or feature set, making migration both complex and potentially risky. The consequences are familiar:

• Unexpectedly high licensing costs
• Reduced negotiating power
• Slower innovation
• Dependence on a single ecosystem

As a result, more organisations are exploring platform-independent infrastructure strategies and alternatives to dominant market leaders. Multi-vendor architectures are becoming an increasingly important part of that conversation. The goal is not simply to reduce dependency, but to make flexibility a core design principle once again.

Cyberattacks continue to increase, while regulatory requirements become more demanding. With frameworks such as NIS2 and DORA, the focus is shifting from simply having security measures in place to demonstrating organisational resilience. Achieving that requires more than a backup solution alone.

Organisations must be able to prove that their infrastructure can recover data and restore systems to a specific point in time. Regular testing and documented recovery procedures are essential to provide that assurance. In practice, the focus is often on protecting and recovering the organisation’s most critical data and systems.

Complicating matters further, ransomware attacks increasingly target backup environments themselves. As a result, capabilities such as air-gapped storage, immutable backups, and disaster recovery have become essential elements of modern IT strategies. Compliance is no longer a box-ticking exercise, but an ongoing process of demonstrating digital resilience.

Almost every organisation sees opportunities in AI, yet many are still grappling with the same question: where do you start? Successful adoption requires far more than technology alone. Without clear use cases, reliable data, and effective governance, AI initiatives often fail to progress beyond experimentation.

That is why conversations rarely start with hardware or GPUs. The more important questions come first: What data will you use? How will you make those data available? Are there regulatory or governance considerations to take into account? And, most importantly, what outcome are you trying to achieve?

AI depends on access to high-quality data, which must be collected, managed, refined, and continuously improved. Building those capabilities is an ongoing process that often delivers value over the medium to long term. Not every project needs to generate immediate results to be successful. In many cases, early initiatives form an essential part of the learning process.

Large-scale data processing in the cloud is not always the most cost-effective option. Many organisations start their AI journey in the cloud before expanding towards private cloud or on-premises environments as requirements mature. Practical first steps include:

• Automating repetitive processes
• AI-assisted service desk operations
• Intelligent analysis of complex documents
• Predictive analytics
• Knowledge management support

IT teams are under increasing pressure as the shortage of specialised skills continues to grow. At the same time, supply chains for hardware and critical components remain unpredictable in parts of the market. The result is increased pressure on innovation, operations, security, business continuity, and scalability.

This all means that IT departments are being asked to deliver more with fewer resources. Maintaining agility increasingly depends on automation, standardisation, and more intelligent ways of managing IT environments. Forward planning and timely investment can help organisations avoid unnecessary cost increases and mitigate supply-related risks. I've seen projects where quotations increased from around €100,000 to €180,000 within just a few months. Understanding which replacements are likely to be required over the coming years makes it possible to evaluate procurement strategies earlier and make more informed investment decisions.

At the same time, many organisations are looking for more than a technology provider. Growing complexity and limited IT capacity are increasing the need for partners that can contribute expertise, ongoing support, and strategic guidance. Automation plays an important role here, reducing the operational burden on internal teams and freeing up resources for higher-value work. In networking, for example, AI can already detect and resolve many first-line issues before users notice a problem or IT teams need to intervene.

Traditionally, organisations treated IT as a capital investment (CapEx), purchasing hardware up front, depreciating it over time and using it for years. Today, however, they are increasingly prioritising flexibility over large upfront investments.

The move towards OpEx-based models offers several advantages, including:

• Predictable monthly costs
• Faster scalability
• Reduced upfront investment
• Greater flexibility as business needs change
• Simplified lifecycle management
• Less impact on the balance sheet

That said, few organisations are looking for a one-size-fits-all approach. Not every workload makes financial or technical sense within a fully cloud-based OpEx model. Instead, many organisations are adopting hybrid consumption models, choosing the financial and technical approach that best fits each individual workload.

One of the biggest challenges in modern IT is overprovisioning, where infrastructure consistently exceeds actual requirements. Many organisations continue to pay for unused storage, excess compute capacity, oversized licensing models, and underutilised cloud resources. For years, that was rarely questioned. Additional capacity came at relatively little extra cost, and planning for growth often felt safer than sizing infrastructure too tightly. As costs continue to rise, however, those assumptions are being challenged.

Rightsizing is about continuously aligning capacity with actual demand. Achieving that requires visibility, monitoring, and flexibility across both infrastructure and commercial agreements. The benefits are clear:

• Lower costs
• More efficient energy consumption
• Improved performance
• Greater predictability
• More sustainable IT operations

Tomorrow’s IT strategy is no longer defined solely by innovation or speed. The priority is gaining greater control over costs, data, security, suppliers and the ability to scale. Achieving that calls for more deliberate data centre decisions, sustained investment in digital resilience, fewer dependencies and a smarter approach to capacity management.

No single blueprint exists. Public cloud, private cloud and on-premise infrastructure can all play a valid role within an organisation’s technology landscape. What matters is choosing technology that supports the business case today while preserving the flexibility to adapt as requirements evolve over time.

Technology is evolving at pace, making it increasingly difficult to assess where your organisation stands today and where opportunities for optimisation exist. Are you paying for capacity you no longer use? Does your infrastructure still support your plans for the years ahead? And where might risks be emerging around continuity, security, and dependency?

ARP’s Digital Health Check provides a clear picture of your current IT environment, helping you identify key areas for attention while uncovering practical opportunities to reduce costs, minimise risk, and align your infrastructure more closely with actual business requirements.