Cloud repatriation – the movement of workloads from public cloud back to on-premises or private infrastructure – is accelerating.
According to a recent QA report, as many as 71% of companies have considered repatriation, driven primarily by escalating cloud costs and unexpected financial complexities.
In short, cloud wasn’t the blue sky many were promised. But this shift fundamentally misinterprets why cloud investments fail.
In this piece, I will show how it’s rather how we’ve failed to understand cloud for what it is. How, along many fronts, treating cloud technologies like bare metal servers has held back true innovation and the kinds of efficiencies offered by true cloud-native approaches.
You only have to look at the huge savings possible with practices like FinOps to see that the potential is there.
So, why are we retreating from savings?
Here Are 3 Reasons Cloud Initiatives Struggle and Why Repatriation Rarely Provides the Answer
Traditional MSPs Lagging in the Cloud-Native Era
Managed Service Providers (MSPs) critically shape cloud success. However, many MSPs remain trapped in traditional IT practices. They’re accustomed to managing servers, patching security vulnerabilities, and performing routine infrastructure maintenance. Cloud-native operations, like serverless computing, container orchestration, and infrastructure automation, demand entirely different skillsets.
Partnering with outdated MSPs results in high costs and low ROI while companies pay premium prices without the benefits.
This is especially true of those larger cloud hosting businesses that pivoted from bare metal to cloud without sufficiently updating their core business model, leading to a high market penetration by those unable to drive real results.
Success demands MSPs fluent in cloud-native practices: automation, Infrastructure as Code (IaC), and dynamic scalability. Without these capabilities, inefficiencies persist, leading businesses to blame the cloud instead of their partnerships.
Skeuomorphic IT Design: Old Habits Die Hard
Another core challenge is “skeuomorphic IT” – replicating familiar on-premises infrastructure in the cloud. Many organisations use a “lift-and-shift” approach, simply moving existing systems without redesign. This shortcut feels comfortable but sacrifices the cloud’s true advantages.
Cloud efficiencies are driven by taking advantage of virtualisation: ephemerality, elasticity, and resource pooling. Skeuomorphic design undermines this, reducing cloud to just another data center – one managed by Amazon somewhere far away. Real ROI emerges only when teams embrace cloud-native designs and cloud-native thinking.
Outdated Governance Frameworks
Another significant barrier to cloud success is outdated governance frameworks that haven’t evolved alongside cloud technology. Traditional governance, designed for stable, predictable on-premises environments, often conflicts with the dynamic, scalable nature of cloud computing.
Companies using legacy governance models struggle to manage rapid provisioning, decentralised decision-making, and the continuous optimisation that cloud demands.
Modernising governance means aligning policies and procedures with cloud capabilities, enabling agility and innovation without sacrificing control. Effective governance in the cloud era includes clear visibility, robust security tailored specifically to cloud environments, real-time cost management, and the flexibility to scale quickly.
Organisations failing to update their governance frameworks find themselves mired in inefficiency and uncontrolled costs, often leading them to mistakenly blame cloud itself.
The Hidden Costs of Repatriation
The QA report highlights another often-overlooked problem: repatriation itself isn’t cheap or easy. Returning workloads to on-premises infrastructure involves significant capital expenditures, increased staffing demands, and renewed responsibility for hardware lifecycle management and security. These hidden costs quickly add up, undermining anticipated savings.
Furthermore, repatriation can stall innovation and agility–essential competitive advantages in today’s fast-moving market.
So, for businesses stuck between a rock and a hard place, what is the answer?
Embracing Cloud Design Principles: The Real Path to ROI
Repatriation occasionally makes sense, such as for regulatory compliance or specialised technical requirements. But generally, repatriation means retreating rather than solving fundamental issues.
True cloud ROI requires culturally, technologically, and organisationally embracing cloud design principles.
This means designing specifically for cloud strengths – leveraging microservices, serverless platforms, containerised architectures, scalable infrastructures, and managed services. The MACH architecture (Microservices, API-first, Cloud-native, and Headless) model is absolutely the application model of the future.
Successful cloud strategies also require partnering with modern MSPs who deeply understand cloud-native operations. They are key in helping organisations without the in-house skills to ride the wave of technological development.
On a governance level, adopting FinOps/GreenOps and other cloud-native governance strategies will help align organisations with the potential of cloud.
Last but not least, no think piece in 2025 would be complete without a reference to the benefits of AI. However, sometimes a cliche is so because it’s the stone-cold truth. Already, AI is making huge improvements to resource management in cloud workloads – a trend that’s only set to continue.
Organisations embracing these comprehensive strategies significantly reduce costs, boost operational efficiency, and accelerate innovation. Instead of retreating, businesses must reassess their cloud approach holistically – partnering effectively, designing intelligently, aligning culturally, and fully exploiting the cloud’s capabilities.
The solution isn’t repatriation. The solution is evolution. Aligning cloud strategies with clear business goals and genuinely adopting modern cloud practices will deliver the ROI businesses seek.
