For a long time, infrastructure decisions could feel relatively binary. You either refreshed your existing hardware or moved workloads to public cloud. But critical IT rarely fits neatly into either choice. Some applications need firmer control over where data sits. Others need predictable infrastructure costs, specific recovery arrangements or continued support for technology that isn’t ready for cloud-native transformation.
That makes workload placement an increasingly important part of cloud strategy. Rather than deciding on a destination first, you can look at what each workload actually needs, including its dependencies, supportability, data location, resilience requirements, operating responsibilities and cost model. From there, you can decide whether it belongs on premises, in public cloud, on a managed platform or across a hybrid environment.
In this article, we look at the practical questions that can help you decide what should stay, move or modernise, from ageing infrastructure and cost predictability to sovereignty, resilience and legacy applications.
The infrastructure refresh cycle deserves a rethink
Replacing ageing infrastructure can solve an immediate problem, but it can also begin another cycle of capital investment, maintenance and eventual renewal. At the same time, moving every application to public cloud may not suit workloads that need greater cost certainty, UK data location or a particular level of operational control.
A managed cloud platform introduces another option, allowing suitable infrastructure resources to be consumed as a managed service rather than requiring another hardware refresh. The most important part is choosing that model because it fits the workload, rather than because it fits a predetermined cloud strategy.
Look at workloads individually
A critical business application may require stronger resilience and recovery than an internal system. Another application may have specific data-location requirements. A legacy application could remain essential to day-to-day operations but be difficult to refactor for a public-cloud environment.
Applying one infrastructure model to all of them creates unnecessary compromises.
A workload-led approach gives you the opportunity to assess factors including:
- Current infrastructure and ongoing cost
- Application dependencies and supportability
- Data-location and governance requirements
- Connectivity
- Availability and recovery requirements
- Ownership and operational responsibilities
Together, these factors provide a clearer basis for deciding what should remain where it is, what can move and what may need wider modernisation.
Sovereignty is about more than the address of a data centre
For workloads where UK data location matters, it’s worth looking beyond the physical location of the infrastructure.
You also need clarity around who operates the platform, who is responsible for each part of the service and how access, lifecycle management and accountability are handled.
These considerations become particularly important when assessing sovereign cloud services. A managed UK platform such as Wavenet Cloud Platform (WCP) can form part of that discussion where workload requirements call for UK data location, defined operational responsibilities and greater infrastructure control.
Cost predictability needs to be considered alongside flexibility
For relatively stable workloads, defined infrastructure resources delivered through a managed subscription model can make infrastructure expenditure easier to plan than a wholly variable consumption model.
Cost, however, should be considered alongside the operating model. Understanding what is included, where provider responsibility ends and where your own responsibilities begin is important when comparing different infrastructure options.
Resilience should reflect how important the workload is
Applying the same resilience model to every application can mean spending too much in one place while leaving gaps somewhere more important. A better starting point is the business impact of losing each workload.
Resilience and recovery requirements should be defined according to workload criticality rather than treated as a standard infrastructure setting. That might mean different approaches to availability, backup, retention, replication and recovery depending on the business impact and recovery requirements of each workload.
And there’s an important distinction between platform resilience and application resilience. Platform availability doesn’t automatically determine the availability of an individual virtual machine, operating system, application or third-party service. The complete service design matters.
Legacy doesn't automatically mean replace
Some of the hardest infrastructure decisions concern applications that still provide business value but weren’t designed for modern cloud environments.
A wholesale transformation may not be practical. Leaving them indefinitely on ageing or unsupported infrastructure isn’t an attractive long-term strategy either.
A managed platform can provide another route for suitable legacy or specialist applications, subject to technical review, supportability and risk assessment.
That creates an opportunity to modernise in stages. You can address infrastructure pressures first while retaining a longer-term plan for the application.
Hybrid cloud works when there's a reason behind each decision
A hybrid strategy shouldn’t mean accumulating different environments without a clear rationale. It works when each workload has been placed deliberately.
Public cloud, managed cloud, on-premises infrastructure and colocation can each play a role within a hybrid strategy where there is a clear reason for placing a workload there.
The aim is to find the right balance of cost, control, performance, resilience and flexibility for the workloads involved.
Start small and build the evidence
A sensible first step is to identify one or two workloads creating genuine pressure. That might be because infrastructure is ageing, support is becoming difficult, costs are a concern, resilience needs attention or governance requirements have changed.
From there, assess what those workloads cost today, what they depend on, how supportable they are, where their data needs to sit and what level of recovery the business actually requires.
That gives you a stronger basis for deciding what should stay, move, modernise or eventually retire, and whether public cloud, managed cloud, on-premises infrastructure, colocation or a combination of these is the right fit.
Find the right home for your critical workloads
The right destination starts with understanding the workload. If ageing infrastructure, resilience, cost, sovereignty or operational complexity is driving change, assess those requirements before deciding where to move next.
Speak to our cloud specialists to explore what should stay, what could move and where the Wavenet Cloud Platform could fit within your wider cloud strategy.
Frequently asked questions
What is workload placement in cloud strategy?
Workload placement means deciding where each application should run based on what it actually needs, rather than picking one destination for everything. You look at dependencies, supportability, data location, resilience, cost and who’s responsible for running it. Then you decide whether it belongs on premises, in public cloud, on a managed platform or across a hybrid environment.
How do I decide whether a workload should stay, move or modernise?
Start with one or two workloads that are causing real pressure, for example because of ageing hardware, rising costs or support issues. Work out what each one costs today, what it depends on, where its data needs to sit and how quickly the business needs it back after an outage. That gives you evidence to decide what to keep, move, modernise or retire. Our cloud specialists can help you through the assessment.
Is a managed cloud platform an alternative to a hardware refresh?
It can be. Instead of starting another cycle of buying, maintaining and replacing hardware, you can use infrastructure resources as a managed service on a subscription basis. This tends to suit stable workloads that need predictable costs, UK data location or more operational control. Choose it because it fits the workload, not because it fits a fixed plan. You can also explore our hardware and licensing options.
What does data sovereignty mean for UK businesses?
It’s about more than where the data centre is. It also covers who operates the platform, who’s responsible for each part of the service, and how access, lifecycle management and accountability are handled. A managed UK platform such as the Wavenet Cloud Platform can help where workloads need UK data location, clearly defined responsibilities and more control over the infrastructure.
Should every application have the same level of resilience?
No. Giving every application the same resilience model can mean overspending in one place and leaving gaps somewhere more important. Base your availability, backup, replication and disaster recovery on the business impact of losing each workload. Remember that platform resilience doesn’t guarantee the resilience of every application running on it. Find out more about business continuity.
