Cost is a significant concern around cloud migrations. Because cost is one of the opaque items in the process, it can be difficult to predict the exact costs.

In 2023, the worldwide infrastructure as a service (IaaS) market grew to $140 billion. In 2024, end-user spending on public cloud services is forecast to grow 20.4% to $675.4 billion, with infrastructure-as-a-service (IaaS) experiencing the highest end-user spending growth at 25.6%. Most of the spending happens without a proper plan in place, resulting in overspending.

However, Cloud costs remain difficult to predict and rein in, and most problems begin with the initial migration. Most businesses try to make a like-for-like comparison between what they have and hope to achieve, which results in a lift-and-shift approach. But this is a dangerous bet, says Louis van Schalkwyk, Head of Technical Operations at Digicloud Africa: “I don’t believe that costs should be a primary view when deciding to migrate. When evaluating options, customers should decide where they will be able to derive the most benefits from their data for their businesses. This often translates to big data and machine learning, which the cloud is suited for. These services can help transform businesses to gain a competitive advantage. Moving your virtual machines from one place to another does not change much and might cost you more on the cloud than on-prem.”

Cloud Vs. On-Prem

Yet, companies will need a frame of reference to understand their cloud costs. According to Gordon Davey, SoftwareONE’s Global Service Line Leader and General Manager for Cloud Services, Azure, most will lean towards what they already own.

“We often use on-premise comparisons with customers to help them start their journey and understand what it might look like. The reality is that you need a reference point to begin with. You need to understand where you are today and what that could look like when you are done.”

Van agrees, noting that most companies will make a cost comparison, whether fair or not. Factors such as sunk costs in current systems could suggest that on-premise is cheaper, especially if you primarily look at hosting virtual machines, are happy with the associated capital investments, and have already invested in the infrastructure, power, and network redundancy. However, the value derived from the cloud doesn’t lie in virtual machines. It is about building scalable solutions that take advantage of the elasticity offered by the cloud. It is also about using managed services that give a chance to move away from managing infrastructure and focus on building solutions rather than building solutions that solve business platforms.

The importance of the first project

Cloud migrations have many different parts and constituencies. Not everything will migrate simultaneously, mainly when you include their dependencies on other systems. Go too broad or deep too fast, and you end up with many cascading peripheral adjustments that could make the migration process more complex, leading to higher costs.

Complexities aren’t exclusive to size. Recognizing that different systems will change a migration’s parameters. Some systems might save money merely after shifting to cloud infrastructure, while others acquire extensive retooling or support from new services. Your teams must anticipate such changing scenarios, and it is helpful to start with a handful of different migration projects.

Businesses should try to scope it down to understand what a migration will look like for one or two applications within their environment. Looking at the scope of an application and a couple of supporting applications will help test the environment. That spectrum of variety can reveal different things. The best results may be a lift-and-shift into the cloud with very little change. However, there will be other applications where the business value for you as a customer is extracted by transforming and modernizing that application.

The whole approach will look completely different in that case. That could result in substantial cost savings. The cost and benefit can differ dramatically; you need to understand those parameters. The approach will also help businesses identify how different applications will change regarding their costs. While many cloud features sound great on paper, not all necessarily represent a permanent arrangement. An application could have different needs in different phases, and providing an optimum cloud solution will also help address this.

Additionally, the public cloud has a perfect business case for organizations looking for temporary or elastic capacity, whether an unexpected computing demand or specific use case-specific SaaS/PaaS. When an organization has a workload that has moved past the proof-of-concept phase and is in production, that workload provides a valuable outcome or business function, and most run 24x7x365. Workloads, such as SAP or Oracle that are always on and have a steady state performance profile no longer benefit from the elasticity of the public cloud.

Monitor value

The best migration framework, such as lift-and-shift or retooling the application or workload, is crucial for cloud costs. A migration might require several frameworks. Each is used as the migration target, and its dependencies change. All these steps will incur different costs and value delivery. Change management is also often overlooked as a recurring cost; every change in migration may require consulting users beforehand and training them. Cloud costs are easier to predict if you accept that a given system will have multiple migration events, possibly requiring different migration frameworks and user engagement.

How does one stay on top of everything?

Essentially, have cost-monitoring tools in place early on. Businesses need the right tools to understand costs and monitor them very closely as they move closer to that apex model. If you don’t keep track of them, they can quickly spiral out of control. So, having the tools in place is important.

But, one should not just rely on that cost management. Also, look for ways to manage the phasing-out of older systems. Otherwise, businesses could end up with dual costs and be unable to recoup the costs of what they were doing previously. Businesses could also use a toolset to monitor processes. However, the future might see a need for an end-to-end view of a system, i.e., a system-wide view at a cost.

Ultimately, cloud migration is not just about moving workloads; it’s about transforming business operations to harness the full potential of cloud capabilities. While cost remains critical, businesses must approach migration strategically-focusing on long-term value, scalability, and efficiency rather than just immediate cost comparisons.

Organizations can optimize cloud spending and avoid unnecessary overhead by carefully planning the first migration project, understanding workload-specific needs, and implementing robust cost-monitoring tools. The key is to view cloud adoption as a journey that requires continuous assessment, adaptation, and alignment with business goals. With the right strategy in place, enterprises can unlock the benefits of the cloud, ensuring not just cost efficiency but innovation and competitive advantage in the digital era.

Sarthak Rohal – Senior Vice President, Product Team