What do artificial intelligence, the Internet of Things, machine learning, and big data analytics have in common? Two things:
- Enterprises are adopting these technologies at an ever-increasing rate – and those that fail to are likely to be left behind.
- They all require public cloud technology to function effectively, at least to some extent.
Organizations are well aware of point two. In August 2018, IDG reported that 73% of organizations have at least one application or workload in the cloud. The same study predicts that by 2021, every company will have at least some part of their IT infrastructure in the cloud.
If you’re one of the few yet to adopt the public cloud, your first step is to choose a Cloud Service Provider (CSP). This means that you are likely choosing between Microsoft Azure, Amazon Web Services (AWS), and Google Cloud Platform (GCP).
Without further ado, a breakdown of analyzing the cost of migrating to the public cloud:
How predictable are your workloads?
Understanding how much compute you require on an ongoing basis will significantly impact the cost of your Cloud migration, as all three of major CSPs incentive committed resource provisioning. For instance, Azure offers 32% savings on 1-year reserved compute contracts, and 56% for 3-year contracts.
Part of the benefit of the cloud is its elasticity and scalability – the cloud provides additional compute power instantly, at a fraction of the cost of investing in new hardware. But most of your workloads will remain relatively stable. Even workloads that fluctuate typically have a minimum computing requirement. By understanding CSP commitment incentives and accurately you determining the predictable elements of your IT usage, you will be able to lower your CSP costs.
Latency and Performance
Next, consider location. Do you have latency-sensitive applications that require close physical proximity to your offices? Depending on your region, Azure, AWS and Google all charge different rates.
Additionally, your Operating System (OS) of choice will impact your costs in the cloud. Azure, AWS and Google offer different cost structures based on which OS you use. While some OS licenses are free (i.e.Ubuntu, Debian), many are not and will begin to add up depending on your chosen provider. If you choose to swithc your company-wide OS to fit better with your chosen CSP, ensure you understand the associated costs and factor them into your cloud migration costs.
Prices change frequently
Like everything IT, CSP prices across services change frequently. The team at Rightscale found that 70% of the 26 CSP price points they analyzed between April and November dropped. AWS dropped the most (24 of 26 prices); Azure dropped 19 out of 26; GCP only dropped 4 out of 26 prices.
The key takeaway here is that you need to continually evaluate CSPs not just for their overall cost, but for the cost of the individual services that are most important to your cloud strategy.
Read also: Migrating to AWS platform
The hidden costs of cloud migration
It’s important to note that most organizations only look at cost as a top-level number, to be paid to your CSP and Managed Service Provider(MSP) over the course of a migration and then in the form of monthly cloud consumption costs.
But this doesn’t take into account the internal costs associated with a cloud migration. Does your team have the skills necessary to perform the migration? Will you have to train existing staff, or hire new talent?
These costs can add up quickly and should be assessed as having equal weight as your operational costs. The good news is that many CSPs will cover these costs as an incentive to get your business. This should be used as a negotiation tactic from the beginning of your relationship with any prospective cloud vendor.
If there’s one overarching takeaway, it’s that choosing the right CSP will depend on your unique business needs. Organizations that take the time to research and match up with a CSP that meets their requirements will be rewarded.