I frequently consult with customers comparing hyperscale public cloud platforms like Amazon Web Services (AWS), Microsoft Azure and Google for their data and applications. The workloads they want to evaluate typically fit into one of three categories:
- A lift-and-shift of virtual or physical workloads into public cloud
- A partial or full refactoring/modernization of an on-premises application where the customer generally owns the source code
- A brand-new, born-in-the-cloud application or startup endeavor
All organizations want to understand how much they’d pay in each of the three clouds, based on monthly consumption, but I’d propose a more comprehensive method I call “The Three T’s of Cloud Selection.” The three T’s represent a necessary convergence of technology, skill set and cost, and all three must be balanced in order to move forward with one cloud platform over another. In this blog, I’m going to review how the Technology, Talent and Total Cost convergence can help you make a decision.
Technology refers to everything above the commodity cloud layer, or the general compute, storage and networking products across every single cloud provider. AWS has EC2, Azure and Google Cloud Platform (GCP) have virtual machines (VMs), and with the exception of some small differences, they largely function in the same manner. As a benefit to you, the customer, the three aforementioned providers view the commodity layer as a race to the bottom, slashing prices and offering Reserved Instances galore.
That’s all well and good, but I encourage you to focus on the rich tools and technologies below this layer. When you start exploring data analytics, machine learning (ML), artificial intelligence (AI), Internet of Things (IoT) and other non-commodity platform technologies, cloud providers begin differentiating themselves. It is critical to review these technologies against your desired business needs and rank the providers in strength accordingly.
Cloud talent in many markets is still hard to find, potentially expensive and difficult to retain. A system administrator in the organization will often have a cloud initiative thrown on their plate and told to execute without fully understanding the deployment models. This is why it is critical to evaluate the availability of talent internally, externally or through third-party consulting against the viability of the technology.
GCP, for example, is an incredibly powerful platform for building AI and analytics applications and counts data-driven companies like Spotify and Snapchat among its customers. Unfortunately, what little cloud talent there is remains focused in AWS and Azure. It is important to validate if you can build this expertise in-house or will need to look outside your organization.
3. Total Cost
Total cost represents the convergence of technology and talent and can be determined by asking the following questions for each platform you’re considering:
- How much are the commodity layer (compute, storage, etc.) and transformational layer (analytics, ML, etc.) consumption components on a monthly basis? In other words, what is going to be on my public cloud bill every month?
- How much will I pay to hire cloud talent, build it in-house via training or hire consultants to help me make the move?
- Who is going to take care of this once it is running, and what are the costs for automation and management?
CDW launched Cloud Consulting Services to work through the entire cloud lifecycle, from platform selection to implementation and management. Request a consultation to get started on your custom cloud solution.