Which Cloud Cost Model is Right for Me?
There’s an endless number of permutations of cloud pricing that you can consider. Which cloud cost model is right for you? There comes a double-edged sword of ultimate flexibility. Swiftly followed by the pain of not having one clear solution or package of services with a one price tag on it.
With the continuously changing economic environment, it’s a great time to get a firm grasp on your current state and which cloud cost model makes sense for your business.
In saying that however, there are a few standard cloud cost models that will fit the majority of business needs. This article will help you figure out if you could be leaving money on the table by not leveraging the cost model that best suits your requirements.
Cloud Cost Models and Pricing
AWS has evolved their cost models over the years with the needs of their thousands of customers over the years. The result is a number of cost models to fit any business looking to leverage the power of cloud computing services.
Here are some different cost models that AWS has to offer:
- Pay-As-You-Go
- Spot Instances
- Reserved Instances
- AWS Savings Plans
Pay-As-You-Go
The most straightforward of the cost models. There’s not additional complexity here, you just pay for the services you use as you go. This may work well for companies that are looking to test the waters with cloud computing. Maybe there’s non-production on-premise workloads that you’d like to run in an ECS container to dip your toe into the cloud. No need to sign up for long term commitment contracts. The resulting downside of this model is that when you start to scale up in any way, it will become expensive and fast!
Spot Instances
Spot instances have become a popular option for many businesses on AWS, in some cases providing up to 90% in savings over On-Demand prices. By utilizing the unused EC2 capacity that AWS has, you’ll be making huge savings while enjoying all the benefits of high-performance computing (HPC) at a fraction of the price. The downside however is that spot instances are not suitable for all applications. If your workloads don’t have space for downtime or extended periods of unavailability, then you’re best looking for other models with higher availability built in.
Reserved Instances
Reserved instances (RIs) provide another option for savings over On-Demand pricing. By reserving instances in advance, you’ll also be doing your Finance department a huge favour by giving them more predictability in the cloud billing coming through. If you know you’re going to need the capacity, RIs may be a great opportunity for you to make some savings on EC2 costs.
AWS Savings Plans
AWS has put together savings plans as a flexible pricing model to save costs under 3 specific areas:
- Compute savings – allowing you to save up to 66% on usage EC2, Fargate, and Lambda
- EC2 savings – allowing you to save up to 72% on EC2 spend
- SageMaker savings – savings for SageMaker, beneficial if you’ve incorporated SageMaker into your machine learning architecture
The only trade-off with AWS savings plans however is that it requires a 1-3 year commitment.
Small to Medium (SME) Businesses
As a small to medium business, it’s likely that you’re still in growth mode. Using available CAPEX to fund new revenue-generating initiatives. It’s probably safe to assume that your operational budget (OPEX) isn’t super high right now. This is where you can utilize cost control levers such as spot instances and Pay-As-You-Go to get the most out of your cloud services while not breaking the bank. Spot instances are great for smaller businesses that have yet to build that steady transactional flow where any downtime is disastrous. Until your business gets to that point, enjoy the cost savings from being able to tolerate some downtime.
Large & Enterprise Organizations
Large enterprises however have much different needs. If you’re managing cloud spend in an enterprise company (2000+ staff), it’s likely that your cloud spend is a lot higher and cloud services are now integrated into your critical business systems. At your size, any downtime can have a huge impact to revenue, with minutes of downtime translating to thousands or even millions in lost revenue. But that doesn’t mean that you can’t find ways to control your costs as well. Utilizing products such as RIs and AWS savings plans can add up to thousands and thousands in savings, especially at the scale of usage that you’re likely to be using as an enterprise organization.
You see, AWS likes guaranteed revenue (don’t we all). So as an enterprise company that has been running for a number of years and will continue to run in the same way and perhaps even scale up over the next few years. It’s safe to say that you’ll need those instances and compute power, so reserving cloud computing power makes total sense and will help bring your cloud service bill down significantly.
There’s No ‘One-Size Fits All’ Cloud Cost Model
But there are many options available to all sizes of organizations. If you’re looking to get a better handle on your cloud spending and trying to find the best cloud cost model to fit your business requirements – there are a number of options out there for you. Just make sure that you understand the needs of your business first. Knowing what level of availability is non-negotiable, what your usage is going to look like 1-3 years from now, and what your available budget looks like will all contribute to finding the right cloud cost model.
If you’re looking for help understanding all the tools, pricing models, and services available to fit your business needs, don’t hesitate to reach out to us! We’ve helped many clients in your shoes find the most cost-effective solutions that allowed them to save costs while continuing to grow their business.
