Cloud Cost: 4 Cost Models and 6 Cost Management Strategies

What is the Cost of Cloud Computing?

Cloud computing costs vary greatly between different services and vendors. While the majority of organizations can accurately estimate the costs of setting up and maintaining their own IT infrastructure, many cannot accurately estimate the costs of shifting their operations to the cloud.

Cloud computing offers greater flexibility and often an improvement in efficiency. Cloud-based on-demand resources enable organizations to quickly scale up and down, and automation capabilities provide a higher level of efficiency. However, many organizations express uncertainty as to how to accurately estimate the costs of dynamic (and often ephemeral) cloud resources.

To properly estimate, control, and optimize cloud computing costs, organizations need to understand the parameters that impact costs, assess their needs, design a strategic plan, and continuously monitor and optimize usage and billing. To help organizations gain control over costs, cloud vendors offer several pricing models and cost management and optimization tools. There are also third-party tools that can help optimize cloud costs.

 

In this article, you will learn:

 

Breaking Down the Cost of Cloud Computing

Main Components of Cloud Costs

There are three main components that determine the cost of cloud computing services:

  • Compute—most cloud providers offer a range of compute instance types, each offering a certain amount of CPU resources, memory, and in some cases, specialized hardware such as fast networking or graphics acceleration. The customer pays according to the number and types of instances used and the duration they are used.
  • Networking—most cloud services bill customers according to the volume of data transferred into the cloud service (ingress), out of the cloud service (egress), or both. There may be special charges for virtualized network services such as static IPs, load balancers and gateways.
  • Storage—cloud providers offer storage as a service. For elastic storage services, customers pay by GB-month of storage actually utilized. For managed storage services, such as managed disks attached to compute instances, customers pay for an entire storage volume, regardless of the amount of storage utilized on the volume.

Calculating Cloud Costs vs. Traditional Infrastructure

Here are the three main categories of costs typically associated with setting up and maintaining on-premises infrastructure:

  • Capital costs—server software, licensing, and hardware, as well as network infrastructure, storage environments, and backup systems.
  • Operational costs—including support for server and network infrastructure, as well as storage warranty, data center facilities, existing system administration labor costs, and IT staff training and turnover.
  • Indirect business costs—including unplanned and planned downtime.

Usually, for every dollar spent on infrastructure upgrades, organizations can expect to spend approximately $2 on managing, maintaining and securing the extended infrastructure.

When calculating their overall cloud migration and adoption costs, organizations need to first conduct a comprehensive audit of existing IT, including both direct and indirect expenses. For example:

  • Direct costs—include hardware, software, management, maintenance, and staff, as well as any physical facility. Direct costs are usually straightforward and easy to estimate.
  • Indirect costs—include loss of productivity which might be a result of several factors like server downtime, loss of customer trust, and reputational damage. Indirect costs are typically more complex to predict and estimate.

 

5 Cloud Cost Models

Here are several common cost models used in the cloud, which you can combine depending on your needs. 

Pay-As-You-Go

In this model, cloud services are billed per actual usage. Cloud services may bill for utilization of computing power, storage, networking, or other resources. The advantage is that you only pay for actual usage, and can scale down resources when needed. The downside is that as you add more resources to your cloud deployment, ongoing costs can quickly skyrocket.

Prepaid/Fixed Subscriptions

In a subscription-based model, cloud customers pay for services upfront. Subscription prices deliver a predetermined package of services for a specified time. The longer the period, the lower the price. 

Subscription pricing is common for cloud services that combine multiple hardware and software elements, like platform as a service (PaaS) and software as a service (SaaS). Most cloud providers also offer subscription-based pricing for customers with high spend, allowing them to enter into a corporate discount plan, where they commit to a certain level of cloud spend and receive a discount on some or all of their cloud services.

Reserved Instances

Reserved instances allow companies to commit to cloud resources for a long period of time, typically 1 or 3 years. The longer the discount, and the more the company is prepared to pre-pay at the beginning of the period, the greater the discount. A three-year term is usually the most cost effective. Cloud providers typically offer discounts of 50-75% compared to pay-as-you-go rates for reserved instances with the same capabilities. 

Reserved instances are suitable for steady state loads and long running systems. However, organizations should not use reserved instances for peak loads. Instead, reserved capacity should be used for core components of the system, and additional capacity required during peaks should be handled using pay-as-you-go or spot instances (see below).

AWS Savings Plan

Similar to reserved instances, Savings Plans are a flexible pricing model that allows organizations to enjoy lower than on-demand pricing, in exchange for a one-year or three-year specific usage commitment. The commitment is expressed in terms of spend per hour on Amazon services.

AWS offers three types of Savings Plans: 

  • Compute Savings Plans - apply to all usage of Amazon compute services usage, including EC2, AWS Lambda and Fargate. 
  • EC2 Savings Plans - applies only to usage of Amazon EC2 instances.
  • SageMaker Savings Plans - applies only to SageMaker usage.

Savings plan offer three payment methods: 

  • No upfront - does not require an upfront payment, bills customers according to actual usage each month. This grants the minimal savings plan discount.
  • Partial upfront payment -  with this option, more than half of your contract is prepaid and the rest is billed monthly, which grants an additional discount.
  • Full upfront payment - the full commitment is paid upfront, which grants the deepest discount.

Spot Instances

Spot instances are usually the lowest-cost computing option, offering discounts of up to 90% compared to pay-as-you-go rates. Spot instances are used by cloud providers to sell off spare capacity. The discount comes with a catch—spot instances can be interrupted at very short notice. 

Ordinarily, spot instances can only be used for workloads that are stateless, fault tolerant, or processes that can be stopped and restarted. Cloud optimization technology like Elastigroup from Spot by NetApp can help you leverage spot instances for demanding, mission critical workloads as well. 

Learn more in our detailed guide to cloud cost models

 

6 Cloud Cost Management Strategies

Cloud cost management (also known as cloud cost optimization) enables businesses to understand and centrally manage the costs associated with cloud technology, maximize the return on investment in cloud technology, and increase efficiency.

Cloud resources are highly dynamic and may be distributed across multiple locations, services and cloud providers. Easy scalability and deployment are compelling advantages of the cloud, but they also make it easy for IT, devops or engineering teams to run resources without understanding the short-term or long-term cost considerations. The cloud makes it difficult to achieve visibility, and enforce decisions and policies about cloud costs. A cloud cost management strategy can help organizations plan for future consumption and costs of cloud services.

Today the majority of companies use multiple clouds, which makes it critical to adopt multi-cloud cost management strategies. These take into account the costs of different public cloud providers and enable cost management of multiple clouds on one pane of glass.

Here are some of the strategies companies can use to manage cloud costs:

  • Budget control—as a first step, organizations must establish budgets for cloud services and ensure teams are aware of them, and cannot exceed the budget for their specific project. 
  • Right sizing—ensuring compute instances, storage volumes and other services are provisioned at the level actually required by the business. It is very common for cloud resources to be provisioned and not fully utilized.
  • Autoscaling—scale resources up and down dynamically according to application demand, ensuring you only pay for extra cloud resources during peak usage.
  • Scheduling—many cloud services are not required 24/7, and can be scheduled to shut down when not needed. For example, services used by a US-based team can be shut down outside US business hours.
  • Detecting unused resources—compute instances, storage volumes, load balancers, snapshots, and many other resources can easily be created and forgotten. Organizations must be able to scan their cloud deployment for unused resources and delete them to conserve costs.
  • Smartly applying discounts—discounted pricing models like reserved instances and spot instances can significantly drive down cloud costs, but they must be used appropriately. Tools like Cloud Analyzer from Spot by NetApp can help you understand which of your applications and workloads is best suited for discounted price models.

Related content: read our guide to cloud savings

 

Cloud Cost Optimization Tools

Unlike traditional IT systems, which were well known and static, cloud cost management cannot easily be managed by spreadsheets and manual lists. Automated tools are available which can retrieve metrics from APIs, report on cloud consumption and costs, and make changes to services as necessary. There are two main types of tools—first party tools provided directly by the cloud provider, and third party tools from external vendors. 

Related content: read our guide to cloud cost optimization (coming soon)

First Party Tools

All public cloud platforms provide cost management tools. These tools are highly integrated with the cloud platform, and are available out of the box without a special deployment effort. Some of these tools are free to use, while others are billed on a pay-per-use model. 

In many cases, these basic tools are the fastest way to start managing your cloud costs. However, they have several limitations, which are addressed by third party tools:

  • Limited functionality—most first-party tools are limited in their ability to identify wasted costs and maximize savings using multiple cost models.
  • Limited to one vendor—most first-party tools only work with one cloud provider, and are not suitable for organizations that require multi-cloud cost management.
  • Conflict of interest—a cloud provider is, in the end, interested in maximizing profits. On the one hand, cloud providers do want to help clients run applications in a cost-effective manner, to increase usage and retention. On the other hand, cloud providers want to maximize consumption of cloud services and may not always offer the optimal solution for the cloud customer.

Third Party Tools

Third-party cost management tools can address the functional limitations of first party tools, and commonly provide multi-cloud cost management. Most of these tools are built to reduce cloud costs across a variety of cloud services and workloads, providing clear return on investment (ROI). 

However, organizations should thoroughly evaluate the functions of cloud management tools across clouds, understanding how their capabilities map to the specific cost models of each cloud, and evaluating their ability to detect, recommend and automate cost optimizations.

 

Optimizing Cloud Costs with Spot by NetApp

Spot by NetApp not only offers comprehensive visibility into cloud spend, but actively manages and optimizes compute deployments with advanced automation that ensures customers get more out of their cloud for the lowest cost possible. In addition to spend visibility Spot by NetApp delivers dramatic cost reduction and increased efficiency for:

Traditional, stateless and stateful workloads

Kubernetes and container workloads

Big data on Kubernetes

RIs and Savings Plans

Managed Service Providers (MSP)

 

Learn More About Cloud Costs

Leveraging Cloud Computing Pricing Models for Greater Cost Efficiency

With over 60% of all cloud costs attributable to compute, focusing on compute infrastructure spend should be of the utmost priority. Get comfortable with all the various public cloud pricing models so you can successfully leverage them all for greater cloud cost efficiency.

Read more: Leveraging Cloud Computing Pricing Models for Greater Cost Efficiency

Cloud Cost Optimization: 15 Ways to Optimize Your Cloud

Cloud cost optimization ensures the most appropriate and cost efficient cloud resources are allocated to each workload or application. Discover 6 cloud cost optimization tips that can help you build a cost-optimized cloud environment and 9 free tools that can reduce your cloud costs.

Read more: Cloud Cost Optimization: 15 Ways to Optimize Your Cloud (coming soon)

Cloud Savings: 9 Ways to Save in the Cloud

There are two aspects to cloud savings. First, using cloud computing can in itself generate cost savings for organizations. Second, once already in the cloud, organizations need to control and optimize cloud costs. Discover 4 ways the cloud itself can generate savings for your organization, and 5 ways to optimize cloud costs once already in the cloud.

Read more: 9 Ways to Save in the Cloud