PCG logo

4 Tricks for Cloud Cost & Performance Optimization


For many companies, their cloud journey begins with an initial assessment of the potential cloud costs. This involves checking whether the migration of various workloads to the cloud will be cost-effective for the IT budget in the long term. Once this has been completed and evaluated as promising, the actual migration follows. Optimization potential already plays an important role here: how can the existing IT infrastructure be optimized in the cloud in terms of performance, power and costs?

Once the transition to the cloud is successfully accomplished, the subsequent phase involves the operation of the environment. The key to success with your cloud environment lies in continuous analysis and ongoing optimization. This becomes particularly evident in the operation of your cloud infrastructure. Even though migration might seem like a monumental task in your cloud journey, experience has shown that it's the operation phase that demands the most attention and care.

A provider of managed services for public cloud environments, often referred to as a Next-Gen Managed Service Provider, can assist you in this regard. They can take over the operational responsibilities and enhance optimization through automation.

In this article, we're giving you a glimpse into four areas of potential optimization. These are aspects you should examine not only at the outset of your cloud journey but also continuously in order to improve your IT KPIs (Key Performance Indicators).

IT areas as enablers for corporate success

A company's IT environment always has to adapt to the current challenges and requirements of the business environment. For example, if a company is working on an internationalisation strategy, the IT should also be able to withstand a global roll-out. If employees are enabled to work from their home offices, the IT must enable access in the same way as via virtual workstations within the specified timeframe.

In short, the IT strategy follows the business strategy in many aspects and therefore needs to be flexible and efficient.

In the past, as well as in the heyday of data centres, this flexibility was often implemented with an investment in partly unused server resources. These were available in the event that new requirements needed to be met quickly and current IT scenarios needed to be mapped. The desired flexibility was linked to investments and long-term contracts, often without knowing whether these would really be needed or ever used. The bottom line: Efficient in terms of performance and availability, but not in terms of cost.

A perfect match: Public Cloud and Optimization

The requirements for flexibility, availability and efficiency can be mirrored in cloud environments. These three characteristics, along with cost transparency and scalability, are among the most frequently cited added values of the public cloud.

In addition to handling the current challenges of an IT environment, the optimization factor also plays an important role when using a public cloud. As is generally known, payment is based on the "pay-per-use principle", i.e. you are charged for the cloud services you have used in a month. Accordingly, it is a high priority to design the cloud environment as closely as possible to the requirements in order to avoid paying for unnecessary capacities.

In addition, within the public cloud of your chosen hyperscaler, you also have nearly unlimited scalability of your resources available. For instance, if you operate a consumer goods retail company with an online shop, it should be able to handle all customer and purchase requests during peak periods. The scalability of your cloud environment is therefore essential for success.

4 Optimization Tricks You Should Know About

To recognize and apply these optimization opportunities for your own public cloud environment, it's important to know the corresponding actions to take at the right points in time. We'll now introduce you to four of these tricks, or optimization techniques.

1. Reserved Instances
AWS Reserved Instances (Amazon Reserved Instances) are virtual servers running on Amazon Web Services (AWS) Elastic Compute Cloud, or EC2, and Relational Database Service (RDS). Companies purchase the instances, which are available in various levels of compute power, at contract prices plus hourly rates.

IT administrators can purchase Reserved Instances directly from Amazon Web Services or from other administrators in the AWS Reserved Instance Marketplace. Reserved Instances purchased through the Marketplace offer cost savings on compute power with shorter remaining terms than new instance types. The contract terms are always one or three years.

To create a Reserved Instance, companies can freely choose from the following categories according to their requirements:

  • one of the different instance types of AWS and
  • an associated operating system,
  • a runtime,
  • a tenancy specification, and
  • a region, and an Availability Zone.

The advantage of using Reserved Instances is that they do not have to be tied to a specific purpose, neither at the time of booking nor during subsequent use. The Reserved Instances are therefore not linked to a single VM for a specific purpose. Rather, it is a specific performance contingent that you can secure for the long term (one or three years) at a discounted price and use, for example, for IT scenarios that cause a relatively predictable load.

Expert tip: Map the general load of your IT, i.e. the recurring resources in the cloud, using Reserved Instances. Here you can benefit from enormous discounts - the longer the contract term, the higher the discount, of course.

2. Savings Plans
Amazon Web Services' Savings Plans model was introduced in 2019 as a flexible pricing model for discounted use of AWS cloud resources.

Savings Plans have many parallels to Reserved Instances, which were already in use long before 2019. The decisive difference between the two optimization measures is that Reserved Instances offer a discounted price on a specific usage capacity, whereas the price discount of a Savings Plan is based on a spending obligation.

The cloud costs incurred can be reduced by up to 66 to 72 percent with Savings Plans from AWS. In return, the customer must enter into a contract term of one or three years for AWS services.

There are currently three different types of AWS Savings Plans available. These three types are:

  • Compute Savings Plans (for use with Amazon EC2, AWS Lambda, and AWS Fargate),
  • EC2 Instance Savings Plans (for EC2 usage),
  • Amazon SageMaker Savings Plans (for use with the Amazon SageMaker cloud machine learning platform).

It is easy to monitor and manage Savings Plans using budgets, budget alerts, and performance reports. These can be set up and customised in your organisation's AWS account. It is important to note that AWS Savings Plans are not capacity reservations. However, on-demand capacity reservations can be used to reserve capacity and achieve discounted pricing within a Savings Plan.

Expert tip: Companies are now increasingly relying on Savings Plans as opposed to Reserved Instances. Here's an example to explain why: In the first year, you migrate one of your applications to the cloud and book Savings Plans for it. In the second year, you optimise this application towards a cloud-native approach and can still continue to use the Savings Plans already booked. This is possible because you run the application within the same Savings Plans type: former Amazon EC2 becomes AWS Lambda, both of which fall under Compute Savings Plans. So you can evolve within the cloud environment and optimise your environment, but still continue to benefit from your discounting.

3. Spot Instances
Spot Instances are a type of instance that can offer you significant cost savings, but they should only be used for specific business cases and use cases in the public cloud.

Spot instances use basically free EC2 capacities, which can be offered and used at a very high discount. The hourly price that is charged for the hourly use of a Spot instance is referred to as the spot price.

The spot price is available for all instance types in all regions and availability zones and depends on the long-term supply and demand of Amazon EC2 instances. For example, your Spot Instance will run whenever capacity is available and the maximum price per hour for your request exceeds the spot price. You will then be charged the discounted Spot price.

Spot instances are a cost-effective choice, provided you are flexible in the timing of your applications and can tolerate interruptions.

Spot instances are suitable, for example, for data analysis, batch processing jobs, background processing, and calculations.

4. Right-Sizing
Right-sizing means finding the optimal cloud configuration to maximise your performance at the lowest cost. In other words, you need to make sure you're not wasting money on something you don't need. With right-sizing, you can ensure that you get the most out of the cloud.

Right-sizing has several benefits, the most obvious of which is saving money. Many businesses choose to move to the cloud because of the potential cost savings. So why not make sure you maximise your savings? Another benefit of right sizing is the elasticity it gives you to adjust your infrastructure to your workload as needed. For example, if you need to increase capacity for a few days and then decrease it again, you can continually adjust your infrastructure to meet your needs so you're not locked into a specific size. That means your cloud shouldn't have a "set it and forget it" mentality. While you don't have to monitor it constantly, it's good to make sure your infrastructure is right-sized to balance your workload needs and your cost savings.

Unlike other optimisation measures, right-sizing can really be an option for anyone looking to maximise performance while saving money.

Expert tip: setting up right-sizing for your cloud infrastructure can be done by a qualified partner. This will ensure that all your requirements are implemented. You can then leave the continuous review to your IT team.

Services Used

Continue Reading

VPC Endpoints: Explanation and Cost Comparison

In this blog post, I will explain what VPC Endpoints actually are. What are the differences between VPC Gateway and Interface Endpoints? What benefits do VPC Endpoints offer? Also, a cost breakdown of how much the VPC Endpoints cost, as well as a comparison of the costs to a NAT Gateway.

Learn more
Case Study
Generative AI
Using AI to Transform Marketing for Small Producers

A detailed case study on Farsk.ai, a Swedish startup leveraging AI to help small producers market their products. It covers the challenges, AI-driven solutions, and the positive impact on user engagement and sales.

Learn more
Case Study
With FSx Netapp for ONTAP to the Oracle database on AWS

The goal was to migrate the databases and associated applications to the AWS Cloud while maintaining high standards of performance, robustness and security.

Learn more
Pain-free migration from Amazon Workdocs

Amazon WorkDocs is being discontinued in 2025. This article explores your options for alternative file storage and collaboration solutions and outlines the right steps to ensure a smooth transition for your business.

Learn more
See all

Let's work together

United Kingdom
Arrow Down