In this blog article I share 7 ways with which you can start optimizing your Azure costs today. In addition, I also share 7 options that Microsoft has for offering Azure credits, so that you can reduce your costs even further.

Cost optimization is a business-focused, continuous discipline to drive spending and cost reduction, while maximizing business value. You should treat cost optimization as a process, rather than a one-time activity. When taking advantage of the various tools in Azure that provides cost optimization recommendations doesn’t have to be a time consuming task. Below are 7 ways to optimize your Azure costs today:

Managing and monitoring spending is facilitated from Azure Cost Management. Here you can specify budgets, set alarms and view per resource, resource group, subscription, tag, etc. where which costs come from. Because costs are often allocated to different departments, you can also make a distinction based on tagging. In this way you can immediately inform the relevant budget holders about his/her costs and have a targeted conversation about this. Being aware of costs in relation to use, and recognizing the patterns for increases and decreases of costs, is the first and most important step in cost optimization.

Obviously, you save costs if you turn off resources when you are not using them. Virtual machines, ExpressRoute circuits, and other resources can be identified by Azure Advisor, which gives you recommendations on which resources to shut down, and how much you would save. You can also apply the concept of “snoozing”, which means turning resources on during usage hours, but turning them off outside of usage hours. This is easy to automate with the help of Azure Automation, or embedded solutions within Azure services, for example Scaling Plans for Azure Virtual Desktop.

Right-sizing is not a service or tool, neither can you access it from the Azure portal. Right-sizing is a process in which the goal is to find the “perfect fit”, or “right size” for a workload, so that costs can be reduced without compromising performance requirements. Or, improve performance to a required level, whilst minimizing the extra costs. It is about finding the most optimal size in relation to needs and requirements. Microsoft offers a variety of tools to support your right-sizing process, such as Azure Advisor, Azure Monitor and Insights, Cost Management and Billing, etc. If you’re considering migrating to Azure from your data center, Azure Migrate’s performance-based evaluation can help you right-sizing from the start.

Right-sizing should be applied especially in migration scenarios from on-premises data centers to Azure. In Azure there really is no reason to apply overcommitment and/or over-provisioning, you provision exactly what you need, when you need it, and you pay for what you use. An additional advantage is that you do not have to commit to a certain generation of hardware (CPU family or chipset), for the duration of the hardware depreciation. Instead of having to spend three to five years with the same hardware generation, you can enjoy the latest generations that Microsoft Azure offers with a few mouse clicks away.

Also, consider Spot Virtual Machines, which allow you to purchase unused computing power with significant cost savings. This is a very cost-effective way for workloads that can handle interruptions and don’t need to run within a specific time frame.

If you are an organization that has licensed their current on-premises workloads with Software Assurance, Microsoft offers the option to bring these licenses into Azure. You then no longer pay for the licenses within the consumption model, so that you do not have to write off the purchased licenses in one go. This can make a significant difference if you want to migrate your on-premises workloads to Azure.

Dynamic upscaling and downscaling can of course also be cost effective. By working with autoscaling, resources become available when you need them, and are deleted when you no longer need them.

Besides virtual machines, there are many more compute services that Azure offers. Azure offers many ways to host your code. Most compute services require less management burden than virtual machines and therefore the total cost of ownership can be much lower. You should operate more cost efficiently by selecting the right compute service for your application, if you can. 

Microsoft has written an extensive article about finding the right compute service.

You can get a discount of up to 72 percent over pay-as-you-go pricing on Azure services, if you make a longer commitment to use certain instances. These reservations can be made for a one- or three-year term. It’s strongly advised to apply reserved instances after considering the optimalization options above, in order to achieve the maximum cost advantage. Reserved instances are bought over the entire contract period, assuming you run the workloads 730 hours a month. Shutting down these resources will not affect the costs anymore, after buying a reservation. It may therefore be cheaper not to make reservations and to work with “snoozing” or “autoscaling”, or a mix of these. On the other hand, in some cases it is already attractive to purchase a reserved instance, even if the instance is only used for a few months. It is recommended that you always calculate this for yourself.

Reserved instances are not limited to virtual machines only. Microsoft offers reserved instances across a variety of products and services within its portfolio, including databases, App Services, and various software plans.

In addition to the 7 ways of optimizing Azure costs, Microsoft provides 7 ways to help you save money through Azure credits:

These credits are spending caps, because they are reduced if they are used to purchase resources that are not offered for free, similar to gift cards from a retail store. 

In this blog article I described 7 ways how you can start cost optimization within Azure today. As icing on the cake, I have listed 7 options that Microsoft offers to reduce your costs even further, using Azure credits. When embracing consumption-driven models, in which you pay according to use, cost control is very important. Instead of periodic investment and capacity plans for the replacement and delivery of infrastructure components, cost management within a cloud context should be thought of as a continuous process. Stop robbing yourself, and adopt the many tools that Microsoft provides to perform good cost management.