TL; DR: iGaming studios can use three cost optimizing strategies to improve their budgeting. Spot instances, Reserved instances, and Savings Plans all offer different approaches to pay for AWS server use. The exact discounts vary from strategy to strategy and are easier to grasp in comparison to on-demand pricing, which is the most expensive way to pay for AWS server space. Since the iGaming industry experiences fluctuating player rates, but also needs a reliable solution that is constantly active, using a combination of several cost optimizing strategies will be beneficial. 

A reliable server is the backbone of smooth multiplayer experience, so it comes as no surprise that the server fee is one of the biggest drivers of cost.  

Savvy iGaming studios will always look for ways to save on this long and important commitment, while cloud providers will use discount offers to attract clients to their server offerings. AWS has excelled at this game, coming up with several alluring cost optimization strategies.  

Today we will look at Spot Instances, Reserved Instances, and Savings Plans. 

Every game strives to be profitable, but within the iGaming industry the stakes are even higher. Cost optimization is crucial in iGaming due to the industry’s competitive nature and the need for continuous growth and user engagement. Operating on thin margins, iGaming companies must efficiently manage costs to consistently expand while also investing in technology, marketing, and user acquisition.  

Additionally, the iGaming industry faces fluctuating demand, with peak usage during certain events or times. Efficient cost management ensures that companies can scale their infrastructure dynamically without overspending during low-traffic periods.

This scalability is vital for maintaining performance and reliability. Lags are still a big problem and a popular complaint for modern day gamers, so being able to invest in scalable resilient servers gives iGaming companies a tangible competitive advantage. 

Regulatory and compliance requirements that are a staple of modern iGaming often impose additional financial burdens. By optimizing costs, iGaming companies can also ensure they have sufficient funds to meet these obligations while still investing in growth and innovation. From every angle, cost optimization is an important priority for iGaming studios to keep in mind, and cost optimization of server fees specifically can make a rather big difference for the overall budget.

Cloud-powered iGaming applications need servers to run on, but not every studio can afford to maintain their own. By far the most popular solution is to rent them from a cloud vendor, and Amazon is often the go-to candidate due to its versatility, scalability, security, and payment options.  

AWS EC2 (Elastic Compute Cloud) instances are virtual servers in Amazon’s cloud computing platform, providing resizable compute capacity to run various apps. You can pay for them on-demand or using Spot Instances, Reserved Instances, or a Savings Plan. Considering the continuous necessity of this service for the iGaming providers, it is advisable to be extra careful in choosing the payment strategy. 

AWS has an on-demand payment option, where users pay an hourly fee for their use of the server. Since you pay-as-you-go, this route might appear to be the most cost-efficient, but it is quite the opposite. While technically you only pay for the server uptime used, the hourly fee for the always available server space is pretty steep.  

Comparatively, the price of on-demand instances is the full 100%, and other payment types can be analyzed as a fraction of that fee, by their discount. This makes on-demand payment an extremely expensive choice for iGaming, but it is a solid investment for other small-scale deployments where you only need the server for a short time and do not want to buy a plan with more hours than necessary.  

Spot Instances, on the other hand, are a cost optimizing strategy. They function as a type of perpetual auction where various users apply bids for Amazon’s spare server space. It’s an affordable solution, but it packs some risks. AWS can reclaim the instances with a two-minute warning when it needs the capacity or when another user has outbid you. This pricing model is a good contender for individual deployment projects, but using it for the entirety of your iGaming app runtime would be its own type of gamble.    

Bottom line: 

• Pricing. Spot Instances offer the largest discounts (up to 90%) compared to On-Demand prices. 
• Use. Best for workloads that can handle interruptions, such as batch processing, data analysis, and CI/CD pipelines. 
• Availability. Can be interrupted, so they require applications to be designed to handle potential downtime. 

This strategy is more nuanced and sophisticated, allowing the user to specify the exact instances that they need. The main categories are type and scope, and the payment options split into All Upfront, No Upfront, and Partial Upfront, depending on how much you are willing to pay right away. While the All Upfront option will amass the best discount, even the No Upfront option will be cheaper than paying on-demand.  

There are three types of Reserved Instances.  

• Standard RIs offer the highest savings and are ideal for steady-state workloads. While they are less flexible, you can modify attributes such as instance size within the same instance family. 
• Convertible RIs provide slightly lower savings than Standard RIs but offer the flexibility to change instance families, operating systems, and tenancies over the term, accommodating evolving application needs.  
• Scheduled RIs allow you to reserve instances for specific time periods on a recurring schedule. They are ideal for predictable workloads that need to run at regular intervals. 

While convertible RIs with their flexibility seem to better fit the tumultuous nature of iGaming, there are ways to creatively combine several strategies to get what you want. For example, Wildlife Studios had great results from combining Reserved Instances with Spot Instances, assessing when to choose what via Cloud Analyzer.  

Reserved instances are also divided by scope into regional and zonal. Every AWS region is a physical piece of land where Amazon has their servers. Each region has several Availability Zones, and, depending on your preferences, you can either commit to a specific zone or to the region as a whole. 

• Regional RIs apply to any matching instance in the specified region. This means that they can be used across different Availability Zones within the same region, allowing for more flexible deployment and easier instance balancing. That being said, quite like Spot Instances, Regional RIs do not reserve capacity and availability.  
• Zonal RIs do reserve capacity in a specific availability zone, operating in a smaller area, but providing reliable deployment. If your iGaming app requires high availability and guaranteed capacity in a specific zone, but does not necessarily need regional flexibility and resilience, this would be a more confident choice.  

When you purchase an RI, AWS applies the discount to any matching instance usage within your account. If you have a Regional RI, it will automatically apply to any instance that matches the criteria (instance type, region, etc.), helping maximize the discount usage. For Zonal RIs, AWS reserves capacity in the specified availability zone and applies the discount to instances launched in that zone. 

While most AWS customers choose their region based on proximity to themselves or to their end users, it would be wise to remember that regions are all differently priced and have different latencies. A data-driven iGaming studio will clearly define their priorities (data transfer, latency, storage, or others) when making their choice. 

Bottom line: 

• Pricing. Reserved Instances offer significant discounts (up to 75%) compared to On-Demand pricing when you commit to using AWS resources for a one-year or three-year term. 
• Use. Ideal for predictable workloads with steady-state usage, such as web applications and databases that require consistent and reliable performance. 
• Availability. RIs are always available and are not subject to interruption.

Introduced in 2019, Savings Plans are the newest addition to AWS Cost Optimization Strategies. With Savings Plans, customers commit to a consistent amount of usage (measured in dollars per hour) over a 1- or 3-year term, in exchange for discounted rates. 

Savings Plans offer more flexibility than Reserved Instances in terms of payment and reservations. RIs require upfront payment for a specific instance type in a region for a 1- or 3-year term, offering a fixed discount. Savings Plans offer discounts on usage across a wide range of AWS services, with flexibility to apply commitments to any instance type, region, or service, without upfront payment. Unlike RIs, Savings Plans adapt to usage changes automatically, providing more flexibility and simplicity in cost management. 

Savings Plans are a strong option if your iGaming platform has varying usage patterns or if you need flexibility across different instance types and services. They offer a balance between cost savings and flexibility without the need to manage specific reservations. 

There are two types of Savings Plans suitable for iGaming Apps: Compute Savings Plans and EC2 Instance Savings Plans. 

• Compute Savings Plans offer a flexible pricing model for AWS services, providing discounts in exchange for a commitment to a consistent amount of compute usage. These plans cover a broad range of AWS services, allowing customers to commit to a wide selection for different goals. 

• EC2 Instance Savings Plans provide discounts specifically tailored to EC2 instance usage. Customers commit to a specific instance family in a region for a 1- or 3-year term, and in return, receive discounted rates for the selected instance type. 

Both types of Savings Plans offer significant cost savings compared to standard on-demand pricing, with the flexibility to adapt to changing usage needs.  

AWS also offers a third option, AWS SageMaker Savings Plan, but it relates more to the use of SageMaker specifically. SageMaker can be used to optimize and improve your iGaming practices through separate machine learning tasks, like training responsible gaming interference models.  

Bottom line: 

• Pricing. Offer significant savings (up to 72%) while providing more flexibility than RIs. They apply to a broader range of usage across instance types, regions, and even AWS services like Lambda and Fargate. 
• Use. Ideal for environments where workload patterns might change, ensuring you still benefit from savings even if your instance usage changes. 

The ultimate selection of the right cost optimization strategy must be informed by your priorities and plans, but some general truths still apply. For example, for iGaming apps, a combination of Spot Instances and Savings Plans would generally add up to a good hybrid cost optimization strategy. 

Spot Instances are ideal for non-critical workloads or tasks that can tolerate interruptions. However, spot instances can be used for rendering jobs and testing environments. iGaming apps also often have fluctuating demand, making Spot Instances a cost-effective choice during periods of low traffic.  

Since iGaming apps may also have varying usage patterns, Savings Plans provide flexibility by offering discounts on a broad range of AWS services, allowing you to apply commitments as needed. This flexibility aligns well with the dynamic nature of iGaming workloads. 

Reserved Instances could also be beneficial if you have predictable usage patterns for specific instance types. However, since iGaming app usage tends to be more variable, the flexibility provided by Savings Plans might be more advantageous. 

Spot Instances, Reserved Instances, and Saving Plans all have their merits, and your iGaming studio will likely benefit the most from using more than one optimization strategy.

Regardless of which route you choose, as long as you choose a strategy you will benefit from a discount and arrive at a more reasonable total cost than if you were using on-demand payment.

Innovecs Games has been helping our partners to improve their iGaming content for over a decade. We are proud owners of Amazon Consulting Partner Status and would be happy to discuss AWS-powered iGaming optimization with you. Feel free to contact us with any questions or reach out for a chat. Maybe we will feature your expertise in our next blog!