Organizations optimizing cloud costs using rate optimization often trade risk, in the form of commitment lock-in or flexibility—the ability to adjust commitment coverage up or down, aligned with usage, for savings. To help mitigate this risk, Amazon Web Services (AWS) recommends deploying a rolling Savings Plans strategy. This advanced commitment management strategy involves purchasing multiple overlapping Savings Plans with staggered start dates. Organizations can improve coverage and mitigate risk by distributing their overall commitment pool across multiple equally sized smaller commitments and purchasing them regularly, such as quarterly.
Even though the rolling Savings Plans approach requires more effort, it allows businesses to adjust their commitment portfolio more frequently, aligned with dynamic and sometimes unpredictable usage patterns.
ProsperOps, an AWS Advanced Technology Partner available on the AWS Marketplace, automates a rolling Savings Plans strategy for AWS customers. Using its proprietary algorithm, ProsperOps continuously analyzes usage and optimizes commitments based on changes, reducing cloud costs and mitigating risk.
Challenges of Manually Managing Rolling Savings Plans
Rolling Savings Plans provide more flexibility, but executing this strategy manually is nearly impossible in an enterprise cloud environment. To successfully deploy the rolling Savings Plan strategy, your FinOps team would need to:
- Analyze usage patterns hourly
- Estimate or forecast future usage*
- Account for upcoming engineering changes, product launches, and/or optimizations
- Purchase and/or renew existing Savings Plans alongside other commitments
All done at the optimal time, but requiring them to work around the clock, 24×7. Keep in mind, Savings Plans are long-term commitments of one or three years, so missing, or over-committing within an active adaptive laddering strategy could have a reverberating impact on your savings outcomes, not to mention reset time and effort invested to build your ladder to that point.
Using ProsperOps Automation To Build an Adaptive Savings Plans Ladder
Building upon the rolling Savings Plans concept, ProsperOps introduced Savings Plan Adaptive Laddering™ for AWS Compute. Adaptive Laddering automates AWS’s recommended laddered Savings Plan approach by dynamically managing the purchase and renewal of AWS Compute Savings Plans, covering Amazon Elastic Compute Cloud (Amazon EC2), AWS Lambda, and AWS Fargate. Our algorithm continuously analyzes usage patterns to ensure optimal coverage, adjusts commitment levels to reduce the risk of overcommitment, and adapts to dynamic compute usage patterns to create an optimal, ongoing ladder of Savings Plan commitment. In short, adaptive laddering helps FinOps teams achieve optimal cost savings while minimizing the commitment lock-in risk of savings plans.
With ProsperOps, customers no longer need to manually monitor their commitments or make educated guesses about future usage. Instead, the platform’s automation ensures continuous optimization without human intervention within the guardrails set by the FinOps team.
Let’s look at a simple example to see how this works.
Drawbacks of a single monolithic Savings Plan approach
A single monolithic 1-year Compute Savings Plan (CSP), shown in green, is batch purchased to cover ~90% of usage. After about 3 months of steady usage, an unanticipated activity occurs (due to optimization or reduced consumption), and usage drops below the CSP commitment amount. Because this is a 1-year single monolithic CSP, there is no recourse except to let a portion of the commitment go unutilized until it expires at the end of the 12-month period.
A single annual CSP fails to adjust as usage drops, leading to underutilization.
AWS has provided guidance on the drawbacks of purchasing a single, annually renewed Savings Plan to cover target commitment levels. While commitment-based discount programs can deliver significant savings, especially in the typical and desirable business growth scenario, they also introduce risk. If usage unexpectedly and sharply declines, falling well below the committed levels, the value proposition of cloud computing’s pay-as-you-go model is undermined. In such cases, expenses may no longer correlate with actual usage, eroding cost efficiency.
Leveraging automation to implement rolling Savings Plans
Adaptive Laddering introduces flexibility by purchasing small rungs or batches of Savings Plan commitment over time (depicted in varying colors), such that a rung of Savings Plan expires every month (and up to four times monthly, depending on volume). Unlike the rigid single monolithic Compute Savings Plan, Adaptive Laddering builds a commitment pattern that adapts to usage changes and thus allows for more aggressive coverage. As usage changes, each rung expiration becomes a decision point—you can let the commitment fall off or recommit to the rung. ProsperOps automation aligns with AWS’s recommended Rolling Savings Plans to mitigate commitment risk and enhance discount coverage.
Staggered Savings Plan commitments align with usage changes, improving flexibility and reducing risk.
How ProsperOps Implements AWS Savings Plan Adaptive Ladders
Once configured and implemented, ProsperOps Savings Plan Adaptive Laddering begins ingesting telemetry from the AWS environment (i.e., cost and usage data) on an hourly basis. ProsperOps continuously analyzes this data to calculate rung size and purchase frequency. ProsperOps’ high-frequency laddering algorithm is intelligently aware of other portfolio discount instruments and normalizes and isolates applicable usage for the adaptive ladder.
Prevailing discount rates per term, along with current and historical usage, enable our platform to determine the optimal coverage point, which is recalculated on an ongoing basis and aligned with each new commitment purchase. Over time, ProsperOps adjusts commitment purchases, rung size, coverage targets, and frequency according to usage patterns.
This ongoing analysis of real-time data, factoring in remaining unutilized commitment and recent fluctuations in usage, helps organizations optimize cloud costs while also avoiding Commitment Lock-in Risk. When usage drops and commitment coverage is high, there is a risk of overcommitment. On the other hand, when usage spikes or increases relatively fast, it is risky to increase commitment coverage without having a safe way to reduce coverage if usage tapers off or declines.
Results with ProsperOps Automated Adaptive Laddering
The following examples illustrate results customers can achieve with ProsperOps automation in various real-world scenarios.
Customer overview
- $1 million in starting monthly usage
- 27% discount rate
- All 1-year discounts
Scenario 1: Annual seasonality in a slightly growing environment
On $1,000,000 in monthly usage, the difference in performance between a daily-executed Savings Plan Adaptive Ladder (SPAL) and quarterly SPAL is $42,651 per month, or $511,817 per year (Supporting data in the seasonal tab here) in an environment that’s slightly growing with annual seasonality.
| Daily SPAL | Quarterly SPAL | Delta | |
| Min ESR | 5.11% | -13.51% | 18.62% |
| Average | 17.51% | 13.25% | 4.27% |
Scenario 2: Weekend cyclicality in a slightly declining environment
On $1,000,000 in monthly usage, the difference in performance between a daily SPAL and quarterly SPAL is $96,357 per month, or $1,156,290 per year in a weekend cyclical environment declining 7.5% month over month.
| Daily | Quarterly | Delta | |
| Min ESR | 20.73% | -0.26% | 20.99% |
| Average | 24.74% | 15.11% | 9.64% |
Conclusion
AWS pioneered the idea of rolling Savings Plans to provide customers with more flexibility in commitment management. With ProsperOps automating this approach, organizations can now optimize costs with even greater savings and efficiency.
ProsperOps Savings Plan Adaptive Laddering for AWS Compute offers a breakthrough approach to automate cloud cost optimization. Organizations benefit from AWS’s flexible commitment structures while leveraging ProsperOps’ automation to remove complexity, reduce commitment risk, and maximize savings. The result? Cloud financial management that is smarter, faster, and fully optimized—without requiring a dedicated team to adjust commitments continuously.
In addition to Savings Plans, AWS provides Reserved Instances (RIs) that offer a discount compared to on-demand pricing. Based on your workload and usage patterns, ProsperOps automation will optimize all AWS discount instruments (Savings Plans and Reserved Instances) to maximize savings and reduce the commitment lock-in risk.
You can learn more about ProsperOps on the AWS Marketplace and schedule a call with their FinOps experts for a free cloud savings analysis.
