Skip to main content

What are you looking for?

Explore our services and discover how we can help you achieve your goals

Client Workspace

Savings plans, reserved instances and committed use discounts: how much cloud usage to commit

Commit only the cloud usage you are sure of: cover the stable floor of the bill with commitment discounts, add more as the baseline proves itself, and leave the variable top on demand. Netbase sets the level by usage stability, matching resource-based reservations to locked workloads and spend-based savings plans to workloads that still change.

Book a solution review See the related solution

By Netbase Commerce and Platform Engineering Desk · Reviewed by David (CEO) · Updated · 10 min read

star

This guide is for CTOs, platform leads and finance owners of a web, SaaS or commerce product on AWS, Azure or Google Cloud who are about to buy their first commitment, or to renew one. It sits beside Netbase's cloud platform engineering service, where cost is one of the platform concerns Netbase designs for, and it picks up where cloud cost controls for SaaS stops: after waste is removed and resources are rightsized.

In this guide

What does a cloud commitment discount actually buy?

A commitment discount buys a lower rate, not capacity. AWS describes a Reserved Instance as a billing discount applied to matching usage, not a physical instance, and the FinOps Foundation compares commitment discounts to coupons that can cover certain running resources. If nothing matching runs in an hour, the coupon for that hour is lost.

The providers sell two shapes. A spend-based commitment promises a fixed hourly spend for one or three years: AWS Savings Plans, the Azure savings plan for compute and Google Cloud's compute flexible committed use discounts. A resource-based commitment promises specific capacity, such as an instance type in a region or a number of vCPUs: AWS Reserved Instances, Azure Reservations and Google Cloud's resource-based committed use discounts.

Both shapes bill the commitment whether you use it or not. Google Cloud states that committed resources are billed until the end of the term regardless of use, and Microsoft states that unused savings plan commitment for an hour expires and does not roll over. Usage above the commitment is billed at on-demand rates.

How do savings plans, reservations and committed use discounts compare?

Instrument Commitment and flexibility Exit route Choose it when
AWS Compute Savings Plans Hourly spend, 1 or 3 years; any EC2 family, size or region, plus Fargate and Lambda None once bought Compute is steady in spend but instance types or services still change
AWS Reserved Instances Instance attributes, 1 or 3 years; Standard can only be modified, Convertible can be exchanged Standard ones can be sold on the Reserved Instance Marketplace One instance configuration will run unchanged for the whole term
Azure savings plan for compute Hourly spend, 1 or 3 years; across eligible compute services and regions Cannot be cancelled or refunded Workloads move between VMs, App Service or containers
Azure Reservations Resource, size and region, 1 or 3 years; exchange within a product family Refunds within a yearly cancellation limit A stable resource where the deeper discount matters
Google Cloud resource-based CUDs vCPU, memory or GPU in a region, 1 or 3 years Cannot be cancelled Predictable, steady-state machine usage in one region
Google Cloud compute flexible CUDs Hourly spend, 1 or 3 years; across Compute Engine, GKE and Cloud Run Cannot be cancelled Container or serverless usage that changes shape

Two notes change the comparison over time. AWS now recommends Savings Plans over Reserved Instances in its EC2 documentation. Microsoft has announced that reservations bought from 1 February 2027 for services that savings plans also cover lose the right to exchange, so an Azure reservation bought after that date is effectively a fixed bet.

The commitment ladder: how much of each usage layer to commit

Split last quarter's usage into four layers by how stable they are, then commit each layer differently. The layer, not the provider, decides the instrument and the term.

  1. Locked floor

    Production databases and core services whose type and region will not change

    Instrument
    Resource-based reservation or resource CUD
    Term
    1 year; 3 years only with no migration or re-architecture on the roadmap
    Entry test
    Same configuration ran every hour of the look-back period
  2. Stable baseline

    Compute that runs every hour but may change family, size or service

    Instrument
    Spend-based plan (Compute Savings Plan, Azure savings plan, flexible CUD)
    Term
    1 year
    Entry test
    The lowest hourly spend of the look-back period, not the average
  3. Shifting baseline

    Usage that is steady this quarter but tied to an open roadmap item

    Instrument
    No new commitment; review next quarter
    Term
    None
    Entry test
    Moves up to layer 2 after a further quarter of stable usage
  4. Variable top

    Peaks, batch jobs, tests and experiments

    Instrument
    On demand, spot or preemptible capacity
    Term
    None
    Entry test
    Never committed

Commit layer 2 at its floor, not its average. The FinOps Foundation advises buying commitments that break even within nine months and spreading purchases through the year, so each purchase can be re-evaluated. Several small purchases staggered across quarters also expire at different times, which keeps a later migration from stranding the whole commitment at once.

Deciding the level: a quarterly routine

  1. Clean first

    Remove idle resources and rightsize before any purchase; a discount on capacity you do not need is still waste.

  2. Pull hourly usage

    Export at least the last full quarter of hourly usage per service, region and instance family.

  3. Layer it

    Assign every usage line to one of the four layers, and write the roadmap item that could move each one.

  4. Read the recommendation

    Compare your layer 1 and layer 2 totals with the provider's own recommendation in AWS Cost Explorer, Azure Advisor or the Google Cloud console.

  5. Buy a slice

    Purchase part of the gap now and the rest in later quarters, in the instrument the layer calls for.

  6. Watch two numbers

    Track utilization (how much of the commitment was used) and coverage (how much eligible usage was discounted) every week.

  7. Re-layer each quarter

    Move usage between layers as the roadmap changes, and let commitments that no longer fit expire.

When should you not buy a commitment?

Do not commit before a cloud migration, a move to containers or serverless, or a database change that is already planned inside the term. Commitments bind to the service, family or region you use today; the roadmap decides whether that is still true next year. During a cloud migration and modernization, stay on demand until the target platform has run a full quarter.

Do not commit to fix a cost problem that is really waste or architecture. If a single tenant, a runaway job or an oversized instance drives the bill, a commitment locks that cost in. Do not take a three-year term only because the discount is deeper; for a product whose architecture is still changing, a one-year spend-based plan is usually the cheaper mistake.

Where AI fits in commitment planning

AI helps with the reading, not the buying. An assistant can sort a quarter of hourly billing data into the four layers, flag usage lines tied to an open roadmap item and draft the purchase memo with the provider recommendation next to the layer totals. Netbase works with the major commercial and open-source AI tools and models, chosen per project.

Engineers and the finance owner still confirm every layer against the roadmap, check the provider recommendation themselves and approve each purchase by name; no model buys or exchanges a commitment.

What delivery record exists, and what does not

Netbase works as offshore development partner on a multi-tenant cloud ERP delivered as SaaS to small and mid-sized businesses for a US client, not named, and builds and runs its own products, such as Cloodo Workspace. Netbase names AWS, Google Cloud, DigitalOcean and Cloudflare as the cloud providers it uses.

Delivery runs in a six-step lifecycle with weekly reviews and KPI dashboards, and most Netbase projects are delivered on fixed-price contracts. No published Netbase record describes a commitment purchase for a client, and no cost saving, discount rate or utilization figure is claimed.

Alternatives: other ways to lower the rate

Alternative Fits when Selection criterion
Stay on demand Usage is young, small or about to move The roadmap can change the service, family or region inside a year
Google Cloud sustained use discounts Compute Engine VMs run most of the month Applied automatically, no commitment needed
Spot or preemptible capacity Batch, CI and fault-tolerant jobs The job can stop and restart without user impact
Architecture change One workload dominates the bill Containers, serverless or caching remove the usage instead of discounting it
Commitment ladder A stable floor exists and the roadmap is known At least one full quarter of hourly usage shows a floor

Teams that want to see where cost practice sits among their other delivery habits can score it with the DevOps maturity assessment scorecard, and the usage data a ladder needs comes from the same pipeline described in observability stack design with OpenTelemetry. When the work also covers pipelines and environments, Netbase's DevOps and platform engineering service is the place to start.

What are the limits of this guide?

  • It sets a method, not a number; provider discount rates and recommendations change and are not restated here.
  • Commitment rules differ by account type and region; check the current terms of your agreement before you buy.
  • The ladder covers compute and databases; storage, network and licence commitments follow their own terms.

Plan the next step with a Netbase consultant

Frequently asked questions

Commit the usage that ran every hour of the last quarter and will not move with the roadmap, measured at its lowest hourly level rather than its average. Buy it in slices across quarters and leave peaks, experiments and anything tied to a planned migration on demand.

Buy spend-based savings plans for usage that is steady in spend but may change instance type, size or service. Buy resource-based reservations only for configurations that will run unchanged for the whole term, and accept that they are harder to exit.

Only for a locked floor with no migration, re-architecture or database change on the roadmap. For a product whose architecture is still changing, a one-year term usually costs less than a three-year commitment the product stops using.

You still pay for the full commitment. The providers bill committed amounts every hour of the term, and unused commitment does not carry over to later hours.

Next step

If you want a second opinion on your commitment level before the next purchase, book a solution review with Netbase.

AI-ready cloud platform engineering for SaaS and commerce workloads AI-ready cloud platform engineering for SaaS and commerce workloads

Netbase provides cloud platform engineering for SaaS and commerce teams that need a secure, repeatable home for their software and its AI features, on AWS, Google Cloud, DigitalOcean or Cloudflare. We design the architecture, write the environments as code, connect managed and AI services and hand over a platform your team can run, built the way we run ours.

Learn More
line
AI-assisted cloud migration with a cut-over plan and a way back AI-assisted cloud migration with a cut-over plan and a way back

Netbase provides cloud migration services for companies running business applications on ageing servers or hosting to move them to AWS, Google Cloud, DigitalOcean or Cloudflare without losing data or control. AI speeds up the code and dependency assessment; every migration still has an inventory, a rehearsed cut-over, cost controls from day one and a tested rollback point.

Learn More
line
AI-assisted DevOps and platform engineering for routine releases AI-assisted DevOps and platform engineering for routine releases

Netbase provides DevOps and platform engineering for SaaS and commerce teams to make releases routine instead of risky. We set up CI/CD pipelines, repeatable environments and observability on AWS, Google Cloud, DigitalOcean or Cloudflare, with AI assistance for tests, reviews and incident triage, so every change ships the same way and problems surface before customers notice.

Learn More
line
SaaS product accelerator: launch an AI-ready SaaS on proven Netbase modules SaaS product accelerator: launch an AI-ready SaaS on proven Netbase modules

A SaaS product accelerator is a set of reusable Netbase modules for accounts, billing, roles and integrations that helps founders and product teams launch subscription software faster, with room for in-product AI from the first release. Reusing these modules can shorten development time, and the approach is proven on Printcart, Netbase's own web-to-print SaaS.

Learn More
line

Work with Netbase remotely

Every Netbase engagement is delivered remotely, so clients in any country and any industry can work with us; Netbase has worked with clients in 80 countries. Contracts and invoices are issued by Netbase JSC, and payment is made by bank transfer to Netbase JSC's company account or PayPal Business. Talk to the CEO, Huy Nguyen (David), on WhatsApp or by email.

Contact Netbase

Discuss a project

Netbase JSC helps organizations design, build, modernize, and operate digital products and AI-enabled business systems.
Project enquiries

project@netbasejsc.com

WhatsApp

+84 937 869 689

Office address

91 Nguyen Chi Thanh, Dong Da, Hanoi, Vietnam

Get in touch

Tell us what you want to build, modernize, or operate.

Tell us what you want to build, modernize, or operate.

Contact Netbase