A commitment coverage target on GCP is the share of your Compute Engine usage you put under Committed Use Discounts, and the correct number is the steady state floor your forecast can defend, not the highest coverage you can buy. Cover the usage you are confident will run for the full one or three year term, and leave the variable layer on demand, where sustained use discounts apply automatically to eligible usage. Coverage that chases the maximum discount strands spend the moment usage shifts, so the target follows the forecast, not the other way around.
This is a GCP compute and commitments discipline. Here is how to set the target, choose the CUD type, and avoid the over commitment trap.
What should the coverage target be?
There is no universal percentage, because the right floor depends on how much of your compute is genuinely steady. Build the target bottom up. Look at the trailing usage in your billing export, identify the baseline that has run consistently for months and is forecast to continue, and set coverage to that floor. The variable layer above it, the bursty and seasonal usage, stays uncovered and earns sustained use discounts on its own. For most estates the defensible floor is a clear majority of baseline compute, but the honest answer is whatever you can stand behind in a renewal conversation. Size it carefully, as covered in sizing CUD coverage without stranding spend.
Spend based or resource based CUDs?
GCP offers two CUD shapes and the choice is a flexibility versus discount trade.
| CUD type | What you commit | Best for |
|---|---|---|
| Spend based | A dollar per hour amount across eligible services | Flexible baseline where the machine mix may change. |
| Resource based | Specific machine resources in a region | A stable, well understood workload in one region. |
Spend based CUDs trade some rate for flexibility and ride through changes in machine family. Resource based CUDs give a deeper discount but lock you to specific resources in a region, so they suit only the most predictable workloads. The distinction matters enough that we cover it on its own in the risk adjusted approach to GCP commitments.
How do sustained use discounts change the math?
Unlike AWS or Azure, GCP applies sustained use discounts automatically to eligible Compute Engine usage that is not already under a CUD. This is why over committing is doubly wasteful on GCP: usage you leave uncovered is not paying full on demand all month anyway, so the marginal benefit of pushing coverage past the steady floor shrinks while the stranding risk rises. The buyer move is to cover the floor with CUDs for the deeper committed rate and let sustained use discounts work the variable layer for free.
A European SaaS company had been advised to cover nearly all GCP compute with three year resource based CUDs to maximise the discount. The billing export showed only part of that usage was truly steady; a meaningful slice was seasonal. Resizing coverage to the defensible floor, using spend based CUDs for the flexible baseline and resource based only for the rock steady core, and letting sustained use discounts handle the variable layer raised realised savings while removing the stranding risk the maximised plan carried. Figures are verified against billing data and anonymised.
Keep coverage honest over time
A coverage target is not set once. Usage drifts, machine families change, and a CUD bought around last year's estate can quietly go underused. Monitor utilization monthly, ladder renewals so no single commitment cliffs at once, and revisit the target each quarter against an updated forecast. GCP Recommender will suggest CUD purchases, but it recommends, it does not decide, so test every suggestion against your own forecast before buying. This standing discipline is what holds the 31 percent median reduction we see in the first 90 days. The full strategy lives in the GCP cost optimization guide, and the cross cloud view in the cloud cost optimization guide.
Frequently asked questions
Set your coverage target with us
We build the forecast, set a CUD coverage target you can defend, and choose the right mix of spend based and resource based commitments so you capture the discount without stranding spend. Our guarantee: we reduce your cloud spend or we reimburse our service fee.
Put a defensible number on your cloud spend.
No provider in the room, no published price list. Tell us your footprint and we will scope the savings against your billing data — we reduce your cloud spend or we reimburse our service fee.
The Cloud Spend Navigator: what changed in cloud pricing, commitments, and FinOps — no vendor spin.