A discount isn't a saving if you commit to the wrong usage

A cloud provider offers a 25% discount in exchange for a commitment. The calculation looks straightforward.

Expected eligible spend: $500,000 per month. Potential discounted cost: $375,000 per month. Potential benefit: $125,000 per month.

Why would Finance say no?

Because the discount is only one side of the decision. The organisation is also making an assumption about the future: we believe we will continue to need enough eligible usage to justify this commitment.

If that assumption is right, commitments can create significant financial value. If it is wrong, the organisation may end up paying for consumption it no longer needs.

The real FinOps question is therefore not "how large is the discount?" It is "how much future usage are we confident enough to commit to?"

Commitments exchange flexibility for a lower rate

On-demand cloud consumption provides valuable flexibility. Use more, and the bill increases. Use less, and the bill decreases.

A commitment changes that relationship. The organisation accepts some form of future consumption obligation in exchange for improved pricing.

The exact mechanics vary across cloud providers and commitment products, but the economic principle is similar: lower rate in exchange for reduced flexibility.

That trade-off is important because future cloud demand is uncertain. Products change. Customers behave differently than forecast. Workloads migrate. Architectures evolve. Engineering finds more efficient ways to operate. Businesses acquire or divest activities.

A commitment decision therefore needs to consider not only today's usage, but the organisation's confidence in tomorrow's usage.

The headline discount can be misleading

Imagine an organisation commits to $400,000 of eligible monthly usage at a 20% lower effective rate. If the full commitment is used, the economics may look attractive.

But six months later, an architecture change reduces the relevant consumption to $300,000. The organisation has improved its technical efficiency. Yet part of the commitment may now be underutilised.

This is the paradox. The cloud environment became more efficient. The commercial position became less efficient.

That is why a commitment cannot be evaluated only by its nominal discount. The organisation needs to consider the effective financial outcome after utilisation.

A large discount on an underutilised commitment may create less value than a smaller discount applied to usage the organisation can confidently sustain.

Coverage and utilisation answer different questions

Two metrics are particularly useful.

CoverageHow much eligible usage benefits from commitment-based pricing? If an organisation has $1 million of eligible usage and $700,000 is covered by commitments, coverage is approximately 70%. Higher coverage can reduce exposure to on-demand pricing. But high coverage is not automatically desirable.

UtilisationHow much of the commitment we purchased are we actually consuming? If an organisation purchased commitments for $700,000 of expected usage but consumes only $560,000 of the committed amount, part of the obligation is underutilised.

The two metrics create a natural tension. Pushing coverage aggressively upward can increase the risk of lower utilisation. Remaining too conservative can preserve utilisation but leave stable usage paying higher rates.

The objective is not to maximise either metric. It is to find an economically appropriate balance.

The commitment decision starts with confidence

Not all cloud consumption has the same level of predictability.

Consider an organisation with $1 million of eligible monthly usage. Its environment might contain:

$500,000 of highly stable consumptionCore production workloads with long histories and no expected material changes.

$300,000 of moderately predictable consumptionWorkloads expected to continue but affected by growth, optimisation or product changes.

$200,000 of uncertain consumptionApplications facing migration, architectural change, potential retirement or volatile demand.

Treating all $1 million as equally predictable would hide important risk.

The organisation may be comfortable making a stronger commitment against the stable base. It may take a more conservative approach to the middle layer. And it may deliberately leave the uncertain portion flexible.

That is not lost savings. It is risk management.

Historical usage is evidence, not certainty

Commitment analysis naturally relies on historical consumption. That makes sense. Stable historical usage can be a strong indicator of future demand.

But it should not be mistaken for a guarantee.

A workload can have consumed the same amount for 18 months and still be scheduled for migration next quarter. A database can appear stable while Engineering is planning to replace it. A rapidly growing service can look highly predictable until a major architecture change alters its cost profile.

Historical data therefore needs to be combined with forward-looking context. Before making material commitments, the organisation should understand planned migrations, architecture changes, decommissioning, product roadmap changes, expected business growth or contraction, major optimisation initiatives, and contractual or strategic changes.

This is why commitment management cannot belong exclusively to Finance or exclusively to Engineering. It requires both views.

Longer commitments require stronger conviction

A longer commitment may offer better economics. It also increases the period over which the organisation needs to be right. That makes duration a risk decision.

Suppose two options exist.

Option AModerate discount. Shorter commitment. Greater flexibility.

Option BLarger discount. Longer commitment. Lower flexibility.

Choosing Option B purely because it offers the largest percentage saving ignores the value of optionality.

The right decision depends on the workload. A mature, stable platform with little expected architectural change may support stronger commitment. A fast-changing environment undergoing modernisation may justify paying a higher rate to preserve flexibility.

The difference in price can be thought of partly as the cost of optionality. Sometimes paying more today protects the organisation from a much larger mistake tomorrow.

Growth forecasts should be treated carefully

Expected business growth can create pressure to commit more aggressively. If customer demand is forecast to grow 30%, why not buy commitments for the expected future consumption now?

Because business growth and cloud consumption do not necessarily move one-for-one. A 30% increase in transactions may create only a 15% increase in cloud demand. Engineering improvements may absorb part of the growth. Product mix may change. New architecture may scale differently. Forecasts can also be wrong.

A strong commitment strategy therefore distinguishes between usage that already exists and is stable, and usage expected to appear in the future. The second category usually deserves a higher level of caution.

Portfolio effects matter

Commitments should not always be evaluated workload by workload. Depending on the commitment mechanism, usage across multiple workloads may contribute to utilisation. That diversification can reduce risk.

One application may shrink while another grows. One environment may migrate while another expands. At portfolio level, aggregate demand can sometimes be more stable than individual workloads.

This is valuable because it changes the question from "will this exact application still consume the same amount?" to "how confident are we that eligible consumption across the relevant portfolio will remain above this level?"

The appropriate level of analysis depends on the commercial mechanism and environment, but the principle is important: risk can sometimes be managed through diversification, not just conservative individual workload forecasts.

Commitment headroom has value

Imagine the organisation is highly confident that eligible usage will remain above $600,000 per month. Current consumption is $800,000. Should it commit all $800,000?

Not necessarily. The remaining $200,000 of on-demand exposure creates headroom. That headroom can absorb optimisation, demand reduction, migrations, architectural change and forecasting error.

The organisation pays a higher rate on that flexible portion. But that does not automatically make the decision inefficient. It is paying for flexibility.

This is similar to many financial decisions: maximising the theoretical return often requires accepting more risk. FinOps should make that trade-off visible.

Evaluate the downside, not only the expected saving

Commitment business cases often focus heavily on the upside. For example: expected annual discount benefit $1.4 million.

A stronger decision also asks: what happens if demand is 10% lower than expected? What happens if a major migration happens six months early? What happens if optimisation removes more usage than planned? What happens if the product strategy changes?

This does not require sophisticated financial modelling for every commitment. But material decisions should consider scenarios:

Expected caseHigh utilisation and strong financial benefit.

Lower-demand caseSome underutilisation, but commitment remains economically attractive.

Downside caseMaterial unused commitment reduces or eliminates the expected benefit.

The organisation can then make an informed decision about whether the potential saving justifies the risk.

Commitment management does not end at purchase

A commitment should not disappear from attention once it has been purchased. The environment continues to change. Usage needs to be monitored. Utilisation can deteriorate. Coverage can become too low. New stable demand may create additional opportunities. Architecture changes may alter the risk position.

Commitment management should therefore be continuous. The organisation should regularly understand: what have we committed? How well are we utilising it? Where is coverage appropriate or insufficient? What material changes are coming? When do commitments expire? How has our demand confidence changed?

This turns commitments from a procurement event into an ongoing FinOps capability.

Finance, Engineering and FinOps see different risks

Each stakeholder brings a different perspective. Finance sees the financial obligation and expected benefit. Engineering sees how architecture and demand may change. Procurement understands commercial terms and contractual constraints. FinOps connects usage, pricing, forecasts and technology context.

The decision is strongest when those perspectives come together.

A commitment should not be purchased because Finance likes the discount. Nor should it be rejected simply because Engineering cannot guarantee future consumption with certainty.

The objective is to determine whether the organisation has enough confidence to accept the financial obligation.

The best commitment strategy may look conservative

There can be pressure to maximise commitment coverage because uncovered usage appears to represent "missed savings". That framing is incomplete.

Some uncovered usage may represent intentional flexibility. If the organisation has consciously decided that a portion of demand is too uncertain to commit, paying the on-demand rate may be the economically correct decision.

This is particularly true during periods of major migration, architectural transformation, uncertain growth, product change or aggressive optimisation.

A FinOps team should be able to explain why some usage remains uncommitted. That is stronger than pursuing a coverage target without considering the risk underneath it.

The decision is ultimately about risk-adjusted value

Commitments can generate substantial savings. They are an essential part of cloud financial management for many organisations.

But the strongest commitment strategy is not the one that produces the largest theoretical discount. It is the one that creates attractive financial value for an acceptable level of demand risk.

That requires understanding: how stable is the usage? How might technology change? How confident are we in growth? How much flexibility do we want to preserve? What happens if our assumptions are wrong? Is the expected discount sufficient compensation for the obligation we are accepting?

Those are financial questions as much as technical ones. And that is exactly why commitments belong in FinOps.

Key takeaway. A cloud commitment is not simply a discount. It is a financial decision about future technology demand.

Higher coverage can create more savings potential. But it can also create greater exposure if demand changes. The right strategy balances discount, utilisation, coverage, duration and flexibility against the organisation's confidence in future usage.

Do not ask only "how much can we save by committing?" Ask "how much usage are we confident enough to commit — and what happens if we are wrong?"

Nooven helps organisations evaluate cloud commitments through both financial and technology context — connecting usage, demand confidence, optimisation plans and commercial options to support better risk-adjusted commitment decisions.