Should I concentrate cloud commitment with one provider or split it?

Concentration unlocks the deepest discount tier but makes you maximally dependent on one provider and weakest at renewal. Splitting across AWS, Azure and Google preserves competition and resilience at the cost of some discount depth. The right answer depends on how portable your workloads truly are.

Does splitting commitment cost me discount?

Usually some. As of June 2026 discount tiers reward concentration, so pooling spend with one provider reaches a deeper tier than splitting the same total across three. The question is whether the leverage and resilience you gain outweigh the discount you give up.

When does splitting make the most sense?

When you have genuinely portable workloads, so the threat to move spend is credible and keeps every provider competing for your next commitment. If workloads cannot practically move, the leverage from splitting is theoretical.

How should I size each provider commitment in a split?

Size each commitment to that provider own certain floor, never to a forecast. Concentrate committed spend only where workloads are sticky anyway, since you lose no leverage there, and keep portable workloads spread.

What is the risk of committing deeply with every provider?

Stacking deep multi year commitments with all three providers at once recreates lock in three times over and removes the leverage that splitting was meant to preserve. Build flexibility into each deal rather than maximising discount on all of them.

Does a split improve my negotiating position?

Yes. Holding live spend with more than one provider lets you bring real benchmarks and a real alternative to each negotiation. A consolidated buyer can only argue against the provider own numbers.

COMMITMENT STRUCTURING · BUYER SIDE

Splitting commitment across hyperscalers

GET A CONFIDENTIAL REVIEW →

PUBLISHED JUNE 16, 2026 · REVIEWED JUNE 16, 2026

Splitting commitment across hyperscalers is the structuring question multi cloud enterprises face: do you concentrate your committed spend with one provider to reach the deepest discount tier, or spread it across AWS, Azure, and Google Cloud to preserve competition and flexibility? Concentration buys discount depth. Splitting buys leverage and resilience. The right answer depends on how genuinely portable your workloads are and how much you value keeping the providers competing for your next dollar.

Each provider will tell you that consolidating with them is the smart move. Of course they will. The deepest tier sits with the largest single commitment. Your interest is broader than any one provider discount curve, and splitting commitment across hyperscalers is often the structure that serves it best.

Splitting commitment across hyperscalers versus consolidating

Discount tiers reward concentration. As of June 2026, an AWS EDP scales its discount with the committed amount (source: AWS EDP program terms), Azure MACC ties discount to a fixed consumption commitment (source: Microsoft MACC documentation), and Google offers CUDs plus custom private pricing for large enterprises (source: Google Cloud documentation). Pooling your spend into one of these unlocks a deeper tier than splitting the same total across three. That is the case for consolidation, and on pure discount math it is real.

The case against is everything that depth costs you. A concentrated commitment makes you maximally dependent on one provider, removes the credible threat of moving spend, and leaves you with the least leverage at renewal. Splitting keeps every provider aware that your next workload, and your next commitment, could go elsewhere.

What splitting actually buys you

Live competitive tension

When you hold meaningful spend with more than one provider, each renewal is a real contest. The credible ability to shift workloads is the strongest lever a buyer has, and it only exists if you have not concentrated everything in one place. This connects to the multi year versus single year tradeoffs, because splitting and shorter terms together keep the maximum number of decision points alive.

Resilience and portability

A split footprint reduces the operational and commercial risk of a single provider problem, whether that is a price change, a service gap, or a strategic shift. But portability has to be real. If a workload cannot practically move because of data gravity, proprietary services, or migration cost, then the leverage from splitting is theoretical and you are paying the higher split price for an option you cannot exercise.

Negotiating power at each table

There is a subtler benefit. When you negotiate with one provider while holding live spend with another, you carry real benchmarks and a real alternative into the room. You can cite what comparable workloads cost elsewhere and mean it, because you actually run them elsewhere. A consolidated buyer has only the provider own numbers to argue against. A split buyer brings outside evidence to every negotiation, which is worth more than most buyers realise.

How to decide the split

Map which workloads are genuinely portable and which are locked to a provider by data or services.
Concentrate committed spend only where workloads are sticky anyway, since you lose no leverage there.
Keep portable workloads spread, so the threat to move is credible at renewal.
Size each provider commitment to its own certain floor, never to a forecast.
Avoid stacking deep multi year commitments with every provider at once, which recreates lock in three times over.

The per provider sizing discipline is identical to single provider sizing, covered in how to size a cloud commitment correctly. The difference is that you are running the exercise three times and weighing the total discount lost against the leverage gained.

A worked illustration

Take a composite enterprise spending fifteen million a year across two providers, roughly ten million on one and five million on the other. The larger provider offers a deeper tier if the buyer consolidates the full fifteen million with them. The extra discount looks attractive. But consolidating would retire the only credible alternative the buyer has, and most of the five million on the second provider runs portable workloads that genuinely could move. A buyer side structure keeps the split, sizes a conservative commitment with each provider to its own floor, and uses the live presence on both to negotiate harder at each renewal. The modest discount given up by not consolidating is repaid many times over by the leverage retained across future negotiations.

Watch the instrument interactions

Splitting adds complexity to the discount stack. Each provider has its own underlying instruments, and they do not behave the same way. As of June 2026, GCP Sustained Use Discounts apply automatically with no commitment and do not double stack with CUDs on the same resource (source: Google Cloud documentation), while AWS commitments sit on top of Reserved Instances and Savings Plans. We compare these in reserved instances vs savings plans vs commitments. Building flexibility into each provider deal, rather than maximising discount on all three, is the safer posture, a theme we develop in building flexibility into a commitment.

Operational cost of a split footprint

Splitting is not free on the operational side, and an honest buyer side view accounts for that. Running meaningful spend across more than one provider means duplicated tooling, broader skills requirements, and more complex governance. These costs are real and they offset some of the leverage benefit. The decision is not discount depth versus leverage alone, it is discount depth versus leverage net of the operational overhead of multi cloud. For some organisations the overhead is modest because they are already multi cloud. For others it is significant.

Weigh the whole picture before you let a provider talk you into consolidating purely for a deeper tier. The discount they offer to pull all your spend into one place is the price they pay to remove your leverage. Whether that trade is worth it depends on your portability, your operational maturity, and how much you value keeping a credible alternative alive, the same factors that drive how to size a cloud commitment correctly on each provider.

One more discipline matters when you split. Stagger the renewal dates across providers rather than letting them all expire together. Staggered renewals mean you always have at least one live negotiation and one credible alternative in play, which keeps pressure on every provider continuously. Coterminous renewals, where every commitment ends at once, hand the providers a coordinated moment and dilute the very leverage that splitting was meant to preserve. Spread the dates the way you spread the spend.

The bottom line on splitting commitment across hyperscalers

Concentration buys discount depth, splitting buys leverage and resilience. Concentrate only where workloads are sticky anyway, keep portable spend spread so your threat to move stays credible, and size each provider commitment to its own certain floor. For the full framework see our cloud commitment structuring and sizing guide. If you are weighing consolidation against a split, a commitment structuring and sizing service will model the discount lost against the leverage gained before you sign.