Cloud migration economics turn on two numbers that buyers routinely confuse: the one off project cost to move, and the recurring cost to run the estate once it has arrived. The buyer takeaway is that the run cost decides whether a migration pays back, and the run cost is set after landing by rightsizing, architecture, and commitment coverage, not by the migration itself. A lift and shift that is never tuned often runs more expensively than the estate it replaced, while the same move followed by deliberate optimisation typically reaches payback within the first year or two. Model both numbers honestly before you commit, and treat provider funding as a discount on the move rather than a reason to make it.
Here is how to break down the true cost of a migration, where to find funding that buyers overlook, and how to build a business case that survives contact with the bill.
What are the real cost categories?
A credible migration model has two sides. The project side is the one off cost of getting there; the run side is the recurring cost of being there. Buyers who only model the project side are surprised by the bill, and buyers who only model the run side underestimate the effort to arrive.- Assessment and planning, including discovery of dependencies that the estate documentation always understates.
- Rework and modernisation, from repackaging applications to redesigning data layers, which is where lift and shift saves time but defers cost.
- Dual running, the period when you pay for both the source estate and the cloud target at once, often the largest and most forgotten line.
- People and change, the engineering time and the training that a migration consumes and that competes with the roadmap.
- The recurring cloud bill itself, which is the number that determines payback and which optimisation, not migration, controls.
Where can buyers find migration funding?
Every major provider funds migrations because a migration locks in years of recurring spend. AWS offers migration programs and credits, the Azure migrate and modernise incentives sit alongside the broader MACC, GCP provides migration funding and credits, and Oracle uses Universal Credits and support offsets to make moving to OCI attractive. Used well, these materially reduce the project cost and the dual running window. The buyer side caution is that funding is paired with a spend commitment, and a commitment is a forecast you are agreeing to be wrong about in only one direction. Size the commitment to a defensible forecast of the optimised estate, not to the credit being offered. Negotiating leverage here comes from a credible forecast, benchmark data, timing against the provider's quarter, and the real option of placing workloads on another cloud, which is exactly the independence we bring to the table.How do you keep the business case honest?
Most migration business cases are optimistic because they compare a tuned cloud estate against an untuned source estate, and because they book the savings before the optimisation work is funded. An honest case does three things. It models the day one cost and the optimised cost as separate states with a dated path between them. It funds the optimisation work explicitly rather than assuming it happens for free. And it states its assumptions about commitment coverage, rightsizing, and architecture so they can be tested against the actual bill. A useful rule is to require the case to hold even if optimisation lands later and smaller than hoped. If the migration only pays back on aggressive assumptions, it is a bet, not a business case, and it should be presented as one.Does the destination cloud change the math?
It does, in ways worth modelling per provider. On AWS the standing wins after landing are Graviton and gp3 migrations plus Savings Plans coverage. On Azure the Hybrid Benefit can change the math sharply for Windows and SQL estates, and reservations plus the Azure Savings Plan cover the rest. On GCP, committed use discounts and sustained use discounts apply, and network tier choice matters. On OCI, license included versus bring your own license changes database economics, egress is materially cheaper than the hyperscalers, and Support Rewards offset Oracle support fees. These differences mean the cheapest destination is workload specific. A database heavy Oracle estate and a Linux microservices estate can point to different clouds, which is why a placement decision should follow the economics of each workload rather than a single default provider.A Fortune 500 retailer planned a lift and shift to hit a datacenter exit deadline and modelled savings against the old estate without funding any post migration optimisation. We rebuilt the case as two states, a day one cost that was honestly higher than the source estate and an optimised cost reached through rightsizing, gp3 and Graviton moves, and Savings Plans coverage, with a funded plan and dates between them. We also restructured the commitment to match a defensible forecast rather than the credit on offer. The honest case still cleared the bar, the exit deadline was met, and the estate reached payback inside the planned window because the optimisation work was funded rather than assumed. Figures are verified against billing data and anonymised.
Frequently asked questions
What does a cloud migration actually cost?
Do cloud providers help fund migrations?
How long until a cloud migration pays back?
Model your migration economics with us
We build buyer side migration business cases across AWS, Azure, GCP, and OCI, model the run cost honestly, and negotiate provider funding sized to a defensible forecast rather than to the credit on offer. Our guarantee: we reduce your cloud spend or we reimburse our service fee, on a Fixed Fee or a no risk Gainshare basis.
Our cross cloud cost guide sets migration in the context of every other lever, and we share new analysis through The Cloud Spend Navigator.
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.