- Market
- Saudi Arabia
- Written for
- CIO, CTO
- Decision
- Understand the one-off and recurring cost of a cloud migration
- Languages
- EN · AR
Information security for cloud delivery

ISO/IEC 27001
ISO/IEC 27001 certified
Move cost and run cost
What sets the one-off move and what sets the monthly bill
Each driver is marked for which cost it moves — the one-off migration, the recurring bill, or both. The two are easy to conflate and expensive to confuse.
| Variable | Effect on price | Why |
|---|---|---|
| How much you change on the way | Move cost up, run cost down | Lifting a system across unchanged is the cheapest move and often the most expensive month, because it runs in the cloud the way it ran in a server room, paying for idle capacity around the clock. Re-architecting a workload to use the cloud properly costs more to migrate and can cost far less to run. The choice between them is the biggest single lever, and it trades a one-off figure against a recurring one. |
| How well the workloads are sized | Sets the recurring bill directly | Cloud bills are dominated by resources that run larger or longer than the work needs. Right-sizing what each workload actually uses, and letting it scale down when idle, is where a monthly figure is genuinely set. A migration that copies the old server sizes across preserves years of over-provisioning as a recurring charge, which is the most common way a cloud bill disappoints. |
| Data volume and movement | One-off transfer, recurring egress | Moving large volumes of data has a one-off cost and a duration that affects the cutover. Then there is a recurring one people miss: providers charge to move data out, so an architecture that constantly ships data between regions, services or back on-premises builds a monthly egress charge into ordinary operation. How data flows after the move matters more to the bill than how much of it there is. |
| Data residency and compliance | Shapes the region and the design | Where data must physically reside, and the controls a regulated workload has to run under, decide which regions and services are available and how the architecture is arranged. Meeting residency and sector rules is a requirement to design for, not a setting, and it can rule out the cheapest configuration — so it belongs in the estimate rather than as a discovery after go-live. |
| Running the two in parallel | A temporary but real double cost | A safe migration usually runs the old and new environments together for a while, so the business is never bet on a single switchover. That overlap is a real, if temporary, double cost, and a plan that ignores it has under-costed the safe path. The alternative — a hard cutover with no parallel period — is cheaper on paper and carries the risk of an outage the operation cannot afford. |
Ask any supplier quoting this
- Does this figure cover only the move, or also the expected monthly running cost?
- For each workload, is it a lift-and-shift or a re-architecture, and why?
- How will workloads be right-sized rather than copied at their current size?
- What ongoing data-egress cost does the proposed architecture build in?
- How do data-residency and sector rules constrain the regions and design?
- Does the plan cost a parallel-running period, or assume a hard cutover?
The bill that stays
A cheap move can be the most expensive decision you make
The fastest, cheapest way to migrate is to lift each system across exactly as it is and change nothing. It gets you into the cloud quickly and it wins on the one-off figure, which is the number most migrations are judged on. It also reproduces every assumption the systems made about a server room — fixed capacity always on, sized for the busiest hour, paid for around the clock — inside a place that bills by the hour for what you provision. The result is a monthly bill that can be higher than the data centre it replaced, and it arrives every month for as long as the systems run.
This is why the honest way to cost a migration is to quote two numbers, not one, and to weigh them against each other. Sometimes a fast lift-and-shift is exactly right — to exit a data centre by a deadline, or for a workload not worth re-engineering. Often the better economics come from re-architecting the few workloads that dominate the bill so they use the cloud as it is meant to be used, accepting a higher one-off cost to lower the recurring one. The decision cannot be made on the move alone, because the move is paid once and the bill is paid forever.
Where the bill is set
The stages at which the monthly cloud bill is actually decided
The recurring cost is not set when the invoice arrives; it is set during the migration, at the points below. Each is a chance to lower the bill or to lock in a high one.
Assess each workload on its own
We look at what each system does and how it is used before deciding how it moves, because the right treatment differs per workload and a single wholesale approach over-spends on some and under-serves others. A per-workload plan that spends re-engineering effort where it lowers the bill.
Right-size before moving, not after
We size each workload to what it actually uses rather than copying the old server dimensions, so the migration does not carry years of over-provisioning into a place that charges for it every hour. A starting bill matched to real usage instead of to the old hardware.
Design the data flows
We arrange where data lives and how it moves so the architecture does not build a recurring egress charge into ordinary operation, and so residency requirements are met by design rather than discovered late. A data layout that meets the rules and does not leak money on movement.
Leave the controls that keep it legible
We hand over sizing, scaling and monitoring so your team can see what each part costs and adjust it, because a cloud bill without cost controls drifts upward quietly and a migration that omits them has finished half the job. A bill your team can read and control rather than only receive.
Our role
What we quote before we agree to move anything
- 01
Both numbers, side by side
We estimate the one-off move and the expected monthly running cost together, so the decision is made on the total of owning the platform rather than on the cheaper number a move-only quote shows.
- 02
Re-architecture where it pays
We re-engineer the workloads that dominate the bill and lift the ones that do not, rather than applying one approach to everything, because that is how the recurring cost is genuinely lowered rather than merely relocated.
- 03
Controls left with your team
We leave the sizing, scaling and monitoring that keep the bill legible and adjustable, so the platform stays affordable after we are gone instead of drifting upward unwatched.
How fast you must be out of the data centre, and how much re-engineering is worth to lower the monthly bill, are yours to weigh; we quote both the move and the run so the trade between them is a decision you make with the numbers in front of you.
Common questions
What buyers ask
Isn't moving to the cloud supposed to save money?+
It can, but not automatically, and a migration done for speed alone often raises the monthly bill rather than lowering it. The cloud saves money when workloads are sized to what they actually use, scale down when idle, and are built to use managed services instead of always-on servers. It costs more when systems are lifted across unchanged, because they keep running as though they were in a server room — full capacity, all the time — inside a place that charges by the hour. Whether the cloud is cheaper for you depends far more on how the migration is done than on the decision to migrate, which is why the running cost has to be estimated before you commit, not discovered on the first invoice.
Should we lift and shift quickly or re-architect as we move?+
Usually a mix, decided per workload rather than as one policy. A lift-and-shift is right when you have to exit a data centre by a deadline, or for a system not worth re-engineering because it is small, stable or on its way out. Re-architecting earns its higher one-off cost for the handful of workloads that dominate the running bill, where using the cloud properly can cut the monthly figure substantially. Applying either approach to everything is the mistake: lifting everything preserves an expensive bill, and re-architecting everything spends engineering effort on workloads that would never repay it. The plan worth paying for is the one that tells you which workloads deserve which treatment and why.
What cloud costs surprise organisations after they have migrated?+
Two recur. The first is over-provisioning carried across from the old environment: servers sized for the busiest hour but paid for every hour, and resources left running when nothing uses them, which is the largest avoidable part of most bills. The second is data egress — providers charge to move data out, so an architecture that constantly ships data between regions, between services, or back to on-premises systems builds a monthly charge into ordinary operation that nobody sized at migration time. Both are set during the migration and both are hard to unwind afterwards, which is exactly why right-sizing and data-flow design belong in the plan rather than in a cost-cutting exercise a year later.
In this cluster
Cost both numbers
A migration plan that quotes the one-off move and the expected monthly running cost together, so the decision is made on both numbers rather than on the cheaper one.
Start the conversation