Skip to content
North Ark

Cloud cost optimisation

Azure spend that follows the design, and stays down once it is fixed.

Cloud costs creep back when savings are one-off clean-ups. We find the savings, make the changes safely through your change process, and keep reviewing spend on a set cadence so it doesn't drift again.

Start with an Azure architecture and cost review
Small server rack alone in a glass-walled room, with cloudy sky reflected in the glass.

Where the savings usually are

  • Idle and orphaned resources: disks, IP addresses, snapshots and test environments nobody switched off
  • Over-sized virtual machines and databases, sized from measured use rather than the original estimate
  • Schedules for non-production resources that don't need to run overnight or on weekends
  • Storage tiers and retention that match how data is used
  • Reservations and savings plans for the workloads you will keep running
  • Licensing benefits for Windows Server and SQL Server that are available but not applied
  • Log and monitoring ingestion, often one of the fastest-growing lines on the bill

How it runs

  1. Visibility: tagging and allocation set up so spend can be traced to owners and applications.
  2. Review: resources, usage and commitments analysed, with savings ranked by value, effort and risk.
  3. Change: savings made through your change process, as code where the resources are managed as code.
  4. Commit: reservations and savings plans bought once usage is right-sized, so you commit to what you need.
  5. Cadence: spend reviewed monthly, with anomalies alerted and the trend covered in the quarterly architecture review.

Cost follows design

The largest savings often come from design changes: a hub network built for more traffic than exists, databases on virtual machines that could be platform services, or backups kept far longer than required. Those changes are scoped and made as planned work, with the savings and the risk set out before anything changes.

Who decides

North Ark finds the savings and makes the changes. Your team decides where the money goes, which trade-offs are acceptable and when commitments are made. Every change is recorded, so finance can see what moved and why.

Questions buyers ask

Do you charge a percentage of the savings?

No. Cost work is part of a Managed Engineering domain or a fixed-scope review, so recommendations are not shaped by how we are paid.

Will savings put performance at risk?

Changes are sized from measured use, tested and made through change control, with rollback. Savings with real risk are shown with that risk so you can decide.

Do you work with AWS or GCP costs?

Our focus is Azure and hybrid estates. We can review GCP alongside Azure where both are in use.

Should we buy reservations now?

After right-sizing, not before. Committing first locks in the waste. We size commitments to the workloads you will keep running.

Can we start with a one-off review?

Yes. The Azure architecture and cost review is a fixed-scope starting point, and ongoing cost work can follow as a Managed Engineering domain.

Related

Start a conversation

Get the Azure bill back under control, and keep it there.

Tell us how your spend has moved and who owns it.

Talk to an Engineer