Contact
All services Service

Cost optimisation

We break cloud and on-premises spend down by system, service and owner, and find idle, oversized and duplicated resources and unused licences. Each cut goes into a reduction plan with its expected savings, effort and risk, reversible changes first.

Describe what you need

What you get

  • A spend breakdown by system, service and owner, including data transfer, idle capacity and licences
  • A list of idle, forgotten, oversized and duplicated resources and unused licences
  • Right-sizing recommendations based on peak, p95 and baseline usage
  • A reduction plan ranked by expected savings, effort and risk, with reversible changes first
  • Budget alerts, tagging conventions and idle-detection policies
  • Written findings for finance and engineering: what, where, how much, and what breaks if it is cut

How it works

  1. 01

    Break down the spend

    We break spend down by system, service and owner from the billing data, including data transfer, licences and idle capacity.

  2. 02

    Find the waste

    We identify idle, forgotten, oversized and duplicated resources and unused licences, and check sizing against measured peak, p95 and baseline usage.

  3. 03

    Rank the reductions

    We list each cut with its expected savings, effort and risk, reversible changes first and irreversible ones last.

  4. 04

    Set guardrails

    We set up budget alerts, tagging conventions and idle-detection policies to keep the bill from growing back.

Problems it solves

  • ProblemThe bill shows totals by category, so nobody can say which system, service or team the growth comes from.

    SolutionWe break spend down by system, service and owner, including data transfer, idle capacity and untracked licences, and name the team each cost belongs to.

  • ProblemForgotten dev and staging copies, old pilots and idle resources keep running and billing, though nothing on them is used.

    SolutionWe check usage and ownership of each resource in scope, including ones that still look in use, and show what each costs and what would break without it.

  • ProblemInstance sizes and SKUs were picked for a worst case years ago, and the unused capacity is still billed.

    SolutionWe base right-sizing recommendations on measured usage: peak, p95 and baseline. Each one states the expected savings and the risk of the change.

  • ProblemLicences are paid for unused seats, unused features and every core of an oversized host.

    SolutionWe compare licences with what is in use, including per-core licences on oversized hosts and bring-your-own-licence (BYOL) options, and add each saving to the reduction plan.

What it covers

  • Azure
  • AWS
  • VMware / VCD
  • On-premises
  • Spend by system, service and owner
  • Finding idle and oversized resources
  • Right-sizing on measured usage
  • Licence review, including BYOL
  • Data transfer and egress costs
  • Budget alerts, tagging and idle detection
Typical scope

Typical engagements run 3–8 days, depending on how many environments and accounts are in scope and how clean the billing data is. A single-subscription review takes less time; several platforms and owners take longer.

Next step

Send a short description of what you need and the systems involved. We reply with proposed next steps.

Describe what you need