Support service
Know what every line of the cloud bill buys
When the bill grows faster than usage and nobody can explain it line by line, the savings are usually in plain sight. We break the cost down, apply the changes and leave a monthly tracking model.
- Duration
- 2 to 6 weeks
- Model
- Analyse and apply
- Output
- Savings applied and a tracking model
Signs the cloud bill is out of control
Cloud cost problems are rarely one big mistake. They are many small decisions that nobody revisits.
- The bill grows every month, faster than customers or traffic.
- Nobody can say which team, product or customer a cost line belongs to.
- Test and staging environments run around the clock at production size.
- Instances were sized for a launch peak years ago and never resized.
- Data transfer and storage costs keep rising, and nobody knows why.
- Reserved capacity and commitments were bought once and never matched to actual use.
- Finance asks engineering to cut costs, and engineering cannot see which cuts are safe.
Cost is an architecture property
Cloud cost is often treated as a finance problem, solved with discounts and budget alerts. Those help, but the largest savings sit in engineering decisions: how services scale, where data moves, what runs idle, and which managed services are used for what.
We start by making cost visible. Resources are tagged and allocated to teams, products and environments, so every line has an owner. Then we look for waste, oversizing, idle capacity, expensive data paths and commitment gaps, and estimate the saving and the risk of each change.
Changes are applied with your team in order of saving and safety, and each one is measured on the next bill. A monthly tracking model keeps the gains: cost per customer or per transaction, reviewed by the people who can act on it.
Scope
Included
- Cost breakdown by team, product, environment and service
- Tagging and allocation rules
- Rightsizing of compute, databases and storage
- Scheduling and removal of idle resources
- Data transfer and storage lifecycle review
- Commitment and reserved capacity plan
- Architecture changes with the largest cost effect
- Monthly tracking model and unit cost metrics
Not included
- Negotiating contracts with cloud providers
- Moving to a different cloud provider
- Ongoing cost management as an outsourced service
- Accounting and invoicing processes
How the work runs
- Week 1Make cost visible. Billing data, the resource inventory and usage metrics are collected, and allocation rules are set so that every cost has an owner.
- Weeks 2 to 3Find the savings. Waste, oversizing, idle capacity and expensive data paths are identified. Each saving is estimated with its risk and effort.
- Weeks 3 to 5Apply the changes. Changes are applied with your team, safest and largest first, and verified in monitoring before the next one starts.
- Final weekKeep the gains. The tracking model, unit cost metrics and review routine are handed over to the people who own the budget and the systems.
What you receive
All changes are made in your accounts, together with your team.
Who this service is not for
- Companies whose cloud spend is small compared with the cost of the analysis. If that is the case, we will tell you in the first call.
- Teams looking only for a discount negotiation. The durable savings are in engineering decisions.
- Organisations that want cuts regardless of reliability. We do not trade stability for a lower bill.
Frequently asked questions
How much can we save?
It depends on how the system is built and run, so we do not promise a percentage in advance. The first week’s breakdown shows where the savings are and roughly how large they are.
Will cost changes affect reliability?
They should not. Each change is assessed for risk, applied in order of safety and verified in monitoring before the next one.
Which clouds do you work with?
AWS, Azure and Google Cloud. We work in your accounts with read access first, and change access only for the changes you approve.
Do you need access to our billing data?
Yes. Read access to billing and usage data is needed for the breakdown. It is granted for the duration of the engagement and removed afterwards.
How is this different from Cloud and Platform Engineering?
This service lowers the cost of what you already run. Cloud and Platform Engineering builds the platform itself, with cost designed in from the start.
Start with a short technical call
Thirty minutes. You share the shape of your bill, and we tell you where the first savings are likely to be.