To track cloud server costs, review daily spend and the monthly forecast alongside each server's status, workload and the websites it supports. For agencies and web teams managing multiple sites, make the review about specific resources and their purpose, rather than a single account total. Use the figures to decide what needs investigation, not as an automatic instruction to reduce capacity.
What should you include in a cloud cost review?
Build a review around five questions. Keep the answers together in your working notes so the person investigating a cost change can also see the operational context.
- Spend: what has the account spent so far, and what is the forecast for the month?
- Resource: which server, database or service needs a closer look?
- Status: is the resource running, stopped or reporting a problem?
- Workload: what do the available utilisation charts show over the period being reviewed?
- Purpose: which websites depend on the resource, who owns the decision and what should happen next?
TLDTrack brings cloud resource status, workload, costs and domain dependencies into one view for connected AWS, Azure and DigitalOcean accounts.
How do you establish which servers you are paying for?
Start with one cloud account and establish a resource list before investigating individual costs. Record a recognisable resource name, the account it belongs to, its intended purpose and a responsible person. Include databases in the review rather than limiting the exercise to web servers.
TLDTrack connects to cloud accounts with read-only access and keeps the discovered server and database list up to date as resources come and go.
For your own review, label the purpose clearly: live client hosting, staging, development or an infrastructure service, for example. Treat an unfamiliar resource as a question for its owner. Do not assume that an unclear name means it is safe to remove.
How should you compare spend and forecasts?
TLDTrack shows daily spend by service in pounds, dollars or euros, a monthly forecast, budget and spike alerts, and estimated costs for individual servers.
Review the account total first, then investigate the services behind any change. Compare equivalent periods and record the dates you used. Keep the forecast labelled as a forecast and per-server estimates labelled as estimates; do not present either as a final invoice amount.
Choose a budget threshold that prompts a useful conversation with the account owner. When an alert arrives, ask whether the increase was planned, which service changed and who can explain it. Avoid treating every increase as waste or every stable total as a reason to stop checking.
How do status and workload inform a cost decision?
TLDTrack checks supported cloud resources every five minutes and alerts when a resource stops or fails. TLDTrack also provides CPU, network and CPU credit charts, with chart periods of 24 hours, seven days or 30 days.
Use those signals to frame an investigation rather than jumping straight to a capacity change. Consider these review paths:
- If spend rises alongside workload, ask the owner whether the extra activity was expected and whether the forecast remains acceptable.
- If spend rises without an obvious workload change, inspect the service-level breakdown and recent account changes before drawing a conclusion.
- If utilisation looks low, review a longer period and confirm the resource's purpose before proposing a smaller configuration.
- If a resource is stopped or failing, resolve its operational purpose and current condition before deciding what to do about its cost.
Which websites depend on the server?
TLDTrack suggests domain-to-server links from DNS and displays the linked server on the domain's page.
Check those suggested links against your own infrastructure records. Before approving a change, identify the affected sites, any database dependencies and the person authorised to accept the risk. For shared resources, agree how you will discuss costs with clients rather than assigning the whole estimate to one website.
How do you turn the review into action?
Set a regular review slot and give each investigation an owner, a next action and a review date. Separate monitoring from approval: ask the responsible technical person to assess any proposed configuration change, document a rollback approach and agree how to check the affected websites afterwards.
Start with one account and one review period. Use TLDTrack's cloud server monitoring page to review the connection options, then compare spend, status, workload and linked domains before choosing your first follow-up action.
