AWS Cost Explorer shows you cost and usage trends across up to 13 months of history with a 12 month forecast, broken down by service, account, region, instance type, and any cost allocation tag you have activated. It also carries the built in Savings Plans and Reserved Instances utilization and coverage reports and the rightsizing recommendations. What it misses is the precise, line item truth: hourly and resource level detail are a paid add on, data transfer and idle network charges sit lumped in a bucket called EC2 Other, and any spend you have not tagged is simply unallocated. Use it to find where to look, then use the Cost and Usage Report to prove what is actually happening.
This is the most used cost tool in AWS and the most misread. Below is what it answers well, where it goes quiet, and the moment you should stop trusting the dashboard and open the raw data.
What does AWS Cost Explorer actually show you?
Cost Explorer is a visual analytics layer over your billing data. In the console it gives you daily and monthly granularity for free, with grouping by service, linked account, region, usage type, instance type, purchase option, and activated cost allocation tags. It carries a rolling 13 months of history and a forecast that projects up to 12 months ahead from your recent run rate.
Three built in reports earn their keep. The Savings Plans coverage and utilization reports tell you how much of your eligible compute sits under a commitment and how much of each commitment you actually consumed. The Reserved Instance equivalents do the same for RIs. The rightsizing recommendations flag idle and oversized EC2. These are genuinely useful starting points, and they cost nothing to read.
For most buyers the honest summary is this: Cost Explorer answers the question of which service, account, or team is driving a trend. That is the right first question. It is rarely the last.
Where does Cost Explorer fall short?
The gaps are not bugs, they are design choices, and each one hides a different class of spend.
Granularity costs extra and rounds early. Daily and monthly views are free, but hourly and resource level granularity is a paid feature billed per usage record (indicative, check the current pricing page). Without it you cannot see a spike that happened between 2am and 4am, and you cannot attribute cost to a single resource id. The console also rounds, so small but numerous line items disappear into totals.
Network and data transfer hide in EC2 Other. NAT gateway processing, inter region transfer, and load balancer data are folded into a usage group most teams never expand. A bill that looks like compute is often a quarter network once you open the raw data.
Allocation depends entirely on tags. Cost Explorer can only split cost by a tag you activated and that the resource actually carries. Untagged spend lands in a no tag bucket, so a team that owns 30 percent of the bill can look like it owns 10 percent. Cost Explorer will not tell you what is missing.
It reports, it does not reconcile. Amortized versus unblended, blended versus net of credits, Marketplace charges, and refunds all behave differently in Cost Explorer than in the invoice, and the tool will not walk you from one to the other. For chargeback you need the line items.
Cost Explorer versus the Cost and Usage Report
The Cost and Usage Report, the CUR, is the source of truth. It is the full set of hourly line items, every resource, every rate, every discount, delivered to S3 and queryable in Athena or your warehouse. Cost Explorer is a curated view on top of that data. The table below is the decision most teams face.
| Question | Cost Explorer | Cost and Usage Report |
|---|---|---|
| Which service drove last month up? | Yes, in seconds | Yes, but slower |
| What ran at 3am on one resource? | Only with paid hourly granularity | Yes, native hourly and per resource |
| Split NAT and transfer from compute | Buried in EC2 Other | Yes, by usage type |
| Chargeback to teams with audit trail | Limited by tags, no line items | Yes, full line items |
| Effective savings rate from first rates | No | Yes |
The rule of thumb: explore in Cost Explorer, prove in the CUR. If a number is going into a budget, a chargeback, or a vendor conversation, it should come from the line items.
A worked example: the gap a dashboard hides
A European SaaS company read its EC2 line in Cost Explorer at roughly 38 percent of the monthly bill and treated compute as the priority. Opening the CUR told a different story. Inside the EC2 Other bucket, NAT gateway data processing and inter region transfer together made up 11 percent of the total, more than the entire database tier. None of it was visible in the default Cost Explorer view, and none of it was tagged to a team. Re routing traffic through VPC endpoints and consolidating cross region chatter cut that 11 percent by more than half. Figures are verified against billing data and anonymised.
The lesson is not that Cost Explorer was wrong. It is that the default view answered the wrong question, and only the line items showed where the money actually went.
How should you use Cost Explorer in a real cost program?
Treat it as the triage layer. Use the free daily views and the commitment reports to spot trends and obvious waste, set up AWS budgets and alerts that work on top of it, and let AWS Cost Anomaly Detection in practice watch for spikes you would otherwise miss. Then, for anything that drives a decision, build it from the CUR so allocation, granularity, and rate detail are all defensible. That two layer habit, fast triage plus auditable proof, is what separates a dashboard from a cost program. It sits at the centre of the broader AWS cost optimization guide.
Frequently asked questions
Get the buyer side AWS cost playbook
We read your Cost and Usage Report the way an independent buyer does, surface the spend Cost Explorer rounds away, and turn it into a plan your engineers will accept. It is the same discipline behind the 31 percent median reduction we see in the first 90 days, with zero provider commissions. Our guarantee: we reduce your cloud spend or we reimburse our service fee.
Put a defensible number on your cloud spend.
No provider in the room, no published price list. Tell us your footprint and we will scope the savings against your billing data — we reduce your cloud spend or we reimburse our service fee.
The Cloud Spend Navigator: what changed in cloud pricing, commitments, and FinOps — no vendor spin.