Cost and carbon
Carbon estimates you can check line by line
GetFinOps estimates kWh and CO₂e for the instances and EBS storage your AWS and GCP scans flag, and exports a GHG Protocol Scope 3 Category 1 report showing every factor it used.
It doesn’t estimate your whole cloud footprint, only resources your scans flag.
The problem
Carbon reporting is often a yearly spreadsheet built far from the teams changing infrastructure. When a figure can’t be traced back to its inputs, nobody can check it, and nobody uses it to decide anything.
How it works
Scan
Your regular scan collects AWS and GCP inventory and runs the rules, which flag the resources worth your attention.
Estimate
For each flagged EC2, RDS, Compute Engine or Cloud SQL instance, or EBS volume or snapshot with a known size, GetFinOps estimates power draw, converts it to monthly kWh with the provider’s PUE, and applies the region’s grid factor.
Report
See totals, top emitters, a regional heatmap and a trend across recent scans, then export the Scope 3 report for your inventory.
23 and 37
AWS and GCP regions with factors, each with its published source and year
Scope 3
GHG Protocol Category 1 report, exported as JSON, CSV or PDF
By hand
each tCO₂e figure can be recomputed from its report row’s kWh, factor and PUE
480
gCO₂e/kWh fallback where none is published; its estimates lower data quality
Capabilities
CO₂e estimates for flagged AWS and GCP instances and EBS storage, exported as a Scope 3 report.
Estimates on findings
Findings for EC2, RDS, Compute Engine and Cloud SQL instances, and for EBS volumes and snapshots with a known size, carry estimated watts, kWh and CO₂e per month next to their dollar figures.
Region factors with sources
Each estimate uses the grid carbon intensity of the resource’s region and the provider’s PUE. Factors ship for 23 AWS and 37 GCP regions, each with its published source and year.
A Scope 3 report
Export a GHG Protocol Scope 3 Category 1 report as JSON, CSV or PDF, with one row per provider, region and service showing kWh, emission factor, PUE and tCO₂e.
Data quality you can see
Every estimate records whether it used a fallback power figure or region factor. The share of resources relying on fallbacks becomes the report’s data-quality figure, next to its exclusions and attestation.
In depth
What gets an estimate
Estimates are attached to findings during each scan. A finding gets one when its resource is an EC2 instance, RDS database, Compute Engine instance, Cloud SQL instance, or an EBS volume or snapshot with a known size, and its region is known. Each estimate stores its inputs, including provider, region, resource type, CPU figure and hours, so the calculation can be traced.
Resources that no rule flagged are not estimated. The totals describe the part of your estate your scans picked out, not your whole footprint.
How an estimate is calculated
Power draw comes from a catalog of idle and full-load watts per instance type, based on Cloud Carbon Footprint coefficients, and is interpolated at the resource’s CPU utilization. A measured figure on the finding is used when there is one; otherwise 50% is assumed. Types outside the catalog use a per-vCPU fallback.
Watts times 730 hours times the provider’s PUE gives kWh per month, and kWh times the region’s grid intensity gives grams of CO₂e. EBS storage uses 1.2 watts per terabyte instead of an instance power figure.
Region factors
Grid intensity is stored per region: 23 AWS regions cited to the AWS Customer Carbon Footprint Tool and 37 GCP regions cited to Google Cloud’s regional carbon data, each dated 2023. PUE is 1.135 for AWS and 1.10 for GCP.
A region missing from the table, or a provider other than AWS or GCP, uses a global fallback of 480 gCO₂e/kWh with a PUE of 1.125, and the estimate is marked as a fallback. Because factors are annual averages, moving a workload to another region changes its estimate, but the time of day does not.
The Scope 3 report
The report follows the GHG Protocol Corporate Value Chain (Scope 3) Standard for Category 1, with operational control as the boundary. It is built from your latest completed scan, or a scan you choose, labelled with a reporting month, and always covers the whole workspace. Reporting months run from the first of one month to the first of the next in UTC, and a workspace with no completed scan still gets a well-formed report with zero totals.
- Monthly totals in grams, kilograms and tonnes of CO₂e, plus kWh.
- One row per provider, region and service: kWh, emission factor, PUE, factor source, tCO₂e and the share of resources using fallbacks.
- A data-quality figure, a list of exclusions, the method and formula, and an attestation that the figures are estimates.
- JSON for other systems, CSV for spreadsheets, or PDF.
On the dashboard
The Sustainability page shows monthly CO₂e and kWh, the ten highest-emitting resources, a regional heatmap, emissions by cloud and a trend across recent scans. It follows the cloud account selector, so you can look at one account or all of them. A reduction figure adds up the CO₂e of the resources the scan flagged, counting each resource once.
Each top emitter carries its rule, resource, region, kWh and whether a fallback was used, and the trend data includes the change between the two most recent scans.
What it does not do
- It doesn’t estimate your whole cloud footprint, only resources your scans flag.
- It doesn’t count a resource twice when several rules flag it; it keeps the estimate with the best evidence.
- It doesn’t estimate object storage, serverless, networking, GCP persistent disks or Kubernetes workloads, and it leaves out embodied emissions from hardware.
- It doesn’t use metered electricity or hourly grid data; figures are modelled from power coefficients and annual averages.
- It doesn’t cover Azure or other clouds.
- It doesn’t raise findings or make changes on carbon grounds alone.
FAQ
Questions
Which GHG scope does this cover?
Scope 3, Category 1: purchased goods and services, for the cloud infrastructure you buy. The report uses the operational control approach across the cloud accounts in your workspace.
Does it measure all of my cloud usage?
No. Estimates cover resources that at least one rule flagged in your latest completed scan, and among those only compute instances, databases and EBS storage. Treat the report as a partial inventory and disclose its exclusions.
Which clouds are covered?
AWS and GCP. Resources in regions without a published factor use a global fallback and count against the data-quality figure.
How accurate are the numbers?
They are estimates, not metered electricity. Power comes from a catalog based on Cloud Carbon Footprint coefficients, and grid intensity from annual regional averages dated 2023. The data-quality figure shows how much relied on fallbacks.
Try free — no credit card
Create a workspace and connect an AWS or GCP account, or book a demo first.