9. Quality Assurance and Oversight
Standards alignment, validation, and review controls.
9.1 Standards Alignment
The grid carbon intensity factors applied by this methodology are provided by GPX Grid, as described in the Greenpixie Methodology for Cloud Emission Measurement, which is aligned to the GHG Protocol, prepared with reference to ISO 14064-1, and independently verified in accordance with ISO 14064-3 at a limited level of assurance. The carbon grade is assigned from those factors on a location-based basis in accordance with the GHG Protocol Scope 2 Guidance.
The scoring layer described in this document produces relative positions within a distribution rather than a greenhouse gas statement. It is not independently verified at the time of publication and should not be presented as such.
9.2 Data Source Verification
The provider price lists are public, and Greenpixie maintains the translation of each list into the common shape the scoring operates on. GPX Grid's methodology and data selection are validated in collaboration with academic partners, as set out in the cloud methodology. Baseline water stress figures are taken from Aqueduct 4.0 as published by the World Resources Institute, and Greenpixie applies no adjustment to them.
9.3 Calculation Controls
Three controls limit the influence of a bad price on a published score, being the use of the median in place of the mean wherever prices are combined, the removal of prices inconsistent with their product category under Section 6.2, and the suppression of a pricing score for a region below the sufficiency level under Section 7.3.
None of the three detects a provider price list error that is internally consistent across all regions of that provider. The methodology holds no price reference outside the published lists, so an error of that kind is outside the reach of these controls.
9.4 Run Diagnostics
Greenpixie records, for every run, the number of prices removed as inconsistent with the consensus for their product category, the product categories excluded in full for persistent inconsistency, the regions whose pricing score was suppressed, the regions matched to a finer Aqueduct area, the regions given a country-level figure under the country fallback, the regions carrying no water stress category, and any region for which no grid carbon intensity factor was found. The methodology team reviews these counts against previous runs, and a movement in any of them indicates a change in the input data rather than a change in cloud prices or in water stress.
Where a score is unavailable, the output field is confirmed empty before publication, so that a missing figure is never published as a number.
The band boundaries used for each provider and for the pooled carbon distribution are recorded and published alongside the scores of the version that produced them.
9.5 Internal Review
Methodology changes, changes to the comparison basket selection rules, a region newly gaining or losing a score, and the adoption of a republished edition of a third-party data source are subject to internal review before publication.