11. Outputs and Interpretation

11.1 Primary Outputs

The methodology produces the following outputs for each region of each supported provider, published as one row per data version:

• Carbon grade, a letter from A+ to F, where A+ is the lowest intensity band.
• Water stress category, one of five categories from Low to Extremely High.
• Cost grade, a value from 1 to 5, where 1 is the least expensive band. Suppressed under Section 7.3 where a region prices too little of the comparison basket.
• Availability grade, a value from 1 to 5, where 1 is the most complete band.

Every output is a band. The figures those bands are assigned from, being the region's average grid carbon intensity, its relative cost percentage, its basket coverage percentage, and its service availability percentage, are internal values. The band boundaries of a version are published with the scores that version produced, and how each band is labelled is a display decision taken outside this methodology, as set out in Section 7.7.

11.2 Intended Use

These outputs are intended to support cloud region selection at the point a workload is placed; comparison between candidate regions of a single provider; confirmation that a region prices the services a workload requires; a first read on a region's catalogue price position ahead of a detailed pricing exercise; and a stated basis for preferring one region over another on environmental grounds.

11.3 Interpretation Guidance

Every published value is a band, so a region's position on any of the four scores is known to the width of its band and no closer, and the published boundaries are what make that width visible. Boundaries are set again from the data of each run. A grade can therefore change between versions with no change to a region's own prices, its own grid, or its own water stress, where the distribution around it has shifted. Carbon boundaries are set across the regions of all three providers together, so a carbon grade can also move when regions of another provider enter or leave the distribution. No published value on the scorecard is stable across versions in the way an underlying figure would be.

Carbon grades are comparable across providers, as their boundaries are set once across all three. Cost and availability grades are not, as theirs are set within each provider separately, and a cost grade of 4 on one provider describes a position in a different estate from a cost grade of 4 on another.

The figure behind a carbon grade is the average of the previous calendar month and does not indicate what a grid is doing now or in the coming hours. A window of one month carries seasonal movement, so a region can change carbon grade between versions on a shift in the generating mix alone. A water stress category resting on a country fallback describes a whole country and is a weaker guide than one matched to a finer area. Every product category in the comparison basket carries equal weight and compute categories are the most numerous, so a region whose compute prices sit above its provider's middle tends towards a higher cost grade even where its storage prices sit below it.

11.4 Non-Intended Use

The outputs should not be treated as certified measurements, contractual guarantees, or regulatory compliance instruments unless explicitly agreed in writing. They should not be used as a forecast or an approximation of a bill, as no customer usage or spend enters their calculation, and cost and availability grades should not be compared across providers.

The water stress category describes the pressure on water supply in the area a region occupies and is not a measure of water consumption. The carbon grade describes the grid, and no emissions total for a workload can be produced from it, nor is it a scheduling signal. Water consumed in cooling a workload and in generating the electricity it draws, together with the emissions of that electricity, are calculated per unit of cloud usage under the Greenpixie Methodology for Cloud Emission Measurement.