How to read these scores
The Fourth Cloud instrument scores functions within layers, not layers as aggregates. A vendor can score 4 on one function within a layer and 0 on another — that profile is more useful than any aggregate.
The 0–4 gradient
| Score | Label | Definition |
|---|---|---|
| 4 | Hyperscaler | AWS/hyperscaler equivalent — fully managed, fully automated. |
| 3 | Strong | Meaningful automation and integration. Narrow, well-understood gaps. |
| 2 | Moderate | Partial coverage within constraints the vendor does not control. |
| 1 | Weak | Addressed for some workload types but not others; manual-assisted. |
| 0 | Absent | Vendor provides nothing. Enterprise owns the function entirely. |
AWS is the benchmark at 4. VCF is the on-prem anchor at FC-2A F1=3 and F2=2. Scores are calibrated across vendors so a 3 means the same thing on every row.
Gap ownership (every score below 4)
| Type | What it means |
|---|---|
| Closeable | Enterprise must acquire new capability — new software, vendor, or contract. |
| Opinion | Primitives exist; enterprise applies configuration without new acquisition. |
| Vendor roadmap | Vendor has announced product intent with a timeline. |
| Structural | Consequence of the on-prem operating model — no near-term close. |
| Mixed | Different ownership for different parts of the function. |
The critical distinction between Closeable and Opinion: Closeable requires the enterprise to acquire new capability — new vendor, new contract, new software. Opinion requires applying existing capability — configuration of platform primitives already shipped. When in doubt: does closing this gap require a new contract or new software the enterprise does not already have? If yes, it is Closeable.
Scoring discipline
- Score functions, not layers. The layer is a grouping. The function score is the finding.
- Every score below 4 gets a gap ownership classification.
- Workload universality applies at every layer. If a function covers AI workloads but not traditional workloads, score it at its actual coverage level.
- The delivery model is part of the score. A managed consumption model (GreenLake, APEX) changes FC-2A F2. Score what the vendor actually sells, not the software in isolation.
- FC-2C receives the strictest scrutiny. The integration vs. coexistence litmus and the workload universality litmus both apply before a score is set.
- Identity plane continuity is a separate, top-level finding. Each layer can score well while the identity chain between layers is broken.
Relationship to the 4+1 AI Infrastructure Assessment
The 4+1 instrument (layer2c.web.app) evaluates AI infrastructure completeness. The Fourth Cloud instrument evaluates operational sustainability across AI and non-AI workloads.
The FC layers do not map to the 4+1 layers. Do not crosswalk the scores. When a vendor scores identically on a 4+1 layer and its FC counterpart, that is a coincidence to explain, not an expected outcome. When the scores diverge, the divergence is the finding.