CS-PMI methodology

The CollectiveSpend Procurement Maturity Index (CS-PMI) is a 0–4 score summarising the maturity of your procurement function across 11 dimensions. This page explains how it's computed and what each tier means.

The scale

Every question is scored on a 0–4 ordinal scale:

• 0 — Reactive: practice is absent or ad-hoc.
• 1 — Emerging: informal, inconsistent, undocumented.
• 2 — Defined: documented; applied in some areas.
• 3 — Managed: documented; broadly applied; reviewed periodically.
• 4 — Leading: documented; universal; continuously improved with measurable outcomes.

Aggregation

Each dimension score is a weighted average of its questions. Weights live in the question library (default 1.0); we use them to dampen questions that are checkpoints rather than capabilities. The overall CS-PMI is the mean of dimension scores — equal-weighted across the 11 dimensions so no single area dominates the headline number.

Confidence-weighted score

For every AI score, the scoring model emits a confidence value (0.0–1.0) based on how strong the cited evidence is. We surface the confidence-weighted score next to the raw score: actual × confidence. It reveals where the model is hedging — useful when a dimension's headline looks healthy but the supporting evidence is thin.

Coverage

Three coverage metrics surface in the assessment header:

• Evidence coverage — % of in-scope questions with at least one cited evidence chunk.
• ISO 9001 coverage — % of ISO 9001-tagged questions scored.
• ISO 20400 coverage — same for ISO 20400.

Use coverage to spot underweight evidence libraries. A score of 2.5 on 50% coverage is a different story than 2.5 on 95% coverage.

Tier mapping

The overall score maps to a named tier so the headline reads clearly:

• 0.0 – 0.99 → Reactive
• 1.0 – 1.99 → Emerging
• 2.0 – 2.99 → Defined
• 3.0 – 3.49 → Managed
• 3.5 – 4.0 → Leading