Published methodology
How every score is computed
Every number the toolkit produces is deterministic arithmetic you can recompute on paper. There is no model, no hidden coefficient, and no generated rating: your team supplies the judgments, and the platform aggregates them consistently. This page lists every formula and threshold in the product.
Four principles
- Recomputable. Any score can be checked by hand from the inputs shown beside it.
- Visible inputs. Every rating traces back to values your team entered, never to values the platform inferred.
- Practitioner-standard scales. The 5x5 risk matrix, percent-of-target KPI progress, and checklist completion are the conventions auditors and M&E frameworks already accept.
- Your judgment, our arithmetic. The platform never rates severity, likelihood, or impact. People do. The platform standardises how those ratings combine.
Community needs priority score
score = (severity + alignment + urgency + feasibility + reputation + sustainability) / 30 x 100
Six dimensions, each rated 1 to 5 by your team: need severity, strategic alignment, community urgency, implementation feasibility, reputation and risk relevance, and sustainability potential. The six ratings carry equal weight. The sum (maximum 30) is scaled to 0-100 and rounded.
Worked example: ratings of 5, 4, 4, 3, 4, 3 sum to 23, and 23 / 30 x 100 = 77, a High priority.
| Range | Status | Meaning |
|---|---|---|
| 70-100 | High priority | Act on it |
| 45-69 | Medium priority | Plan for it |
| 0-44 | Lower priority | Monitor |
Risk rating
rating = likelihood (1-5) x impact (1-5)
The standard 5x5 risk matrix used in ISO 31000-style registers. Your team rates likelihood and impact; the product multiplies them into a 1-25 rating.
| Range | Status | Meaning |
|---|---|---|
| 15-25 | Red | Escalate and mitigate |
| 8-14 | Amber | Actively manage |
| 1-7 | Green | Monitor |
KPI progress
progress = (actual - baseline) / (target - baseline) x 100, clamped to 0-100
Standard percent-of-target-achieved. Progress below zero shows as 0, above target as 100. If target equals baseline, progress is 100 once the actual reaches the target, otherwise 0.
| Range | Status | Meaning |
|---|---|---|
| 80-100 | Green | On track |
| 40-79 | Amber | At risk |
| 0-39 | Red | Off track |
Budget utilisation and variance
utilisation = actual / approved x 100; variance = approved - actual
Reported raw, with no rating applied: your governance thresholds decide what over- or under-spend means.
Evidence completeness
completeness = items collected / items on the checklist x 100
Each programme carries an evidence checklist your organisation defines (photos, attendance sheets, sign-offs, and so on). The dashboard flags a portfolio below 80 percent.
Strategy completion
completion = sections with substantive content / 12 x 100
The strategy workbook has twelve sections. A section counts as complete once it holds substantive content (twelve characters or more), so a placeholder word cannot inflate the score.
Disclosure readiness score
score = checks passed / 10
Ten pass/fail checks on the live portfolio, mirroring what sustainability disclosure standards (the IFRS S1 pillars for social topics, SASB community-relations topics, GRI 413) and their assurers expect from the community-investment domain:
- An active strategy is in place
- The strategy is at least 80% complete
- Every active programme has at least one KPI
- Every KPI has a target distinct from its baseline
- Evidence completeness is 80% or above
- Every open risk has a mitigation and an owner
- At least one scored needs assessment exists
- Every stakeholder has an engagement cadence and a last-engagement date
- Every active programme has budget lineage
- A monthly report covers the current or previous month
The score signals how ready this portfolio is to feed an organisation's wider sustainability reporting. It is not a compliance certification, and it covers the social-investment domain only.
About the thresholds
The band boundaries above (70/45 for needs, 15/8 for risk, 80/40 for KPIs, the 80 percent evidence flag) are fixed, published defaults chosen to match common practice. They are conventions, stated openly rather than tuned per customer. If your governance framework requires different boundaries, tell us: configurable thresholds are on the roadmap and prioritised by real demand.
Questions or challenges to the methodology are welcome at hello@csitoolkit.app.