| name | impact-diligence |
| description | Use when evaluating a fund manager, direct deal, or portfolio company against impact criteria — especially for impact-aligned, regenerative, or natural-capital transactions in VC, PE, real assets, or fund-of-funds contexts. Applies the Impact Management Project (IMP) five-dimensions framework, suggests IRIS+ metric IDs from the GIIN catalog, maps to UN SDG targets, and runs an additionality stress-test. Output is structured for IC memos and LP-facing reports. |
Impact diligence — IMP five dimensions + IRIS+ mapper
When to use this skill
Trigger this skill when any of the following appear in the conversation:
- A fund manager's impact thesis is being assessed
- A direct deal or co-invest opportunity claims impact alignment
- A portfolio company's impact report needs to be evaluated for credibility
- A new investment is being scored against an impact lens (regen, climate, biodiversity, livelihoods, etc.)
- The user asks "is this real impact?" or "what's the actual impact thesis here?"
If the conversation is about measuring already-deployed impact (vs evaluating a prospective deal), still use this — the framework applies symmetrically.
What this skill produces
A structured impact diligence assessment with five components:
- IMP five-dimensions scorecard (What / Who / How Much / Contribution / Risk)
- IRIS+ metric suggestions — concrete metric IDs from the GIIN catalog that match the deal's impact claims
- SDG mapping — specific SDG targets (not just goals), with confidence levels
- Additionality stress-test — three questions the investor must answer
- Verification quality assessment — how the manager is measuring, and whether it's credible
The IMP five dimensions — the core framework
The Impact Management Project (IMP) framework is the de facto language of institutional impact investing. Every dimension must be addressed for an impact claim to be credible.
1. WHAT — what outcome is occurring, and how important is it?
Questions to answer:
- What specific outcome does this investment produce? (Not "improve health" — which health outcome, measured how?)
- Is the outcome positive, negative, or both? (Most real interventions have both.)
- Is the outcome valued by the stakeholders experiencing it?
- Is it aligned with the SDGs / Paris Agreement / Kunming-Montreal Global Biodiversity Framework?
Red flags:
- Outcomes described only as activities or outputs ("planted 1M trees" tells you nothing about survival, ecosystem services, or beneficiary outcomes)
- No mention of negative outcomes — every real intervention has trade-offs
- Outcomes are sponsor-defined but never validated with stakeholders
2. WHO — who experiences the outcome?
Questions to answer:
- Specifically which stakeholders (geography, demographics, baseline conditions)?
- How underserved are they? (IMP uses an A/B/C/D classification roughly matching expected counterfactual conditions)
- Is the data disaggregated (gender, age, income decile, geography)?
Red flags:
- "Communities" or "smallholders" used without specifics
- No baseline data
- Aggregate metrics that hide distribution (e.g., average income went up but median stayed flat)
3. HOW MUCH — how much of the outcome occurs?
Three sub-dimensions:
- Scale: how many stakeholders experience the outcome
- Depth: how significant the change is per stakeholder (degree)
- Duration: how long the change lasts
Red flags:
- Reports scale (number reached) but not depth (per-stakeholder change)
- No durability assessment — does the outcome persist after the investment exits?
- Headline numbers without unit economics
4. CONTRIBUTION — would this have happened anyway?
The counterfactual question. Three lenses:
- Enterprise contribution: does the investee organization cause the outcome, or just associate with it?
- Investor contribution: does this specific investor add anything the enterprise couldn't get elsewhere? (Patient capital? Networks? Expertise? Concessionality?)
- Stakeholder contribution: would the stakeholders have achieved the outcome through another route?
This is the most contested dimension and the most often hand-waved.
Red flags:
- No counterfactual analysis at all
- Counterfactual assumes the investee wouldn't have raised capital from anyone (rarely true)
- Investor-contribution argument relies entirely on "we have an impact thesis" (so do many others now)
5. RISK — what could prevent the outcome?
Nine types of impact risk (per IMP):
- Evidence risk — outcome may not occur because the evidence base is weak
- External risk — outcome may not occur because of factors outside the investee's control
- Stakeholder participation risk — stakeholders may not engage as needed
- Drop-off risk — outcome reverses after intervention
- Efficiency risk — outcome occurs but at higher cost than alternatives
- Execution risk — investee fails to implement
- Alignment risk — investee's incentives drift away from the impact thesis
- Endurance risk — investee can't sustain the activities long enough
- Unexpected impact risk — negative impact materializes that wasn't anticipated
For a credible impact thesis, the top 3–4 risks should be named and a mitigation should exist.
IRIS+ metric mapping
IRIS+ (Impact Reporting and Investment Standards Plus, from the GIIN) is the most-used catalog of impact metrics in institutional investing. ~600 metrics, organized by Impact Category → Strategic Goal → Core Metrics Set.
How to use: Once the WHAT and HOW MUCH dimensions are clear, suggest 3–5 IRIS+ metric IDs the manager should be reporting. Format: [IRIS+ ID] — [Metric name].
Example mapping for a regenerative ag fund:
PI8330 — Land Directly Controlled: Sustainably Managed
OI1118 — Greenhouse Gas Emissions Reduction (CO2e MT)
PI3389 — Smallholder Suppliers
OD8060 — Soil Organic Carbon Change (if measuring outcomes)
The asset folder includes a CSV of the most relevant IRIS+ categories for the user's typical deal types (regen, natural capital, climate, real assets). See assets/iris-plus-relevant.csv — drop in the full GIIN catalog if needed.
SDG mapping — targets, not just goals
Don't map to "SDG 15: Life on Land." Map to specific targets like "SDG 15.3: combat desertification, restore degraded land and soil." The targets are where the meaningful claims live.
Provide confidence levels:
- High — direct measurable contribution
- Medium — plausible contribution but not directly measured
- Aspirational — alignment in narrative only
Be willing to say "no SDG target maps cleanly" when that's true — better than forcing it.
Additionality stress-test — three questions
If a manager can't answer these, the impact thesis is weak:
- Financial additionality: Would this enterprise have raised capital on the same terms from a non-impact investor? If yes, what does the impact capital add?
- Behavioral additionality: Does the impact capital change the enterprise's behavior (governance, reporting, target-setting, mission lock)?
- Outcome additionality: Would the outcome have occurred without this enterprise existing? (Hardest test — often we can only argue acceleration not additionality.)
Verification quality — the credibility ladder
Rank the manager's MRV (measurement, reporting, verification) approach:
- Self-reported, no third party — lowest credibility, useful only for direction
- Self-reported with third-party review of methodology — middling
- Outputs verified by third party (e.g., B Lab GIIRS) — good
- Outcomes verified by independent third party with site-level sampling — strong
- Outcomes verified + peer-reviewed publication / regulatory standard — gold standard
Most impact funds today sit at level 2–3. Level 4–5 is where the genuinely defensible theses live.
Output structure for memos
When producing the diligence assessment, format as:
## Impact Diligence Assessment — [Fund / Deal Name]
### IMP Scorecard
- WHAT: [outcome description] — [score: clear / partial / vague]
- WHO: [stakeholders, baseline] — [score]
- HOW MUCH: [scale / depth / duration] — [score]
- CONTRIBUTION: [counterfactual] — [score]
- RISK: [top 3 risks + mitigations] — [score]
### IRIS+ Metrics (suggested)
- [Metric ID] — [Name] — [Why relevant]
- ...
### SDG Mapping
- SDG X.Y — [confidence: High / Medium / Aspirational] — [evidence]
### Additionality Stress Test
- Financial: [answer]
- Behavioral: [answer]
- Outcome: [answer]
### Verification Quality
- Current level: [1-5]
- What would move them up a level: [specific recommendation]
### Bottom Line
- Impact thesis strength: [Strong / Adequate / Weak]
- Top 2 things to push on in next conversation: [specific]
- Memo language recommendation: [precise phrasing for IC]
When NOT to use this skill
- For pure financial analysis with no impact dimension — use the financial-analysis skills instead.
- For carbon-credit-specific evaluation — use
carbon-credit-quality.
- For TNFD / nature-related risk specifically — use
tnfd-leap.
- For taxonomy disambiguation in regen ag — use
regen-glossary first, then return to this.
References
The supporting assets/ folder contains:
imp-five-dimensions.md — full IMP framework reference
iris-plus-relevant.csv — curated IRIS+ metrics for typical deal types
sdg-targets-reference.md — full SDG targets list (not just goals)
If those files aren't present, the skill still works from the body above but with reduced specificity on metric IDs.