AI4LDWe IlluminateILDF 2.0 Tools
ILDF 2.0 Phase 4 · Evaluation: Broader Impact

KPI & Impact Measurement: Contribution, Not Attribution

Connect an instructional design to the KPIs stakeholders actually care about — with the causal chain, proxy distance, and rival explanations named for every candidate indicator before the meeting, not during it. This drafts a measurement architecture. It never manufactures the evidence itself.

Causal-Chain Ledger
Contribution, Not Attribution
Metric Fixation Guard
The Core Tension

Why this needs a guardrail, not just a dashboard

AI affordance
A model can map an instructional design to the organization's measurable indicators in minutes, draft the full causal chain for each one, and enumerate the rival explanations a skeptical CFO will raise — before the meeting, not during it.
Named risk
False Precision — an impact dashboard that looks rigorous because it is numeric, while every number is a proxy several causal steps from the learning it claims to measure.
Guardrail
Every KPI carries its causal chain, its proxy distance, and a plain statement of what it cannot show. Contribution framing only — attribution (“training caused this result”) requires a comparison design this tool will name but cannot conjure.
Human judgment domain
Which organizational outcomes actually matter, what the politics of measurement are in your organization, and whether to act on a number. The tool drafts the architecture; you own the claim.
What this tool does not replace: A real evaluation design with comparison conditions, or the conversation with stakeholders about what success means.
What I can and can't do: I draft a measurement architecture — indicators, causal chains, proxy distances, data sources, and measurement windows — from what you tell me about the design, the outcomes, and the organizational goal. I never invent baseline values, percentages, or targets, and I never claim the design caused an organizational result. Contribution language only.
Optional — this tool works fully standalone. Pick or create a project only if you want this architecture saved to the shared Design Journey (and, when prior cycles exist for it, offered as context below).
Draft the Architecture

Describe the design, the outcomes, and the goal

Free text — bring what you actually have. A thin input produces a thinner, more honest ledger (more “far” proxies, more named rival explanations), not a padded-out dashboard.

The design itself: format, scale, audience, and where it sits in the organization.

0 / 3000 · minimum 60

Paste from the Formative Evaluation Designer if you have a cycle for this project (see above), or write the outcomes you actually committed to.

0 / 1200 · minimum 20

In their words — even if it's vague. A vague goal becomes the first ledger entry: operationalizing it.

0 / 800 · minimum 10

Named sources only — this tool never invents the numbers themselves from this field.

0 / 1500
This drafts architecture, not evidence
This drafts measurement architecture — indicators, chains, windows. It does not and cannot generate evidence, baselines, or targets.

Nothing you enter is stored on our servers. Your text is processed by Anthropic's Claude to generate this tool's response. See Privacy for details.