Measuring Impact

Purpose

A scenario for recording and reporting the economic impact of an AI initiative after it goes into production (the "Business Impact" stage).

Core Ideas

• Impact is measured according to a predefined methodology and metrics (agreed upon at the idea/prototype stage).
• Attribution matters: how to attribute the achieved result specifically to the AI solution (A/B tests, control groups, calculation models).
• The results are used for portfolio reporting, scaling decisions, and case studies.

How It Works

1. Baseline measurement plan: which metrics, data source, recalculation frequency; attribution method (see value-realization, economic-impact-model).
2. Data collection: obtaining actual figures for business metrics and technical KPIs.
3. Impact calculation: applying the economic impact model, comparing against the target value.
4. Documentation: recording the actual impact in the initiatives registry; if needed, a case for case studies.
5. Gate 5: decision to close the initiative as having met its goal, to scale, or to adjust.

Portfolio metrics are updated in line with portfolio-kpis.