Skip to main content

AI Initiative Economy

Purpose​

This document describes how AI Conveyor defines, estimates, and accounts for the target and actual economic impact of AI initiatives.

Key ideas​

  • Impact types: revenue growth, cost reduction, risk reduction (including regulatory), improvement of customer metrics — each with defined units of measurement and a calculation method.
  • Target impact: estimated at the idea/experiment stage; refined at the prototype stage; recorded in the registry for comparison against the actual figure.
  • Actual impact: calculated using an agreed methodology after going to production; attribution (A/B tests, control groups, calculation models) for a well-grounded attribution of impact to the initiative.
  • Aggregation: consolidated impact across the portfolio and broken down by departments/domains for reporting.

How it works​

The model is used when filling in the "target/actual impact" fields of the initiative entity and in the measure-ai-impact playbook. Calculations and assumptions are documented for audit and for alignment with finance/business. Results feed into portfolio KPIs (portfolio-kpis) and cases. Value realization details are in value-realization.