Quantifying Non-Profit Impact in the Age of Artificial Intelligence

This methods preprint develops Net Attributable Value (NAV) for non-profit support to micro and small enterprises, combining explicit counterfactuals, contribution margins, attribution, displacement, persistence, survival and costs. It separates private contribution value from a displacement-adjusted producer-and-provider account. A stylised Indian printing-shop appraisal and a Monte Carlo comparison of revenue estimators illustrate the framework. All numerical exercises use simulated data and illustrative assumptions; the paper does not establish realised enterprise impact or a causal benefit of AI. Portfolio precision and detection thresholds are conditional on the specified data-generating process. The proposed measurement protocol and evidence grades require field validation. The deposit includes the corrected manuscript in PDF and editable DOCX, analysis.py, results.json, README.md and requirements.txt. The supplied code is a newly documented implementation of the stated model; the initial manuscript’s original script was unavailable. The revised manuscript reports the recomputed outputs and discloses AI assistance and the author’s organisational interest. This is a preprint and has not been peer reviewed.