Impact evaluation
When tools are used in prevention, separation support or client work, organizations may want evidence of change at program level. This page summarizes how evaluation can be approached without turning consumer tools into surveillance. It is not intended as a statistical methodology manual.
Partnership with evaluoi.ai
We work with evaluoi.ai to define, measure and evaluate outcomes together with the people the work concerns, where that collaboration fits the program’s goals and ethics review.
evaluoi.ai is described on this site as an Impact Intelligence platform for organizations.
Aggregates, not individual surveillance
Impact evaluation is intended to rely on aggregated data and designs that reduce the risk of identifying individual users. Measurement is not intended for individual surveillance, profiling or clinical assessment of end users.
Individual user data from consumer applications is not intended to be shared with the user’s organization.
Transparency with participants
Where evaluation is part of a program, participants should understand what is measured, for what purpose, and who is responsible for safeguarding. Organizational deployments should align consent and data flows with applicable law and professional standards.
Limits of claims
We avoid overstating causal claims. Evaluation is intended to support learning and proportionate decisions—not to replace ethical judgment or regulatory obligations on the organization’s side.