Skip to content
Evidof

Learn with EvidofOS

Build and maintain your evidence library

Keep claims connected to sources, dates and limitations so colleagues can reuse them responsibly.

Published version 2 · Updated

Start with a claim

Write a specific statement you expect to use in a proposal or campaign. Identify its supporting document, record or approved source. People-served counts need a reporting period and a definition of served. Stories need permission and context. A program description supports what you do but does not by itself prove an outcome.

Organize the source

Open Evidence and add the permitted source using the available upload or record workflow. Use a recognizable title and connect it to the correct organization or program. Record the date, period and context in the available fields. Distinguish a public link, an internal document and text intended for publication.

Inspect the source and confidence behind profile information. These records describe fictional BrightPath training data; they are not evidence of real outcomes.

Inspect the source and confidence behind profile information. These records describe fictional BrightPath training data; they are not evidence of real outcomes.

Inspect the source and confidence behind profile information. These records describe fictional BrightPath training data; they are not evidence of real outcomes.

Inspect the source and confidence behind profile information. These records describe fictional BrightPath training data; they are not evidence of real outcomes.

Review extracted facts

Compare suggestions with the original. Check names, units, dates, totals and whether a figure is a goal or result. Preserve disagreements between sources for the responsible person to resolve. Do not average incompatible figures or choose the more impressive number because it reads better.

Reuse with boundaries

Select evidence relevant to the specific campaign or application. Summarize only what it supports; keep private records out of public copy. Internal planning evidence may not be suitable for donor publication. Review unsupported implications as well as numbers: a service description does not establish medical effectiveness or guarantee a beneficiary outcome.

Keep evidence current

Assign someone to periodically check frequently reused records. Update stale references through the reviewed process, preserving history where provided. Before an important publication, reopen selected sources and confirm that their dates and meaning still fit. If validation fails, repair the claim or source relationship instead of repeatedly pressing Approve.