research-repository

name	research-repository
description	Maintain a reusable research repository with study provenance, participant and method metadata, findings, evidence, tags, decision links, confidence, supersession, and privacy controls.

Research Repository

Use when user research should accumulate into organizational knowledge rather than disappear into individual reports.

Procedure

1. Define the repository's users and retrieval needs before choosing taxonomy or tooling.
2. Store each study with decision context, research questions, method, participant criteria, dates, product or prototype revision, researcher, and limitations.
3. Preserve findings separately from raw or restricted evidence and link each claim back to the study or evidence that supports it.
4. Tag by durable concepts such as user segment, job, workflow, product area, behavior, barrier, or market context rather than temporary project code names alone.
5. Record confidence, conflicting findings, and supersession so newer research can update knowledge without erasing history.
6. Link findings to product decisions or artifacts where useful so teams can see whether evidence influenced action and what happened afterward.
7. Apply appropriate access, retention, consent, and deletion controls to transcripts, recordings, personal data, and sensitive customer material.
8. Periodically audit for stale taxonomy, duplicate studies, unsupported summaries, inaccessible evidence, and knowledge gaps worth researching next.

Decision rules

• A repository should make evidence easier to reuse, not become a dumping ground for slide decks.
• Raw participant data may need tighter access than synthesized findings.
• Old research is not automatically invalid, but product and market context should be visible.
• Never erase contradictory evidence merely to keep one canonical insight tidy.

Quality gate

The repository is healthy when studies are discoverable by real decision questions, findings retain source and confidence, sensitive evidence has appropriate controls, superseded knowledge remains traceable, and teams can identify both accumulated evidence and meaningful research gaps.