Synthesize findings across multiple research sessions into one clear narrative
Ten sets of notes aren't a finding. They're raw material for one.
- Use it for
- Researchers and PMs with notes from several interviews or sessions who need one synthesis.
Research synthesis often stalls at compiling — a document listing what each session or interview said, organized by participant instead of by insight. That's useful raw material and a bad final deliverable: nobody reading ten interview summaries back-to-back can extract the actual pattern themselves, which is exactly the work synthesis is supposed to have already done. Real synthesis finds the themes that recur across sessions and states them as findings, with the individual sessions as supporting evidence, not the main content.
This prompt does that cross-session pass explicitly rather than compiling notes session by session.
When not to use this
If you only have one or two sessions, there isn't enough data yet for real cross-session synthesis — a single session summary is the honest output at that stage. This is for when you have enough sessions that real patterns should be emerging.
Did this work?
You end up with a small number of clear, evidence-backed themes that someone could act on — not a long list restating what each individual session said.
Tested on claude-opus-5. Evidence status is draft; it moves to battle-tested only on recorded runs, never by hand.