Analyze All Your Research at Once

The scenario

You've completed a round of user research — 5 to 10 interviews, a couple of survey exports, maybe some support ticket data. Each one has useful insights, but no one has time to read them all and synthesize the patterns. Instead of spending a full day pulling themes across documents, you'll process everything at once and get a unified research synthesis in 20 minutes.

Example output

You'll get two things: individual snapshots for each input file, and a unified synthesis across everything.

The unified synthesis looks like this:

Research Synthesis: Enterprise Onboarding Experience

Sources: 7 interviews, 1 survey (n=84), 1 support ticket export Date: Jan 30, 2026

Top Themes (by frequency)

1. Time-to-value is too long (mentioned in 6/7 interviews, 62% of survey)
   • Users expect to see value in the first session
   • Trial users who don't create something in 24 hours rarely come back
   • "I signed up on Monday and by Wednesday I still hadn't figured out what to do first"
2. Team adoption is the real blocker (5/7 interviews)
   • Individual users love the product but can't get their team to switch
   • No team-level features makes it hard to justify the cost
   • "I use it every day but my team doesn't because there's no shared workspace"
3. Reporting and ROI proof (4/7 interviews, 45% of survey)
   • PMs need to show leadership that the tool is worth the cost
   • No built-in metrics or usage reports
   • "My boss asked how much time it's saving us and I had no answer"

Opportunity Prioritization

Opportunity	Evidence Strength	User Impact	Effort Estimate
Guided first-session experience	Strong (6/7)	High	Medium
Team workspace + shared templates	Strong (5/7)	High	High
Time-saved reporting dashboard	Moderate (4/7)	Medium	Medium
Integration with existing tools	Moderate (3/7)	Medium	Low

Contradictions and Tensions

• Power users want more customization; new users want more guidance
• Enterprise buyers want security features; individual users want speed

Recommended Next Steps

1. Prototype a guided first-session flow and test with 3 trial users
2. Scope team workspace — start with shared templates only
3. Build a simple "time saved" calculation into the product

How to create this in 20 minutes

Step 1. Find the input folder. In your file browser, open: discovery → inputs

Step 2. Add your research files to this folder. Drag and drop or copy them in. Supported formats:

• Interview transcripts or notes (.md, .txt)
• Survey exports (.csv)
• Support ticket exports (.csv, .md)
• Any text-based research document

Name your files descriptively — Claude uses the file names to label sources in the synthesis. For example: interview-sarah-chen-enterprise-pm.md is better than notes-1.txt.

Step 3. In the chat panel, type:

/research-synthesis-engine

Step 4. Claude asks what you're trying to learn — the research question or opportunity area you're exploring. For example: "We're trying to understand why enterprise trial users don't convert to paid." If you're not sure, say "find the top themes across all the research."

Step 5. Claude processes each file individually (creating snapshots) and then synthesizes patterns across everything. This takes a minute or two depending on how many files you added. Results save to:

• Individual snapshots + unified synthesis: discovery/outputs/

If something doesn't look right

Tell Claude what to change. For example:

• "The themes section is too broad — break it into more specific clusters"
• "I think you missed insights about pricing from the third interview"
• "The opportunity prioritization doesn't account for our engineering constraints — we can't do anything 'High' effort this quarter"
• "Add a section comparing what enterprise users said vs. SMB users"
• "The evidence strength ratings seem off — Sarah and Mike both mentioned time-to-value, that should be 'Strong'"

Troubleshooting

What's next?

• Build Personas — Turn your synthesis into behavioral personas
• Create a PRD — Use your research findings to write an evidence-based PRD