Skip to main content

UX Research Synthesis — Templates & Reference

Reference material for ux-research-synthesize skill. This file contains format specifications, templates, and checklists extracted from SKILL.md for token optimization.

Theme Card Structure​

Use this structure for each theme in the Key Findings section (ordered by Priority Score descending):

### {Priority Badge} {Interpretive Theme Name}

**Insight**: {WHY this matters — business impact, user need, strategic implication}

**Evidence**:
- "{Quote 1}" — P{ID}, {context}, {source tag}
- "{Quote 2}" — P{ID}, {context}, {source tag}
- "{Quote 3}" — P{ID}, {context}, {source tag}

**Participant Coverage**: {n} participants ({participant_ids})

| Metric | Value |
|--------|-------|
| Severity | {Nielsen 0-4} ({frequency}% x {impact description} x {persistence: one-time/repeated}) |
| Confidence | {High/Medium/Low} |
| Validation | {Corroborated/Single-source} |

Priority Badges​

BadgeCriteria
Red CRITICALSeverity 4, High Frequency
Orange HIGHSeverity 3+, Medium+ Frequency
Yellow MEDIUMSeverity 2, any Frequency
Green LOWSeverity 0-1

Recommendation Format (Problem-Solution)​

Use this structure for each theme with actionable recommendation:

### {Priority Badge} {Recommendation Title}

**INSIGHT**: {User problem discovered from theme}

**SO WHAT**: {Why this matters to business — ROI, user impact, strategic value}

**NOW WHAT**: {Specific, implementable action — 1-2 sentences max}

**SUCCESS METRIC**: {How to measure improvement — concrete KPI}

| Metric | Value |
|--------|-------|
| Priority | {Must-have / Need / Nice} |
| Effort | {Low / Medium / High} |
| Impact/Effort | {Quick Win / Big Bet / Fill-In / Money Pit} |
| Theme | {Link to theme number} |

Executive Brief Format​

Auto-generated 1-page standalone summary from main report. Maximum ~300-400 words, no methodology details, no raw data, no jargon, standalone (reader needs no other context).

# {Study Name} — Executive Brief

> "{Powerful user quote from most impactful finding}"

## Highlights

- {What's working well — finding 1}
- {What's working well — finding 2}

## Lowlights

1. **{Issue 1 title}** — {Impact description}
2. **{Issue 2 title}** — {Impact description}
3. **{Issue 3 title}** — {Impact description}

## Next Steps

- **{Action 1}** — {Expected impact} — {Timeline: Quick Win/Big Bet}
- **{Action 2}** — {Expected impact} — {Timeline}
- **{Action 3}** — {Expected impact} — {Timeline}

---

| Priority | Count |
|----------|-------|
| Critical | {n} findings |
| High | {n} findings |
| Medium | {n} findings |

Full report: `{path_to_main_file}`

Quality Checklist (Pre-Write Validation)​

Verify every item before preview. If any check fails, revise report before preview.

Executive Summary​

  • Executive Summary <= 1 page and stands alone
  • Opens with powerful user quote (emotionally resonant)
  • Has highlights (what's working) AND lowlights (issues)
  • Has clear next steps (actionable bullets)

Themes​

  • Every theme has interpretive name (not just topic label)
  • Every theme has "Insight" explaining WHY (not just WHAT observed)
  • Every theme has 2-3+ quotes from different participants
  • Every theme has participant coverage noted (IDs listed)
  • Theme count: 3-8 (flag if outside optimal range)

Recommendations​

  • Every recommendation has concrete action (not vague "improve UX")
  • Every recommendation links to specific finding/theme
  • Recommendations include priority + effort estimates
  • Recommendations include Impact/Effort quadrant classification
  • Every recommendation has "INSIGHT/SO WHAT/NOW WHAT" structure
  • Every recommendation has success metric (how to measure)

Evidence & Traceability​

  • All research questions have corresponding findings
  • Participant coverage balanced (no single participant >25% of evidence)
  • Every claim traces to verbatim quote with participant ID
  • Participant IDs are consistent throughout (P1-P{N})
  • Quotes include context (task, timing, source tag)

Methodology & Limitations​

  • Methodology note states analysis approach clearly
  • Limitations stated clearly (sample size, scope, confidence caveats)
  • No hallucinated findings (all claims sourced from data)

Cross-Study Mode Only​

  • Temporal validity checked (when were studies conducted)
  • Methodology consistency noted across studies
  • Contradictory findings across studies are flagged and explained