AI prompts for expert review: scope planning, finding analysis, report generation

AI prompts for expert review

These prompts cover the key stages of an expert review workflow: defining scope and evaluation frameworks, categorizing and prioritizing findings, generating the report, and cross-referencing expert findings with analytics data. Each prompt includes bracketed placeholders — replace them with your project specifics.

Prompt 1: Prepare an expert review scope document

I am preparing to conduct an expert review of [product name], a [product type] targeting [user segments].

Business context: [key goals, known problem areas, recent changes]
Available data: [analytics, prior research, support tickets, if any]
Scope: [full product / specific flows — list them]

Please help me:
1. Draft an evaluation scope document listing the exact flows and screens to review
2. Recommend 2-3 evaluation frameworks appropriate for this product type (e.g., Nielsen's heuristics, WCAG, cognitive laws, domain-specific guidelines)
3. Create a severity rating scale with definitions for cosmetic, minor, major, and critical issues, tailored to this product's context
4. Suggest 5-8 key tasks to walk through as a target user, based on the product type and user segments
5. Identify any areas where analytics data should be checked before the walkthrough to focus the review on high-traffic or high-drop-off areas

Prompt 2: Analyze and categorize review findings

I have completed an expert review and documented the following findings:

[Paste the issue log with: location, description, violated principle, severity, and screenshot references]

Please:
1. Group findings by theme (e.g., navigation, content clarity, form design, error handling, consistency, accessibility)
2. Within each theme, rank findings by severity (critical → cosmetic)
3. Identify patterns — are multiple findings symptoms of the same root cause? If so, flag the root cause
4. Suggest a severity-effort matrix placement for each finding (high/low severity × high/low effort to fix)
5. Highlight the top 5 "quick wins" — high severity issues that are likely low effort to fix
6. Draft a one-paragraph executive summary of the review's overall assessment

Prompt 3: Generate expert review report from findings

I need to write a professional expert review report for [product name]. The audience is [product team / leadership / client].

Here are my structured findings:
[Paste categorized and prioritized findings from Prompt 2]

Strengths I identified: [list of things the product does well]
Context: [product stage, team size, known constraints]

Please generate a report with:
1. Executive summary (3-5 sentences: overall assessment, top findings, recommended immediate actions)
2. Approach section (what was reviewed, which frameworks, how severity was rated)
3. Findings section organized by severity (critical first), each with: issue title, location, description, principle violated, severity, recommendation, annotated screenshot reference
4. Strengths section highlighting what works well and should be preserved
5. Next steps section recommending whether to proceed to usability testing, and which flows to test

Prompt 4: Compare expert review findings with analytics data

I have expert review findings for [product name] and access to analytics data for the same flows.

Expert review findings (by flow):
[Paste findings organized by user flow]

Analytics data:
[Paste relevant metrics: drop-off rates, task completion rates, error rates, session recordings summary, or heatmap observations]

Please:
1. For each expert-identified issue, check whether the analytics data supports or contradicts its severity rating
2. Identify any analytics anomalies (high drop-off, rage clicks, unexpected navigation patterns) that the expert review did not flag
3. Produce a validated priority list that combines expert judgment with behavioral evidence
4. Recommend which findings should be tested with real users before acting on them